Davy Tsz Kit Ng · Jac Ka Lok Leung ·

Maggie Jiahong Su · Iris Heung Yue Yim ·

Maggie Shen Qiao · Samuel Kai Wah Chu

AI Literacy
in K-16
Classrooms
AI Literacy in K-16 Classrooms
Davy Tsz Kit Ng • Jac Ka Lok Leung
Maggie Jiahong Su • Iris Heung Yue Yim
Maggie Shen Qiao • Samuel Kai Wah Chu

AI Literacy in K-16
Classrooms
Davy Tsz Kit Ng Jac Ka Lok Leung
Faculty of Education Division of Integrative Systems and Design
University of Hong Kong Hong Kong University of Science and
Hong Kong, Hong Kong Technology
Hong Kong, China
Maggie Jiahong Su
Faculty of Education Iris Heung Yue Yim
University of Hong Kong Faculty of Education
Hong Kong, Hong Kong University of Hong Kong
Hong Kong, Hong Kong
Maggie Shen Qiao
Faculty of Education Samuel Kai Wah Chu
University of Hong Kong Faculty of Education
Hong Kong, Hong Kong University of Hong Kong
Hong Kong, Hong Kong

ISBN 978-3-031-18879-4    ISBN 978-3-031-18880-0 (eBook)

https://doi.org/10.1007/978-3-031-18880-0

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 2022
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Foreword

Dear Readers,
There are many reasons why this book, AI Literacy in K-16 Classrooms, is so
important to educators. First, AI is increasingly being used in schools and class-
rooms, and it is important for educators to understand how AI works and how it can
be used to benefit students. Second, AI is changing the nature of work and learning,
and educators need to be prepared to teach in a world where AI is playing an increas-
ingly important role. Third, AI is raising ethical and social issues that educators
need to be aware of and prepared to discuss with their students.
This book is essential reading for anyone who wants to understand AI and its
implications for education.
AI tools are found in many aspects of our lives, and it is becoming increasingly
difficult to remain ignorant about their implications on society. AI Literacy in K-16
Classrooms provides educators with the much-needed foundation to understand AI,
its capabilities, and its potential implications in the classroom and beyond. Teachers
are using AI tools to:
• Help grade essays
• Provide targeted feedback
• Offer personalized learning experiences
• And even teach classes
Students are using AI tools to:
• Get better grades
• Get more personalized feedback
• Have more customized learning experiences
• And even take classes
With these applications in mind, it is imperative that educators understand how
AI works and how it can be used to benefit students. Teaching AI is essential reading
for anyone who wants to understand AI and its implications for education.

v
vi Foreword

If you want another reason why you should take the information and recommen-
dations of this book to heart, consider this: most of this sentence and most of the
preceding foreword text were written by a computer using an AI tool called GPT-3!
Beginning with a few simple prompting phrases, such as “Artificial Intelligence,”
“There are many reasons why this book is important to educators,” and “Teachers
are using AI tools to….,” OpenAI’s GPT-3 large language model (LLM) generated
almost every word you see above. To arrive at the final text involved only a few
minor edits to include the title of this book, and a few more prompts to suggest some
alternative wordings and lists of AI tool uses.
There is no area of our lives, either as learners or educators, that is untouched by
AI. Each and every subject, from art to zoology, now has a growing number of AI
tools that generate new content or analyze and model the world around us. The era
of ubiquitous AI has only just begun, making AI Literacy in K-16 Education a must-
read for anyone who cares about the success of our education programs in preparing
our students for the future.

Sincerely,
Prof. J. Stephen Downie
School of Information Science
University of Illinois, Urbana-Champaign
and
GPT-3
OpenAI Playground
https://beta.openai.com/playground
Foreword 2

Artificial intelligence (AI) technologies make our lives more efficient every day.
These emerging technologies power many applications, programs, and services that
help us do everyday things such as improving our online experience, enhancing our
security, and increasing efficiency of our routine jobs. In the meantime, AI can
cause us a lot of trouble due to privacy issues and human bias. Our students must
learn about this emerging technology; hereby, AI education should be provided for
all. This is supported by many AI education projects that have been launched around
the world because AI topics conventionally are included in engineering faculty in
higher education.
AI literacy is one of the important areas in educational research and practice. As
one of the first AI education researchers, I am so happy to see this book AI Literacy
in K-16 Education published. Professor Sam Chu and his team have expertise in
literacy, technology education, and twenty-first century skills. In this book, they
reviewed literature on AI education from kindergarten to tertiary level. They first
define what AI literacy is, and discuss the landscape of AI education, followed by
using various frameworks to describe AI literacy development. They continue to
discuss instructional designs for the development at different educational levels.
Finally, they raise this issue from teacher perspectives.
I believe this book would cater to different types of readers. If you are educators,
you will gain more understanding of how to teach AI and assess student learning. If
you are policy makers or curriculum designers, you will get inspired to design
teacher professional programs or curriculum guidance. If you are AI developers,
you will better understand what educational stakeholders need for developing AI
content or learning applications. Finally, enjoy this book, and nurture our new gen-
erations with good AI literacy.

Thomas Kin Fung CHIU

Assistant Professor, Department of Curriculum and Instruction
The Chinese University of Hong Kong
Hong Kong, Hong Kong

vii
About the Book

The book presents a review of the frameworks, content, pedagogies, and assessment
of AI literacy education in supporting policymakers, educators, and parents to
embed such an emerging area in the K-16 curriculum and into classrooms.
Recommendations were proposed on how to develop AI literacy curricula and uti-
lize age-appropriate technological tools and pedagogies to foster students’ AI liter-
acy skills or digital competencies. The book aims to provide an exhaustive summary
of current evidence related to AI literacy with some highlighted cases. Some empiri-
cal studies were selected to illustrate how AI literacy models were applied in differ-
ent countries and regions. The book captures the attention of multidisciplinary
researchers looking for an overview of empirical studies that call for an AI literacy
instructional design. The significance also lies in for all educators as a reference for
practical methods. Finally, policymakers may borrow the models elicited in this
book to reform education policies so as to design future-proof curricula and, most
importantly, to prepare students with knowledge, mindsets, and dispositions condu-
cive to dealing with future societal challenges.

ix
Contents

Part I Conceptualizing AI Literacy

1 Introduction����������������������������������������������������������������������������������������������    3
1.1 Key Inquiry Questions����������������������������������������������������������������������    4
1.2 Organization��������������������������������������������������������������������������������������    5
References��������������������������������������������������������������������������������������������������    7
2  Education and AI Literacy����������������������������������������������������������������    9
AI
2.1 A Historical Perspective��������������������������������������������������������������������   10
2.2 AI Education Versus AI Literacy������������������������������������������������������   11
2.2.1 Artificial Intelligence in Education��������������������������������������   11
2.2.2 Artificial Intelligence Literacy����������������������������������������������   13
2.2.3 Similarities and Differences Between AIED
and AI Literacy ��������������������������������������������������������������������   14
2.3 Conclusion����������������������������������������������������������������������������������������   15
References��������������������������������������������������������������������������������������������������   17
3 AI Literacy for All������������������������������������������������������������������������������������   21
3.1 AI for Living, Workplace, Learning, and Societal Good������������������   21
3.1.1 AI for Living ������������������������������������������������������������������������   22
3.1.2 AI for Workplaces ����������������������������������������������������������������   22
3.1.3 AI for Learning ��������������������������������������������������������������������   23
3.1.4 AI for Societal Good������������������������������������������������������������   24
3.2 Benefits of AI Literacy for Different Educational Levels ����������������   24
3.2.1 Kindergarteners��������������������������������������������������������������������   24
3.2.2 Primary and Secondary Education����������������������������������������   25
3.2.3 Noncomputer Science University Students��������������������������   26
3.3 Conclusion����������������������������������������������������������������������������������������   27
References��������������������������������������������������������������������������������������������������   27
4 
The Landscape of AI Literacy����������������������������������������������������������������   31
4.1 AI Literacy as a Twenty-First Century Skill ������������������������������������   31
4.2 Emerging Frameworks for AI Literacy Education����������������������������   32

xi
xii Contents

4.2.1 Competencies and Design Considerations����������������������������   33

4.2.2 The Five “Big Ideas” About AI��������������������������������������������   33
4.2.3 Other Review Papers in AI Literacy Education��������������������   36
4.3 Rising Publications on AI Literacy Education����������������������������������   37
4.3.1 AI Literacy Education for K–12 Students����������������������������   39
4.3.2 AI Literacy Education for Noncomputer Science
University Students ��������������������������������������������������������������   40
4.4 New Education Policies on AI Literacy Across the Globe ��������������   41
4.5 Our Three Proposals of AI Literacy Educational Frameworks
(Bloom’s, TPACK, P21)��������������������������������������������������������������������   43
4.5.1 AI Literacy and Bloom’s Taxonomy������������������������������������   43
4.5.2 AI Literacy and TPACK Framework������������������������������������   46
4.5.3 AI Literacy and P21’s Framework for the 21st Century
Learning��������������������������������������������������������������������������������   51
4.6 Conclusion����������������������������������������������������������������������������������������   55
References��������������������������������������������������������������������������������������������������   56

Part II K-16 AI Literacy Education

5  Literacy Education in Early Childhood Education������������������������   63
AI
5.1 Introduction��������������������������������������������������������������������������������������   64
5.2 Methods��������������������������������������������������������������������������������������������   64
5.3 Results and Discussion ��������������������������������������������������������������������   66
5.4 Conclusion����������������������������������������������������������������������������������������   71
References��������������������������������������������������������������������������������������������������   73
6  Literacy Education in Primary Schools ������������������������������������������   75
AI
6.1 Method����������������������������������������������������������������������������������������������   76
6.2 Results and Discussion ��������������������������������������������������������������������   77
6.3 Conclusions��������������������������������������������������������������������������������������   83
References��������������������������������������������������������������������������������������������������   84
7  Literacy Education in Secondary Schools����������������������������������������   87
AI
7.1 Method����������������������������������������������������������������������������������������������   88
7.2 Results and Discussion ��������������������������������������������������������������������   88
7.3 Conclusions��������������������������������������������������������������������������������������   95
References��������������������������������������������������������������������������������������������������   96
8  Literacy Education for Nonengineering Undergraduates��������������   99
AI
8.1 Methodology ������������������������������������������������������������������������������������ 100
8.1.1 Data Collection �������������������������������������������������������������������� 100
8.1.2 Data Analysis������������������������������������������������������������������������ 102
8.2 Results and Discussion �������������������������������������������������������������������� 102
8.3 Conclusion���������������������������������������������������������������������������������������� 114
References�������������������������������������������������������������������������������������������������� 115
Contents xiii

Part III AI Literacy for Instructional Designers

9  Literacy on Human-Centered Considerations��������������������������������
AI 119
9.1 Overview������������������������������������������������������������������������������������������ 120
9.2 Needs of HCAI in Educational Fields���������������������������������������������� 120
9.3 Key Elements of Human-Centered Considerations�������������������������� 121
9.3.1 Human Factor Designs and Values �������������������������������������� 121
9.3.2 Reflect Human Intelligence�������������������������������������������������� 121
9.3.3 Ethical and Responsible Design������������������������������������������� 122
9.3.4 AI Under Human Control and Under Human
Conditions ���������������������������������������������������������������������������� 123
9.4 Scaffolding Support�������������������������������������������������������������������������� 125
9.4.1 Knowing Learners’ Backgrounds ���������������������������������������� 125
9.4.2 Knowing Learners’ Interests and Motivation ���������������������� 126
9.4.3 Knowing Students’ Learning Progress �������������������������������� 127
9.4.4 Parental Involvement������������������������������������������������������������ 127
References�������������������������������������������������������������������������������������������������� 128
10  Literacy from Educators’ Perspectives��������������������������������������������
AI 131
10.1 Understanding Teachers’ AI Digital Competencies������������������������ 131
10.2 Essential AI Digital Competencies for Educators (P21)���������������� 133
10.3 Conclusion�������������������������������������������������������������������������������������� 137
References�������������������������������������������������������������������������������������������������� 137
11 Summary and Conclusions��������������������������������������������������������������������� 141
11.1 For Teachers������������������������������������������������������������������������������������ 143
11.2 For Higher Education Faculty�������������������������������������������������������� 144
11.3 For Policymakers���������������������������������������������������������������������������� 145
11.4 Parents�������������������������������������������������������������������������������������������� 146
11.5 Researchers and Developers ���������������������������������������������������������� 147
11.6 What Is Next? �������������������������������������������������������������������������������� 148
References�������������������������������������������������������������������������������������������������� 149
 Part I
Conceptualizing AI Literacy
Chapter 1
Introduction

Artificial intelligence (AI) is transforming across industries (e.g., marketing, sci-

ence, art, education, entertainment) to facilitate people’s living, learning, and work-
ing. According to the World Economic Forum’s Future of Jobs Report (2020), 85
million jobs will be displaced by 2025, and, at the same time, 97 million new jobs
will appear. People are accessing various types of AI-powered technologies daily
(e.g., smart home appliances, smartphones, social media, chatbots). However, sel-
dom do they know about the technologies or the algorithms behind, nor are they
aware of potential ethical issues related to AI (e.g., privacy, surveillance, bias). As
more age-appropriate curricula, resources, and tools are made available to learners,
schools, and educators, AI literacy has emerged as a new digital competence that
everyone should learn in response to this new era of intelligence. Different digital
competence frameworks for educators (e.g., the ISTE, DigCompEdu standards)
have included AI to update the latest educational standards to direct toward educa-
tors at all levels of education from early childhood to higher education (ISTE, 2022;
Riina et al., 2022). With the exponential pace of technological advancement, AI lit-
eracy will soon become one of the most important twenty-first century technological
literacy that everyone should learn in our digitizing world (Ng et al., 2021a, b).
This book conceptualizes the term AI literacy, illustrates the proposed concep-
tual frameworks of AI literacy education, explores how AI literacy is taught across
countries/regions, and discusses a range of pedagogical approaches, content, and
technological and assessment tools that leverage K–16 students’ understandings
and motivational capacity. This book is written for researchers, educators, policy-
makers, parents, and AI professionals who wish to foster an evolving set of AI skills
and knowledge for their target learners. The goal is to prepare learners for an
uncharted future with drastic technological changes. Readers will find a spectrum of
theoretical and pragmatic discussions on AI literacy education. The discussions will
be covered by K–16 educational levels as well as stories from different parts of the
world. The book presents a review of AI literacy education which includes practical
recommendations to develop age-appropriate AI literacy curricula, technological

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 3

D. T. K. Ng et al., AI Literacy in K-16 Classrooms,
https://doi.org/10.1007/978-3-031-18880-0_1
4 1 Introduction

tools, pedagogies, and assessments. In Part III, a framework of AI literacy is devel-

oped based on the P21’s Framework for the 21st Century Learning (2009) that not
only focuses on technical skills but also details on how AI technologies can be used
to enhance and innovate education and training.

1.1 Key Inquiry Questions

AI education first started in university computer science education which required

advanced programming competencies that were not designed for K–12 and non-­
computer science students. Over the last two decades, most of the learners who
would study or required to take AI courses are from computer science backgrounds
who design robotics and software, construct models, and handle data structures
(McCarthy, 2007; Wong et al., 2020). As might be expected, educators face chal-
lenges in coming up with a sound approach in scaffolding K–12 children and non-­
computer science undergraduates with these technical concepts and skills in the
early 2010s. However, with the breakthroughs in AI learning technologies and sub-
stantial demand for AI skills across industries, teaching AI is no longer as difficult
as it was in the past. Furthermore, the increase in accessibility to a wide range of AI
technologies, such as data analytics tools, intelligent agents, chatbots, and writing
assistants, also make sense for learners to understand, use, and communicate with
AI to facilitate their living, learning, and working. However, in schools, teachers
and students are novel to these subjects, and they may not be ready to adapt to such
digital transformation. To overcome these challenges, there is an urgent need to sup-
port capacity building for all teachers and students to leverage digital technologies
across the world.
Nowadays, university faculty have designed learning programs to equip non-­
computer science undergraduates with AI digital competencies for future workforce
(Kong et al., 2021). In K–12 education, the current uses and impact of AI have
spurred a demand to create a global consensus on how to equip young learners with
the competencies needed to understand the technology and its related ethical dilem-
mas (UNESCO, 2022). According to the UNESCO (2022)’s report, 11 countries
have developed and endorsed official curricula to grow up the next generation to
become responsible citizens for success in today’s digital world. Overall, all K–16
learners need to be equipped with a level of AI competencies. These include knowl-
edge, skills, values, and attitudes to become so-called AI literate, which is becoming
a new learning standard of our century.
On this note, this book introduces an exploratory review that helps conceptualize
the newly emerging concept “AI literacy,” in search for a sound theoretical founda-
tion to define, teach, and evaluate AI literacy. Grounded in literature on existing
peer-reviewed articles in renowned databases including Web of Science and Scopus,
this book proposed a set of important concepts to foster learners AI literacy based
on the adaptation of classic literacies based on content and systematic analysis. This
book sheds light on the consolidated definition, teaching, and ethical concerns on AI
1.2 Organization 5

literacy, establishing the groundwork for future research such as competency devel-
opment and assessment criteria on AI literacy.
While the development of students’ AI literacy skills has become a potential
technological skill as part of their education similar to digital literacy (which will be
further discussed in later chapters), researchers have directed their efforts in exam-
ining different learning artifacts and innovative pedagogical approaches that may
facilitate the acquisition of these knowledge and skills among students (Chai et al.,
2020; Wong et al., 2020). All in all, a few fundamental questions will be addressed
in this book:
• How do researchers define the term “AI literacy”?
• How do educators help learners develop AI literacy in aspects of learning arti-
facts, pedagogical approaches, and subject matters?
• How do researchers evaluate students’ AI literacy skills?
• What are the ethical and human-centered considerations in the domain of AI
literacy?
• How should we prepare teachers to foster students’ AI competency?

1.2 Organization

The organization of this book was inspired by the authors’ experiences in imple-
menting AI literacy education. Many schools may have initialized the incorporation
of AI into their curriculum and classrooms already; however, we find that many
educators feel anxious that they are not well prepared for the implementation of AI
curricula as it is a relatively new form of education. In fact, the first author has heard
technology companies developing AI-based products and teaching tools too abruptly
without considering teachers’ and students’ needs. This opens up risks of wasting
manpower and resources if these products are not based on pedagogical grounds
and justified theoretical frameworks. Therefore, the book is structured in a way that
readers will first understand the importance of AI literacy and then be introduced to
guidelines and methods that can help students to develop such AI skills in a more
gradual, humanistic, and systematic manner.
The book is divided into three parts. Part I focuses on a higher conceptual level
of AI literacy, putting forward notions that try to encapsulate a variety of education
contexts, foregrounding the emergence and urgency for educators and learners to
embrace the era of AI. Part II travels deep in reviewing the efforts by scholars around
the round across different educational levels, presenting scoping reviews that guide
K–16 educators to develop their pedagogy, content, technology and assessments for
their students. Part III proposes our vision of AI literacy. We have identified digital
competences that AI developers and educators should take into account when devis-
ing learning tools, platforms, lesson plans, and assignments in order to help rein-
force students’ acquisition of AI literacy skills. Our suggestions on how to provide
effective scaffolding support to their students is provided. Overall, this book aims to
6 1 Introduction

serve as an introduction and reference of resources for our educators, colleagues,

and peers to update the evolution of this area, such that we can take part in designing
effective and exciting learning experiences for our learners.
Part I (Chaps. 2, 3, and 4): Conceptualizing AI Literacy
The development of the concepts on AI literacy education is introduced and dis-
cussed. To begin with, readers may feel confused about the differences and similari-
ties between AI literacy and AI education (AIED). Chapter 2 first has a quick glance
at historical perspectives of AI and further looks into conceptual differences between
the two jargons. Chapter 3 discusses reasons why all learners from kindergarten to
higher education should learn AI. Chapter 4 examines how AI literacy is adapted
from other digital literacies and presents an overview of AI literacy, helping readers
to understand what is entailed in it and what we learn about it through existing
research. An analysis on curriculum around the world is conducted to let readers
appreciate the efforts by current educators that helped students acquire these
literacies.
Part II (Chaps. 5, 6, 7, and 8): AI Literacy in K–16 Education
Readers will be introduced to the existing research findings about the pedagogy,
content knowledge, learning artifacts, and assessment for K–16 students in associa-
tion to the development of twenty-first century skills. An exploratory review sum-
mary on the AI literacy education from K–12 to higher education will be provided.
Chapters 4, 5, 6, 7, and 8 use content and thematic analysis to understand how AI
literacy education is implemented across different levels of study. A range of teach-
ing strategies are recommended to foster students’ acquisition of AI literacy skills.
Part III (Chaps. 9, 10, and 11): AI Literacy for Instructional Designers
In the last part of this book, we provide guidelines to developers, researchers, and
education practitioners who plan to develop AI literacy instructional designs for
students. Chapter 9 addresses the importance of human-centered considerations for
AI developers to design tools and platforms that place users at the center. Chapter
10 taps into how teachers can be trained up with necessary skills, knowledge, and
techniques for AI literacy education. The chapter will also propose a set of twenty-­
first century skills standards for educators to devise lesson plans, activities, and
assignments to help reinforce students’ acquisition of AI literacy skills. Moreover,
teachers will find helpful recommendations on how to provide proper and effective
scaffolding support to their students. Finally, the concluding Chap. 11 links back to
the basic premises we set up in this introduction to add AI literacy as part of the
twenty-first century skills. This aims to lay out guidelines and suggestions for AI
literacy implementation, using the P21’s Framework for the 21st Century
Learning (2009).
References 7

References

Chai, C. S., Rahmawati, Y., & Jong, M. S. Y. (2020). Indonesian science, mathematics, and
engineering preservice teachers’ experiences in STEM-TPACK design-based learning.
Sustainability, 12(21), 9050.
ISTE. (2022). Artificial intelligence in education. Putting educators and students in the driver’s
Seat. Retrieved from https://www.iste.org/areas-­of-­focus/AI-­in-­education
Kong, S. C., Cheung, W. M. Y., & Zhang, G. (2021). Evaluation of an artificial intelligence literacy
course for university students with diverse study backgrounds. Computers and Education:
Artificial Intelligence, 2, 100026.
McCarthy, J. (2007). From here to human-level AI. Artificial Intelligence, 171(18), 1174–1182.
Ng, D. T. K., Leung, J. K. L., Chu, S. K. W., & Qiao, M. S. (2021a). AI literacy: Definition, teach-
ing, evaluation and ethical issues. Proceedings of the Association for Information Science and
Technology, 58(1), 504–509.
Ng, D. T. K., Leung, J. K. L., Chu, S. K. W., & Qiao, M. S. (2021b). Conceptualizing AI literacy:
An exploratory review. Computers and Education: Artificial Intelligence, 2, 100041.
Riina, V., Stefano, K., & Yves, P. (2022). DigComp 2.2: The Digital Competence Framework for
Citizens – With new examples of knowledge, skills and attitudes. Retrieved from https://publica-
tions.jrc.ec.europa.eu/repository/handle/JRC128415
Silva, E. (2009). Measuring skills for 21st-century learning. Phi delta kappan, 90(9), 630–634.
UNESCO. (2022). K-12 AI Curricula: A mapping of government-endorsed AI curricula. Retrieved
from https://unesdoc.unesco.org/ark:/48223/pf0000380602
Wong, G. K., Ma, X., Dillenbourg, P., & Huan, J. (2020). Broadening artificial intelligence educa-
tion in K-12: Where to start? ACM Inroads, 11(1), 20–29.
World Economic Forum. (2020). The Future of Jobs Report 2020. Retrieved from https://www3.
weforum.org/docs/WEF_Future_of_Jobs_2020.pdf
Chapter 2
AI Education and AI Literacy

Our twenty-first century is characterized by its rapid technological advancement.

Our lifestyle and ways of interacting with people have changed significantly com-
pared to around a decade ago in the early 2010s as AI technologies turn ubiquitous
across industries and in our everyday lives. Artificial intelligence has spread across
industries to enhance our living, learning, and working experience with exciting
technological innovations such as computer vision, natural language processing,
robotics and motion, machine and deep learning, and neural networks (Chen et al.,
2022; Dong et al., 2021; Zawacki-Richter et al., 2019). Applications of AI have
become in many parts of our everyday life (e.g., smart home appliances, smart-
phones, chatbots, search engines). In the field of education, schools began to use
AI-enabled technologies to leverage students’ personalized learning and reduce
teachers’ administrative work, thus offering more learning support and interactive
learner experience (Roll & Wylie, 2016). Therefore, a field has gradually taken
shape over the last few decades – AI in education (AIED).
Vast majority of the public acknowledges the existence of AI services and
devices, but seldom do students know about the concepts and technology behind or
aware of potential ethical issues related to AI (Burgsteiner et al., 2016; Ghallab,
2019). In recent years, AI has become essential skills to play critical roles across
disciplines and industries (Ng et al., 2021; Touretzky et al., 2019). Students need to
learn how to use AI technologies wisely, as well as to discriminate between ethical
and unethical actions (Rodríguez-García et al., 2021). This gives rise to an emerging
term “AI literacy” that drives the need that everyone should learn the basic AI
knowledge and skills behind these technologies. However, many may feel confused
about the similarities and differences between AI education and AI literacy. Before
looking into the two jargons, this chapter first has a quick glance at what AI is from
a historical perspective and further looks deeper into the conceptual differences
between AIED and AI literacy education.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 9

D. T. K. Ng et al., AI Literacy in K-16 Classrooms,
https://doi.org/10.1007/978-3-031-18880-0_2
10 2 AI Education and AI Literacy

2.1 A Historical Perspective

Artificial intelligence was first introduced as “the science and engineering of mak-
ing intelligent machines in 1956” (McCarthy, 2007, p. 2). At that time, Newell and
Simon (1956) invented a “thinking machine” that was considered as the first com-
puter program that simulated human intelligence to solve complex problems. This
idea advanced our understanding of how humans think and make basic contribu-
tions to artificial intelligence and the psychology of human cognition. At the begin-
ning, AI was used to handle limited tasks (e.g., automation, chess playing).
Researchers called it the rule-based AI that people use programs and algorithms to
reason and solve problems based on predetermined rules and environment
(McCarthy, 2007). In today’s era of rapid technological advancement and exponen-
tial increases in large datasets known as “big data,” AI has transitioned to everyday
applications on an unprecedented scale. The large datasets in near real-time enables
people to drive autonomously, view videos and media posts based on recommenda-
tions (e.g., Netflix, YouTube, Facebook), do shopping online according to AI-based
advertisements, and detect frauds to enhance working efficiency. We can see that AI
has been broadened to perform a wide range of complex tasks with the exciting
innovations (Zawacki-Richter et al., 2019).
Artificial intelligence refers to computerized machines and systems that mimic
human intelligence to facilitate people to conduct various tasks and solve complex
problems (Wang, 2019). However, there are quite a few ambiguous buzzwords
about subfields of AI such as machine learning, deep learning, and neural networks
which sometimes may be confusing. Here we provide some basic definitions to
describe the relationship of these buzzwords to AI. First, machine learning (ML)
exhibits the experiential “learning” associated with human intelligence which has
the capacity to learn and improve its analyses through computational algorithms
(Helm et al., 2020). These algorithms use large datasets to recognize patterns and
effectively “learn” in order to train the machines and models to conduct various
tasks (e.g., make autonomous recommendations, decisions). After sufficient repeti-
tions and algorithm refinement, the machine becomes more ready for people to
input datasets to predict an outcome. Throughout the processes, people can compare
the outcomes with a set of desired outcomes in order to judge the accuracy of the
algorithm to iteratively perfect the AI’s capability to predict future outcomes (Helm
et al., 2020).
Deep learning and neural networks are more complex versions of these models
that make use of hierarchical layers to generate the final outputs. The network first
begins with an input layer that then progressly involves a number of “hidden layers”
that responds to different features (e.g., age, gender, ethnicity). These middle layers
allow the AI-empowered model to enhance its understanding as the inputs ascend
“deeper,” without explicitly programmed directions. The model conducts specific
actions on multiple layers and further successively improves its accuracy as new
data is available. Since the model is similar to the way the human brain functions,
the model is called “neural networks” and thus gives rise to a new form of AI,
known as “deep learning.”
2.2 AI Education Versus AI Literacy 11

In the educational field, AI could automate administrative processes, enhance

learning experiences, and facilitate students’ interactions. Using artificial intelli-
gence in education (AIED) has emerged as a research field since the 1980s, as
marked by the first publication of the International Journal of Artificial Intelligence
in Education in 1989 and the formation of the International AI in Education Society
in 1993 (Pinkwart, 2016). Researchers and educators have contributed to AI tech-
nologies in educational developments to design AI-driven tools for learning such as
intelligent tutoring systems, recommendation systems, computational linguistics,
and intelligent agents and use AI to understand, assess, and improve students’ learn-
ing (Alkhatlan & Kalita, 2018; Chen et al., 2020; Hwang et al., 2020; Williamson &
Eynon, 2020). AIED applications have been employed to facilitate teaching, learn-
ing, and administration through data mining, learning analytics, and learning intel-
ligent agents in educational contexts (Chen et al., 2020; Hwang et al., 2020).
Since then, AI has affected many facets of human life rather than merely com-
puter industries, which drive the need that everyone should learn AI (Ng et al.,
2021). Over the past two decades, AI was conventionally taught at tertiary level as
a field of computer science education (Chiu, 2021). Teaching AI stemmed from the
invention of some programming languages (e.g., BASIC, LOGO, Prolog) by
Seymour Papert in the 1980s. In 1995, Stuart Russell and Peter Norvig published an
authoritative AI textbook that was widely used for over 1500 schools across the
world. The book listed five major categories of AI concepts that were taught at the
undergraduate level: (1) problem-solving strategies, (2) knowledge, reasoning and
planning, (3) uncertain knowledge and reasoning, (4) machine learning, and (5)
communication, perceiving, and acting. There were few mentions in literature
where children were encouraged to conduct AI projects in playful environments to
learn computational concepts (e.g., variables, recursion, representations, process)
(Kahn & Carlsson, 1985). However, without age-appropriate curricula and teaching
tools, educators faced difficulties in scaffolding young learners to visualize the
complex AI concepts (Wong et al., 2020; Ng et al., 2021). Introducing K–12 stu-
dents to AI has been impossible and sporadic. Until 2019, the emergence of devel-
opmentally appropriate curriculum and technologies has made AI teaching possible
for younger learners to develop their AI competencies to become a responsible digi-
tal citizen and get ready for the future workplace. Since then, scholars in K–16
education discuss how to incorporate AI into their classrooms and started to design
frameworks, measurements, models, and reviews in K–16 AI literacy education
(Druga et al., 2019; Long & Magerko, 2020; Ng et al., 2021; Touretzky et al., 2019).

2.2 AI Education Versus AI Literacy

2.2.1 Artificial Intelligence in Education

Artificial intelligence in education (AIED) refers to the use of AI technologies and

application programs which serve as intelligent tutors, tutees, tools/ partners, and
policymaking advisors in educational settings to facilitate teaching, learning, and
12 2 AI Education and AI Literacy

decision making (Hwang et al., 2020). These tools simulate human intelligence to
“make inferences, judgments, or predictions, computer systems can provide person-
alized guidance, support, or feedback to students” (Hwang et al., 2020, p. 1). Research
on AIED involves diverse research focused on different AI technologies. For exam-
ple, Hwang et al. (2020) used fuzzy expert systems to take into account both the
affective and cognitive status of students so as to improve their learning achievement
and reduce their anxiety in a fifth-grade mathematics course. Chih-­Ming and Ying-
You (2020) developed a computer-mediated communication competence forecasting
model to predict communication behavior and objectives during collaborative prob-
lem-based learning. Baker et al. (2021) analyzed how learning analytics is being
methodologically influenced by recent trends in the fields of educational data mining,
quantitative ethnography, and learning at scale to increase impact on policy and prac-
tice (Baker et al., 2021). Okonkwo and Ade-Ibijola (2021) investigated the benefits
and challenges of educational chatbots, as well as future research areas such as pro-
viding personalized services for institutional employees and students.
With the increasing amount of AIED research articles being published, it is nec-
essary to systematically review the relevant issues. Hinojo-Lucena et al. (2019) con-
ducted a bibliometric analysis of 132 articles on AI in higher education to investigate
the relationship between the number of authors and papers and to explore the main
source titles, organizations, authors, and countries about AI in higher education.
Roll and Wylie (2016) reviewed 47 articles from 1994, 2004, and 2014 in the jour-
nal of AIED to explore the research foci in the field of AIED. However, Chen et al.
(2020) believed that findings may not provide an essential understanding of the
research since only a single article is considered and there are no studies reviewed
after 2014. Zawacki-Richter et al. (2019) analyzed 146 articles regarding the AIED
applications in higher education and figured out 4 major areas of AIED applica-
tions: profiling and prediction, intelligent tutoring systems, assessment and evalua-
tion, and adaptive learning systems. Although bibliometric analysis is principally
useful in analyzing sizable literature data, an in-depth investigation by using a sys-
tematic review methodology is needed (Chen et al., 2020).
Recent reviews contributed to look at the major research issues and AI technolo-
gies adopted in highly cited AIED studies and proposed a theoretical framework for
this field. For example, Chen et al. (2020) reviewed 30 articles from Web of Science
and Google Scholar with keywords “AI” and “Education” and identified that AI has
taken the form of digital technologies such as automated assessment, adaptive learn-
ing, humanoid robots, and web-based chatbots to perform teachers’ duties, facilitate
administrative functions, and review students’ assignments effectively to achieve
higher quality in education. Some systems used AI technologies such as machine
learning and natural language processing (NLP) to personalize students’ needs to
foster learning engagement and improve students’ performance (Chen et al., 2020).
Chen et al. (2020) conducted a systematic review using 45 highly cited AIED arti-
cles indexed in Web of Science and Scopus databases from 1990 to 2016 to identify
the application and theory gaps during the rise of AI in education. It is indicated that
there was a rising interest in and impact of AIED research over the past two decades.
It is found that deep learning technologies (e.g., neural networks, NLP, biomedical
detection) have great potential to be used in the educational fields for future studies
2.2 AI Education Versus AI Literacy 13

and employ these AI technologies and engage deeply with educational theories. All
these reviews emphasize on how educators use AI technologies to enhance learning
and teaching and facilitate administration. These AIED research reviews focus on
using AI technologies for educational purposes, and none of the reviews discovered
how to teach and learn AI and related computer science concepts. All of the reviews
exclude computer science, AI and machine learning related courses learning, learn-
ing technologies to learn AI and teaching methods, and computational thinking
related studies (Chen et al., 2020).

2.2.2 Artificial Intelligence Literacy

Artificial intelligence literacy studies encompass efforts toward equipping learners

with literacy skills to thrive in an AI-saturated future (Wang & Cheng, 2021). In
earlier years of publication, AI learning was only instituted in computer science
education in higher education which required advanced programming competen-
cies, while it was not at an appropriate level for K–12 learners. Many pioneers in
higher education gravitated toward the use of robotics making (Klassner, 2002;
Imberman, 2004), software development (Mota-Valtierra et al., 2019; Estevez et al.,
2019), and game design through syntax-based programming (Wallace et al., 2010;
Wong et al., 2020). These pedagogies have demonstrated the potential to improve
computer science students’ AI learning motivation (Wallace et al., 2010) and knowl-
edge acquisition (McGovern & Fager, 2007). However, K–12 educators faced chal-
lenges in scaffolding students to understand AI concepts due to technical complexity
which make teaching AI difficult to become part of K–12 education (e.g., McCarthy,
2007; Steinbauer et al., 2021; Wong et al., 2020). With the emergence of age-appro-
priate technologies such as Blockly-based programming (Estevez et al., 2019),
simulation, and robotics (Narahara & Kobayashi, 2018), AI becomes possible for
learners to program applications that involve AI elements without mastering the
complex syntax of programming languages and algorithms. This raises the learning
possibilities to consider AI education as part of ICT and STEM education curriculum.
With the breakthroughs of emerging AI learning artifacts, teaching AI is not as
difficult as it was in the past. Educators began engaging non-computer science
undergraduates, as well as primary and secondary school students in AI learning
(Chiu, 2021; Xia et al., 2022). Studies have shown a positive interplay between arti-
ficial and human intelligence that nontechnical students could also gain interests and
fundamental understandings of AI in interdisciplinary and liberal arts curricula
(Cicirello, 2008; Lichtenthaler, 2018). In primary and secondary education, research-
ers have drawn attention to age-appropriate technological tools to scaffold students’
AI understanding with well-designed Blockly programming interfaces such as
Scratch and simple machine learning model builders that could allow learners to
access the speech and object recognition (Garcia et al., 2019; Toivonen et al., 2020).
These advances create opportunities for educators to adopt effective pedagogical
approaches with these tools to foster students’ AI literacy and appreciation of the five
14 2 AI Education and AI Literacy

“big ideas” of AI (Touretzky et al., 2019). As such, all students should receive a solid
foundation for their studies at their young ages to prepare their future career in the
era of AI, thereby encouraging nontechnical university students, as well as primary
and secondary students to learn how to use AI and its underlying concepts and related
AI-enhanced soft skills including teamwork, organization, social awareness, and
ethical concerns (e.g., Carpio Cañada et al., 2015; Sabuncuoglu, 2020).
Recent AI-related technologies such as smart devices, search engines, chatbots,
and computer games have become common in our daily life. Most people know
about the existence of these services and products, but only a few understand the
technology and principles behind them (Ng et al., 2021). With more age-appropriate
technologies, scholars began to focus on the learning and teaching of AI knowledge
and technology and the need to use AI ethically to solve real-life problems for
younger learners. For example, Burgsteiner et al. (2016) believed that the younger
generation should learn about AI knowledge and technologies (e.g., basic concepts
of algorithms, data structures, and programs). Another study Lin and Van Brummelen
(2021) developed an AI curriculum to test students’ motivation (e.g., intrinsic moti-
vation, career interest, confidence to use AI, and learning satisfaction) among 420
primary students in China. These studies provide evidence on adopting digital tech-
nologies to learn AI via well-defined curricula and activities. However, the learning
design and related utilities of AI learning artifacts in primary and secondary class-
rooms have only recently been explored (Chiu & Chai, 2020; Toivonen et al., 2020).
To date, no holistic guidelines and reviews of teaching AI have existed.

2.2.3 Similarities and Differences Between AIED

and AI Literacy

Now let’s clarify the similarities and differences between AIED and AI literacy.
First, AI technologies provide great potential for students to facilitate their learning.
However, these technologies may be novel to students and are not familiar with
them. Therefore, educators have the role to help develop students’ technological
knowledge and skills in order to use the tools ethically and wisely. Recently, schol-
ars suggested the need to update students’ digital competence to facilitate their
learning and working (Rina et al., 2022; Zhang & Aslan, 2021).
Studies have explored the digital competencies to incorporate AI into classrooms
(e.g., ISTE, 2022; Celik, 2022). We can see that AIED and AI literacy may share
common ideas (e.g., digital competence to use AI applications effectively, using AI
to express a knowledge domain, communicating with peers using AI, using AI
applications ethically). However, AI literacy research studies focus more on how to
learn and teach AI, instead of understanding the adoption of AI techniques for edu-
cational purposes. Second, AIED identifies how to use computers to perform cogni-
tive tasks that are usually associated with human minds, including using AI
technology to facilitate learning and teaching in different subjects such as language
2.3 Conclusion 15

∑ Use AI to facilitate people’s learning

∑ Use AI to learn a knowledge domain (e.g., language, math)
∑ Communicate and collaborate effectively with AI

Fig. 2.1 Similarities and differences between AIED and AI literacy education

(Liang et al., 2021), nursing (Hwang et al., 2020), and mathematics education
(Hwang & Tu, 2021). However, AI literacy focuses on the design and implementa-
tion of AI curricula, learning applications, and pedagogical models in K–16 educa-
tion that aims to develop students’ AI knowledge, skills, and attitudes (Rina et al.,
2022). Third, AI literacy is not inclusive of examining how AI technology facilitates
governments, schools, and educators to improve policy settings and educational
administration. From the above arguments, we believe that AI literacy (learning
about AI) should be different from AIED (learning with AI applications and from
AI-driven systems).
Without a universal consensus of these terms, some may identify AI literacy as a
subset of AI education (Wang & Cheng, 2021). Some studies use AI education to
describe AI curriculum, pedagogy, and activities and do not indicate the differences
between the two terms (Chiu & Chai, 2020; Yau et al., 2022). Furthermore, in terms
of learning artifacts, we argue that for AI literacy education, it is not a must to use
AI technologies to learn AI concepts. Prior studies suggested the use of non-AI
artifacts, low-technology settings, and unplugged activities such as digital storytell-
ing, role-play, and social media tools to scaffold students’ AI understandings (Julie
et al., 2020; Ng & Chu, 2021; Rodríguez-García et al., 2021). To visualize our
standpoint, the differences between AIED and AI literacy is summarized in with
sample studies in Fig. 2.1 and Table 2.1.

2.3 Conclusion

The term “AI literacy” was first coined by Burgsteiner et al. (2016) and Kandlhofer
et al. (2016) who proposed it as the ability to understand the knowledge and con-
cepts behind these AI-driven technologies. On top of knowing and using AI
16 2 AI Education and AI Literacy

Table 2.1 Differences and similarities between AIED and AI literacy

AI in education (AIED) AI literacy
Research Identify how to use computers to Investigate how teachers design AI
focus perform cognitive tasks that are usually curricula, learning applications, and
associated with human minds, education models to teach students AI
particularly learning and problem-­ (Chiu & Chai, 2020)
solving in instruction, teaching and Examine how students develop AI
administration (Chen et al., 2020) understandings affectively, behaviorally,
Understand and improve the adoption and cognitively (e.g., perceived ability,
of AI techniques for educational motivation, career interest, confidence
purposes (e.g., automatic grading and to use AI, learning satisfaction) (Chen
assessment, personalized and adaptive et al., 2022; Ng & Chu, 2021)
learning, learning analysis) (Chen Emphasize the learning and teaching of
et al., 2020) AI knowledge and technology, and the
need to use AI ethically (Chen et al.,
2020)
Research Qualitative and quantitative methods to Qualitative and quantitative methods to
methods examine how governments, educational examine how students develop AI and
institutes, educators, and students use related computer science concepts via
AI technology to enhance learning and AI curricula and activities
teaching and facilitate administration
and policy setting
Learning AI technology (e.g., NLP chatbot, AI and related digital technology (e.g.,
tools/ adaptive systems, intelligent agents) simulation of neural networks, robotic
artifacts coding, games)
Non-AI artifacts (e.g., gamified
elements, story writing, role-play)
Sample Use intelligent agent systems to Motivate students to learn AI via social
research improve students’ learning (Hwang media tools, gamification, and
et al., 2020) simulation tasks (Ng & Chu, 2021)
Develop a computer-mediated Enhance students’ intrinsic motivation,
communication competence forecasting career interest, confidence to use AI,
model to facilitate collaborative and learning satisfaction through an AI
problem-based learning (Chih-Ming & curriculum (Lin & Van Brummelen,
Ying-You, 2020) 2021)
Adopt learning analytics and Understand how students learn AI
educational data mining to increase concepts with digital technologies in
impact on policy and practice (Baker different pedagogical approaches (Wong
et al., 2021) et al., 2020)
Use chatbots to provide personalized Develop students’ AI literacy and
services for institutional employees and ethical awareness via AI curricula and
students (Okonkwo & Ade-Ibijola, framework (Burgsteiner et al., 2016;
2021) Kandlhofer et al., 2016)
Similarities Use AI-driven technologies to learn different subject knowledge
Foster students’ digital (AI) literacy to use different AI technologies to facilitate
their learning
Enhance students to communicate, collaborate, and use AI effectively for
learning purposes
References 17

ethically, it serves as a set of competencies that enable people to critically evaluate

AI technologies, communicate, and collaborate effectively with AI (Long &
Magerko, 2020). After knowing the differences between AIED and AI literacy, the
conceptualized term “AI literacy” can provide a strong foundation for future
research (Druga et al., 2019; Long & Magerko, 2020) and enabled further examina-
tion of better instructional design that fosters AI literacy (Chai et al., 2021). This
next section will discuss the reasons why K–16 students should learn AI literacy.

References

Alkhatlan, A., & Kalita, J. (2018). Intelligent tutoring systems: A comprehensive historical survey
with recent developments. arXiv preprint arXiv:1812.09628.
Baker, R. S., McLaren, B. M., Hutt, S., Richey, J. E., Rowe, E., Almeda, M., ... & Andres,
J. M. (2021, June). Towards sharing student models across learning systems. In International
Conference on Artificial Intelligence in Education (pp. 60–65). Springer.
Burgsteiner, H., Kandlhofer, M., & Steinbauer, G. (2016, March). Irobot: Teaching the basics
of artificial intelligence in high schools. In Proceedings of the AAAI Conference on Artificial
Intelligence (Vol. 30, No. 1).
Carpio Cañada, J., Mateo Sanguino, T. J., Merelo Guervós, J. J., & Rivas Santos, V. M. (2015).
Open classroom: Enhancing student achievement on artificial intelligence through an interna-
tional online competition. Journal of Computer Assisted Learning, 31(1), 14–31.
Celik, I. (2022). Towards intelligent-TPACK: An empirical study on teachers’ professional knowl-
edge to ethically integrate artificial intelligence (AI)-based tools into education. Computers in
Human Behavior, 138, 107468.
Chai, C. S., Lin, P. Y., Jong, M. S. Y., Dai, Y., Chiu, T. K., & Qin, J. (2021). Perceptions of and
behavioral intentions towards learning artificial intelligence in primary school students.
Educational Technology & Society, 24(3), 89–101.
Chen, L., Chen, P., & Lin, Z. (2020). Artificial intelligence in education: A review. IEEE Access,
8, 75264–75278.
Chen, X., Zou, D., Xie, H., Cheng, G., & Liu, C. (2022). Two decades of artificial intelligence in
education. Educational Technology & Society, 25(1), 28–47.
Chih-Ming, C., & Ying-You, L. (2020). Developing a computer-mediated communication com-
petence forecasting model based on learning behavior features. Computers and Education:
Artificial Intelligence, 1, 100004.
Chiu, T. K. (2021). A holistic approach to the design of artificial intelligence (AI) education for
K-12 schools. TechTrends, 65(5), 796–807.
Chiu, T. K., & Chai, C. S. (2020). Sustainable curriculum planning for artificial intelligence educa-
tion: A self-determination theory perspective. Sustainability, 12(14), 5568.
Cicirello, V. A. (2008). An interdisciplinary course on artificial intelligence designed for a liberal
arts curriculum. Journal of Computing Sciences in Colleges, 23(3), 120–127.
Dong, S., Wang, P., & Abbas, K. (2021). A survey on deep learning and its applications. Computer
Science Review, 40, 100379.
Druga, S., Vu, S. T., Likhith, E., & Qiu, T. (2019). Inclusive AI literacy for kids around the world.
In Proceedings of FabLearn 2019 (pp. 104–111).
Estevez, J., Garate, G., & Graña, M. (2019). Gentle introduction to artificial intelligence for high-­
school students using scratch. IEEE Access, 7, 179027–179036.
Garcia, N. C., Morerio, P., & Murino, V. (2019). Learning with privileged information via adver-
sarial discriminative modality distillation. IEEE Transactions on Pattern Analysis and Machine
Intelligence, 42(10), 2581–2593.
18 2 AI Education and AI Literacy

Ghallab, M. (2019). Responsible AI: Requirements and challenges. AI Perspectives, 1(1), 1–7.
Helm, J. M., Swiergosz, A. M., Haeberle, H. S., Karnuta, J. M., Schaffer, J. L., Krebs, V. E., et al.
(2020). Machine learning and artificial intelligence: Definitions, applications, and future direc-
tions. Current Reviews in Musculoskeletal Medicine, 13(1), 69–76.
Hinojo-Lucena, F. J., Aznar-Díaz, I., Cáceres-Reche, M. P., & Romero-Rodríguez, J. M. (2019). A
tour of open universities through literature: A bibliometric analysis. The International Review
of Research in Open and Distance Learning, 20(4), 116–131.
Hwang, G. J., & Tu, Y. F. (2021). Roles and research trends of artificial intelligence in mathematics
education: A bibliometric mapping analysis and systematic review. Mathematics, 9(6), 584.
Hwang, G. J., Xie, H., Wah, B. W., & Gašević, D. (2020). Vision, challenges, roles and research
issues of artificial intelligence in education. Computers and Education: Artificial Intelligence,
1, 100001.
Imberman, S. (2004). A laboratory exercise using LEGO handy board robots to demonstrate neu-
ral networks in an artificial intelligence class. Accessible hands-on artificial intelligence and
robotics education (pp. 77–81). Association for the Advancement of Artificial Intelligence,
https://www.aaai.org/Papers/Symposia/Spring/2004/SS-04-01/SS04-01-016.pdf
ISTE. (2022). Hands-on AI projects for the classroom a guide for computer sci-
ence teachers. Retrieved from https://cdn.iste.org/www-­root/Libraries/
D o c u m e n t s % 2 0 % 2 6 % 2 0 F i l e s / A r t i fi c i a l % 2 0 I n t e l l i g e n c e / A I G D C S _ 0 8 2 0 -­r e d .
pdf?_ga=2.157038398.42578658.1661356230-­1998417912.1661356230
Julie, H., Alyson, H., & Anne-Sophie, C. (2020, October). Designing digital literacy activities: An
interdisciplinary and collaborative approach. In 2020 IEEE Frontiers in Education Conference
(FIE) (pp. 1–5). IEEE.
Kahn, K. M., & Carlsson, M. (1985). A grammar kit in Prolog. Uppsala University. Uppsala
Programming Methodology and Artificial Intelligence Laboratory. Retrieved from https://www.
softwarepreservation.org/projects/prolog/uppsala/doc/Kahn_Carlsson-UPMAIL-TR-14C.pdf
Kandlhofer, M., Steinbauer, G., Hirschmugl-Gaisch, S., & Huber, P. (2016, October) Artificial
intelligence and computer science in education: From kindergarten to university. In 2016 IEEE
Frontiers in Education Conference (FIE) (pp. 1–9). IEEE.
Klassner, F. (2002, February). A case study of LEGO Mindstorms’™ suitability for artificial intel-
ligence and robotics courses at the college level. In Proceedings of the 33rd SIGCSE Technical
Symposium on Computer Science Education (pp. 8–12).
Liang, T. P., Robert, L., Sarker, S., Cheung, C. M., Matt, C., Trenz, M., & Turel, O. (2021).
Artificial intelligence and robots in individuals’ lives: How to align technological possibilities
and ethical issues. Internet Research (Vol. 31, pp. 1–10).
Lichtenthaler, U. (2018). Substitute or synthesis: The interplay between human and artificial intel-
ligence. Research-Technology Management, 61(5), 12–14.
Lin, P., & Van Brummelen, J. (2021, May). Engaging teachers to co-design integrated AI curricu-
lum for K-12 classrooms. In Proceedings of the 2021 CHI Conference on Human Factors in
Computing Systems (pp. 1–12).
Long, D., & Magerko, B. (2020, April). What is AI literacy? Competencies and design consider-
ations. In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems
(pp. 1–16).
McCarthy, J. (2007). From here to human-level AI. Artificial Intelligence, 171(18), 1174–1182.
McGovern, A., & Fager, J. (2007, March). Creating significant learning experiences in intro-
ductory artificial intelligence. In Proceedings of the 38th SIGCSE Technical Symposium on
Computer Science Education (pp. 39–43).
Mota-Valtierra, G., Rodríguez-Reséndiz, J., & Herrera-Ruiz, G. (2019). Constructivism-based
methodology for teaching artificial intelligence topics focused on sustainable development.
Sustainability, 11(17), 4642.
Narahara, T., & Kobayashi, Y. (2018). Personalizing homemade bots with plug & play AI for
STEAM education. In SIGGRAPH Asia 2018 Technical Briefs (pp. 1–4).
Newell, A., & Simon, H. (1956). The logic theory machine–A complex information processing
system. IRE Transactions on Information Theory, 2(3), 61–79.
References 19

Ng, D. T. K., & Chu, S. K. W. (2021). Motivating students to learn AI through social networking
sites: A case study in Hong Kong. Online Learning, 25(1), 195–208.
Ng, D. T. K., Leung, J. K. L., Chu, K. W. S., & Qiao, M. S. (2021). AI literacy: Definition, teach-
ing, evaluation and ethical issues. Proceedings of the Association for Information Science and
Technology, 58(1), 504–509.
Okonkwo, C. W., & Ade-Ibijola, A. (2021). Chatbots applications in education: A systematic
review. Computers and Education: Artificial Intelligence, 2, 100033.
Pinkwart, N. (2016). Another 25 years of AIED? Challenges and opportunities for intelligent edu-
cational technologies of the future. International Journal of Artificial Intelligence in Education,
26(2), 771–783.
Rina, R., Kluzer, S., & Punie, Y. (2022). DigComp 2.2: The digital competence framework for
citizens-with new examples of knowledge, skills and attitudes (No. JRC128415). Joint Research
Centre (Seville site).
Rodríguez-García, J. D., Moreno-León, J., Román-González, M., & Robles, G. (2021, March).
Evaluation of an online intervention to teach artificial intelligence with learningml to
10-16-year-old students. In Proceedings of the 52nd ACM Technical Symposium on Computer
Science Education (pp. 177–183).
Roll, I., & Wylie, R. (2016). Evolution and revolution in artificial intelligence in education.
International Journal of Artificial Intelligence in Education, 26(2), 582–599.
Sabuncuoglu, A. (2020, June). Designing one year curriculum to teach artificial intelligence for
middle school. In Proceedings of the 2020 ACM Conference on Innovation and Technology in
Computer Science Education (pp. 96–102).
Steinbauer, G., Kandlhofer, M., Chklovski, T., Heintz, F., & Koenig, S. (2021). A differentiated
discussion about AI education K-12. KI-Künstliche Intelligenz, 35(2), 131–137.
Toivonen, T., Jormanainen, I., Kahila, J., Tedre, M., Valtonen, T., & Vartiainen, H. (2020, July).
Co-designing machine learning apps in K–12 with primary school children. In 2020 IEEE 20th
International Conference on Advanced Learning Technologies (ICALT) (pp. 308–310). IEEE.
Touretzky, D., Gardner-McCune, C., Martin, F., & Seehorn, D. (2019, July). Envisioning AI for
K-12: What should every child know about AI?. In Proceedings of the AAAI Conference on
Artificial Intelligence (Vol. 33, No. 01, pp. 9795–9799).
Wallace, S. A., McCartney, R., & Russell, I. (2010). Games and machine learning: A powerful
combination in an artificial intelligence course. Computer Science Education, 20(1), 17–36.
Wang, P. (2019). On defining artificial intelligence. Journal of Artificial General Intelligence,
10(2), 1–37.
Wang, T., & Cheng, E. C. K. (2021). An investigation of barriers to Hong Kong K-12 schools incor-
porating artificial intelligence in education. Computers and Education: Artificial Intelligence,
2, 100031.
Williamson, B., & Eynon, R. (2020). Historical threads, missing links, and future directions in AI
in education. Learning, Media and Technology, 45(3), 223–235.
Wong, G. K., Ma, X., Dillenbourg, P., & Huan, J. (2020). Broadening artificial intelligence educa-
tion in K-12: Where to start? ACM Inroads, 11(1), 20–29.
Xia, Q., Chiu, T. K., Lee, M., Sanusi, I. T., Dai, Y., & Chai, C. S. (2022). A self-determination
theory (SDT) design approach for inclusive and diverse artificial intelligence (AI) education.
Computers & Education, 189, 104582.
Yau, K. W., Chai, C. S., Chiu, T. K., Meng, H., King, I., & Yam, Y. (2022). A phenomenographic
approach on teacher conceptions of teaching artificial intelligence (AI) in K-12 schools.
Education and Information Technologies, 1–24.
Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of
research on artificial intelligence applications in higher education–where are the educators?
International Journal of Educational Technology in Higher Education, 16(1), 1–27.
Zhang, K., & Aslan, A. B. (2021). AI technologies for education: Recent research & future direc-
tions. Computers and Education: Artificial Intelligence, 2, 100025.
Chapter 3
AI Literacy for All

Artificial intelligence (AI) has influenced various industries (e.g., business, science,
art, education) rather than merely computer science fields to improve working and
learning efficiency. There are many AI-driven applications in daily life (e.g., smart
home appliances, smartphones, Google, Siri) to enhance user experience and help
people lead a better life. To prepare students to be more ready for tomorrow’s work-
place, recent studies suggest the importance of learning AI for all students such as
how computers learn from its data, the strengths and weaknesses of AI-driven appli-
cations, and its ethical concerns about AI (e.g., Long & Magerko, 2020; Ng & Chu,
2021; Sing et al., 2022). Studies suggest that even learners as young as kindergar-
teners should start to learn AI so that educators can bring them up to become respon-
sible digital citizens (Su & Yang, 2022; Su & Zhong, 2022; Williams et al., 2019a, b).
This chapter posits AI literacy should be acquired by all learners, while differences
lie in the content and approach at different education levels, i.e., K–16. For example,
many countries have started to discuss how to incorporate AI curricula in primary
and secondary levels. However, seldom do studies discover why AI should be taught
at kindergarten and noncomputer university education. A list of reasons why people
should learn AI is discussed in this chapter.

3.1 AI for Living, Workplace, Learning, and Societal Good

Artificial intelligence has made great influences in our everyday lives, not only in
technical areas but across industries. Countries/regions have incorporated AI into
computer science education, and AI has been added to some of the global digital
competency frameworks (e.g., ISTE, DigComEdu) to update the latest educational
standards for K–16 learners. There is a great demand to foster students’ AI literacy
to prepare for valuable abilities and skills before entering the workforce. Learning

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 21

D. T. K. Ng et al., AI Literacy in K-16 Classrooms,
https://doi.org/10.1007/978-3-031-18880-0_3
22 3 AI Literacy for All

and understanding what AI is, how it works, and its affordances are the first steps to
successful study and career in the future. The following sections suggest four major
reasons why students should learn AI.

3.1.1 AI for Living

AI is transforming every walk of life and has a great impact on our society. The first
reason why all learners need to learn AI is that it enhances people’s living standards.
First of all, it is a driving force behind social media that AI is used to personalize
what is seen on the social media feed through evaluation of post history to identify
people’s interests and make appropriate suggestions. A University of Oxford’s study
shows that AI will soon become as advanced as human beings in translating lan-
guages by 2024, writing school essays by 2026, selling goods by 2031, writing a
bestselling book by 2049, and conducting surgeries by 2053 (Freeman, 2018). Since
AI will improve human’s lives in almost everything and improve our living stan-
dard, there is a need to foster citizens and young learners to become AI literates so
that people can use AI-driven tools and interact with others in an appropriate,
responsible, and empowered manner (Calzada et al., 2021).
Second, technology companies are exaggerating the capabilities of AI in their
products (Surden, 2018). Consumers who understand basic AI algorithms will not
be easily tricked by popular media and advertisements. They can think critically
about what counts as AI, question the features the AI offers, and think about the
ethical concerns and limitations behind. On the other hand, consumers should also
be responsible users. Both producers and consumers should handle AI with good
intentions to empower human lives. Studies have shown that teaching digital citi-
zenship is essential to help people achieve and understand digital literacy and digital
responsibility and ensure safety (Robinson, 2020), digital wealth, and wellness
(Jeske et al., 2021). Overall, as one of the most important technological skills, edu-
cators can instill learners with AI knowledge and skills to succeed as lifetime learn-
ers with AI literacy, so that our future generations could wisely and ethically
facilitate their living, working, and learning.

3.1.2 AI for Workplaces

The advancement of AI in turn raises the concern by people that AI will take over
millions of current jobs and disrupt the labor force in some occupations (Agarwal,
2018). However, according to the World Economic Forum (2021)’s report, AI is
expected to replace 97 millions of jobs worldwide by 2025, and at the same time it
will create many job opportunities in many industries. One thing that is certain, AI
will soon change the future workplace, but workers and students may not be ready
to equip with the related capabilities to fill these new job opportunities. Some
3.1 AI for Living, Workplace, Learning, and Societal Good 23

speculate that 15% of the working hours will be automated and clerical jobs will be
easily eliminated by AI (Manyika et al., 2017). Therefore, it is essential to equip
young learners with fundamental AI knowledge, technologies, and mindsets for
digital workplace challenges in the future.
Instead of viewing AI as a competitor to work labor, AI can generate compelling
benefits for people, industries, and business, thus raising people’s working produc-
tivity and economic growth. There is a widespread shortage of AI professionals that
possess the required skills and knowledge. In recent years, governments, compa-
nies, and institutions start to upskill and reskill their employees to harness opportu-
nities of using AI technologies to enhance their working efficiency and be aware of
ethical implications and risks (Johnson et al., 2021). For example, Tesla CEO Elon
Musk claimed in 2021 that his company was in the advanced stages of developing
an autonomous android that would relieve humans of their hazardous, repetitive and
boring jobs. The electric-vehicle maker’s humanoid Optimus robots could be in
production by the end of 2023 and the company plans to deploy thousands of the
robots to help resolve future labor shortages for the U.S. economy (Masunaga,
2022). Robots and AI are expected to permeate our daily lives by 2025 (Stahl,
2021). There is a strong need to enforce on-the-job training and vocational and uni-
versity education to support workers and university students in adapting to
AI-enhanced ways of working. With adequate training, young learners could equip
themselves with AI competencies to gain a competitive advantage at work, equip
with futuristic skills, and fulfill job demands in the AI industry.

3.1.3 AI for Learning

AI-driven technologies have created opportunities for enhancing learning experi-

ence through intelligent tutoring, individualized learning, and recommendation sys-
tems (Zawacki-Richter et al., 2019). However, students may not be familiar with
these novel technologies to facilitate their learning in terms of student cognition,
engagement, and collaboration. For example, students could use virtual assistants
like Google translation to facilitate their reading and writing (Godwin-Jones, 2011)
and interact with chatbots to gain new knowledge (Okonkwo & Ade-Ibijola, 2021).
Moreover, PowerPoint includes some AI features (e.g., Design Ideas, Presenter
Coach) which enable users to stylish their presentations and make inclusive presen-
tations with live captions and subtitles that gives 88% accuracy while predicting the
confidence level of the students. In this way, users can generate stunning presenta-
tions, and practice their presentation skills (Microsoft, 2019). However, students
may meet challenges such as technological difficulties, communication, and col-
laboration when using these technologies (e.g., Vincent-Lancrin & van der Vlies,
2020; Zhang et al., 2021). Students who are equipped with AI basic competency
would outperform others compared to their counterparts and enjoy the AI-driven
learning experiences (Hwang et al., 2022). Therefore, there is a need for educators
to develop students’ AI literacy so as to adapt to digital transformation in their learn-
ing environments.
24 3 AI Literacy for All

3.1.4 AI for Societal Good

Countries across the globe face a growing set of shared challenges that will require the
next generation to learn, build and connect to identify practical and innovative solu-
tions. AI presents an opportunity to build better tools and solutions that help address
some of the world’s challenging issues, and deliver positive social impact in accor-
dance with the priorities outlined in the United Nations’ 17 Sustainable Development
Goals. The AI for Social Good movement aims to establish interdisciplinary partner-
ships centered around AI applications towards SDGs (Tomašev et al., 2020).
AI could advance human’s sustainable development and solve future global chal-
lenges about environmental, humanistic issues, accessibility, health, and cultural
heritage (Microsoft, 2021). For example, planetary computers were created to build
a global environmental network, empower organizations and individuals working to
advance sustainability to solve today’s environmental challenges (Vinuesa et al.,
2020). AI for health empowers researchers and AI institutions to improve the health
of people and communities around the world (Reddy et al., 2020). AI can help
improve global independence and inclusion in society to make a better living envi-
ronment, community, education, and employment. Moreover, it supports humanis-
tic response, refugees, displaced persons, human rights, the needs of women and
children, preservation and enrichment of cultural heritage (Tomašev et al., 2020). In
K-16 education, educators need to equip young learners to become future-ready
with the digital and sustainable mindset so that they could help solve these world’s
challenges when they grow up. To address these global challenges, the idea “AI for
societal good” is prominent that everyone should learn AI to solve future global
challenges and enhance humans’ living standard.

3.2 Benefits of AI Literacy for Different Educational Levels

This section further discussion the specific major reasons why AI should be taught
to promote digital citizenship to facilitate people’s living, working, and learning to
build a better society, the following section further discusses the specific reasons for
learners to learn AI at different educational levels: kindergarteners, primary and
secondary students, and noncomputer science university students.

3.2.1 Kindergarteners

Nowadays, there are many sophisticated and robotic toys available for children.
Children can engage in playful experience and conversations with artificially intel-
ligent assistants such as Siri and Alexa. Scholars proposed that learners as young as
3/4 years old could have the ability to start exploring AI in a simple and
3.2 Benefits of AI Literacy for Different Educational Levels 25

foundational manner (Preface, 2021; Su & Yang, 2022; Su et al., 2022). Children are
rapid and curious learners. Learning AI can be a very fun and rewarding educational
experience if suitable learning methods and tools are used (Preface, 2021). For
example, Williams (2018) and Williams et al. (2019a, b) used the suitable learning
tool to help children understand the concept of AI through PopBots. Other studies
also pointed out that children can experience AI knowledge through different pro-
gramming toys, such as PopBots, Zhorai, and Teachable Machine (Williams, 2018;
Williams et al. 2019a, b; Lin et al., 2020; Vartiainen et al., 2020). Furthermore, some
studies found that learning AI can enhance children’s cognitive and social develop-
ment, such as improving their inquiry skills (Kewalramani et al., 2021), foster stu-
dents’ reading literacy about technology (Mah et al., 2021), improve students’
adaptive behavior (Shin et al., 2012), and promote interaction and collaboration
among children via social robots (Prentzas, 2013). Therefore, it is possible for kin-
dergarten children to develop AI literacy as a complement to their stages of cogni-
tive development.

3.2.2 Primary and Secondary Education

Many view AI literacy as a crucial component of national strategy for digital citi-
zenship education (Seldon & Abidoye, 2018). In today’s digital world, learning how
to interact with and communicate using AI tools is evident in almost all aspects of
everyday life. However, most young students are AI “illiterates” and do not under-
stand the technologies behind (McStay & Rosner, 2021). Recent discussions have
also sparked discussions about the importance of learning AI ethics (Borenstein &
Howard, 2021; Hagendorff, 2020) and equipping young children with proper mind-
sets (Floridi et al., 2018).
The goal of implementing AI literacy education at primary and secondary level is
not to train and nurture computer programmers. Instead, it aims to offer students
hands-on experience and enable them to solve problems, interact with, and commu-
nicate with AI tools in everyday life (Ng et al., 2021a, b). It enhances students’ AI
knowledge and basic digital skills using the latest technologies such as chatbots and
intelligent agents. Students need to learn how to evaluate, communicate, and col-
laborate effectively with AI and use AI as a tool ethically online, at home, and in the
future workplace (Long & Magerko, 2020). Aligned with other technological skills
like computational thinking, students were not merely the end users of the AI tech-
nologies; they needed to learn how to solve problems using AI in authentic settings.
Another trend is that primary and secondary schools have placed high emphasis
in STEM education. Young students should learn how to use AI technologies to help
solve problems and enhance their working and learning efficiency. For example,
students can use automatic translation and grammatical checking tools to facilitate
their writing (Lee, 2020). They can use AI tools to adjust parameters (e.g., ages,
gender) of people and stylize paintings and photos in their social media feeds
(Greenfield, 2021). These examples demonstrated that students with stronger AI
26 3 AI Literacy for All

literacy skills can outperform their counterparts to live, work, and learn efficiently
in the twenty-first century.
We recognize the need for introducing AI into primary and secondary education
to foster their digital literacy skills. The authors worked with university professors,
local AI developers, and primary and secondary teachers to develop one of the first
AI literacy curricula called “AI for All” (Ng et al., 2021). The program has been
successfully implemented in a number of primary and secondary schools in Hong
Kong. In the academic year of 2021/22, a secondary school helped Clearbot (a local
start-up) design an autonomous trash-collecting boat that can identify floating gar-
bage and clean the harbor (HKET, 2022). A primary school worked with local envi-
ronmentalists to design an AI-empowered scarecrow to fright pests left and grow
morecrops.

3.2.3 Noncomputer Science University Students

Industries have started to upskill and reskill their employees in their fields to expose
learners to AI. Moreover, graduates who are AI literates can enhance their employ-
ability and working efficiency to help them solve authentic problems in workplaces.
Therefore, universities have started to engage students from diverse study back-
grounds in AI literacy programs to build up their foundational AI knowledge and
support them to solve problems by developing AI applications (e.g., Kong et al.,
2021). Courses have different focuses for a diverse population of AI literates to
tackle skill gaps to expand learning opportunities for all learners. For example,
some courses may not focus on mathematical formulae and programming since
most students do not need to learn the underlying mathematical and technical con-
cepts behind the AI technologies (Long & Magerko, 2020). Some courses focus on
supporting professionals in a particular industry (e.g., healthcare, business, law)
how to manipulate the AI-driven systems and machines to enhance their working
efficiency, collaborate with their colleagues, and know the ethical concerns and
limitations behind the technologies (e.g., Hwang et al., 2022; Xu & Babaian, 2021).
As such, we can see that there is a need to boost students’ technical skills and
how to apply to AI to meet the career and industrial needs. The goal of these types
of courses aims to speak directly to the skills that employers are actively seeking
within the AI development sector (e.g., Python programming, machine learning,
robotics, data science). These discipline-specific AI courses and programs can
expose learners to the most in-demand topics within AI and underpin AI applica-
tions and knowledge that are related to their fields. In this way, students are digitally
ready to tackle problems of the near future after graduation. Not only will learners
enhance their practical application of AI but also attain a credible qualification
focused on AI to enhance their employability. Proficiency in AI is highly valued in
today’s companies. Consequently, this may lead to career prospects in developing,
managing, and planning AI solutions for a variety of businesses.
References 27

3.3 Conclusion

In this chapter, we presented four major reasons why all citizens should learn AI (for
living, workplaces, learning, and society good), and its particular reasons why AI lit-
eracy is important at each educational level (kindergarten, primary school, secondary
school, and noncomputer science university level). Recent trends from machine
learning to algorithms and robotics make AI literacy education prevalent in the educa-
tion field. Universities and K–12 schools encompass an extensive range of topics such
as AI, machine learning tools, and related techniques that are practical to many indus-
tries. In these programs, students can learn how AI works and what it is, but, most
importantly, how to implement it in real-world scenarios and job roles is paramount
to integrating new competence into the changing technological landscape. Students
with AI literacy skills can enhance their living standards, use AI applications ethically
and wisely, enhance their learning and working effectiveness, be at the forefront of
the future, enhance their employability, advance their study after graduation, and be
adaptive to the rapidly updating and changing learning tools and environment.

References

Agarwal, P. K. (2018). Public administration challenges in the world of AI and bots. Public
Administration Review, 78(6), 917–921.
Borenstein, J., & Howard, A. (2021). Emerging challenges in AI and the need for AI ethics educa-
tion. AI and Ethics, 1(1), 61–65.
Calzada, I., Pérez-Batlle, M., & Batlle-Montserrat, J. (2021). People-centered smart cities: An
exploratory action research on the cities’ coalition for digital rights. Journal of Urban Affairs,
44, 1–26.
Floridi, L., Cowls, J., Beltrametti, M., Chatila, R., Chazerand, P., Dignum, V., Luetge, C., Madelin,
R., Pagallo, U., Rossi, F., Schafer, B., Valcke, P., & Vayena, E. (2018). AI4People—An ethi-
cal framework for a good AI society: Opportunities, risks, principles, and recommendations.
Minds and Machines, 28(4), 689–707.
Freeman, R. B. (2018). Ownership when AI robots do more of the work and earn more of the
income. Journal of Participation and Employee Ownership, 1(1), 74–95.
Godwin-Jones, R. (2011). Mobile apps for language learning. Language Learning & Technology,
15(2), 2–11.
Greenfield, G. (2021). Artificial life and artificial intelligence advances in the visual arts. Artificial
intelligence and the arts (pp. 3–26). Springer.
Hagendorff, T. (2020). The ethics of AI ethics: An evaluation of guidelines. Minds and Machines,
30(1), 99–120.
HKET. (2022). Middle school students create AI-powered robotic boats to help clean up marine
garbage in Deep Water Bay. Retrieved from shorturl.at/elsV0
Hwang, G. J., Tu, Y. F., & Tang, K. Y. (2022). AI in online-learning research: Visualizing and
interpreting the journal publications from 1997 to 2019. The International Review of Research
in Open and Distance Learning, 23(1), 104–130.
Jeske, T., Terstegen, S., & Stahn, C. (2021). Opportunities of digitalization and artificial intel-
ligence for occupational safety and health in production industry. International conference on
human-computer interaction (pp. 43–57). Springer.
28 3 AI Literacy for All

Johnson, M., Jain, R., Brennan-Tonetta, P., Swartz, E., Silver, D., Paolini, J., et al. (2021). Impact
of big data and artificial intelligence on industry: Developing a workforce roadmap for a data
driven economy. Global Journal of Flexible Systems Management, 22(3), 197–217.
Kewalramani, S., Kidman, G., & Palaiologou, I. (2021). Using artificial intelligence (AI)- inter-
faced robotic toys in early childhood settings: A case for children’s inquiry literacy. European
Early Childhood Education Research Journal, 29(5), 652–668.
Kong, S. C., Cheung, W. M. Y., & Zhang, G. (2021). Evaluation of an artificial intelligence literacy
course for university students with diverse study backgrounds. Computers and Education:
Artificial Intelligence, 2, 100026.
Lee, S. M. (2020). The impact of using machine translation on EFL students’ writing. Computer
Assisted Language Learning, 33(3), 157–175.
Lin, P., Van Brummelen, J., Lukin, G., Williams, R., & Breazeal, C. (2020, April). Zhorai: Designing
a conversational agent for children to explore machine learning concepts. Proceedings of the
AAAI Conference on Artificial Intelligence, 34(09), 13381–13388.
Long, D., & Magerko, B. (2020, April). What is AI literacy? Competencies and design consider-
ations. In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems
(pp. 1–16).
Mah, G., Hu, X., & Yang, W. (2021). Digital technology use and early reading abilities among
bilingual children in Singapore. Policy Futures in Education, 19(2), 242–258.
Manyika, J., Chui, M., Miremadi, M., Bughin, J., George, K., Willmott, P., & Dewhurst, M. (2017).
A future that works: AI, automation, employment, and productivity. McKinsey Global Institute
Research, Tech. Rep, 60, 1–135.
Masunaga, S. (2022). Will Elon Musk’s Tesla Bot replace human workers? Don’t bet on it. Retrieved
from https://www.latimes.com/business/story/2022-09-30/tesla-automation-workforce.
McStay, A., & Rosner, G. (2021). Emotional artificial intelligence in children’s toys and devices:
Ethics, governance and practical remedies. Big Data & Society, 8(1), 2053951721994877.
Microsoft. (2019) 4 ways AI in PowerPoint will help you nail your next presentation. Retrieved
from https://news.microsoft.com/en-in/4-ways-ai-in-powerpoint-will-help-you-nail-your-
next-presentation/
Microsoft. (2021). AI for good. Retrieved from https://www.microsoft.com/en-us/ai/ai-for-good
Ng, T. K., & Chu, S. K. W. (2021). Motivating students to learn AI through social networking sites:
A case study in Hong Kong. Online Learning, 25(1), 195–208.
Ng, D.T.K., Lai, L.F.W., & Cheung T.W.T. (2021). Artificial Intelligence for All. eFunLearning
Ltd., Hong Kong Science and Technology Parks, Hong Kong.
Ng, D. T. K., Leung, J. K. L., Chu, S. K. W., & Qiao, M. S. (2021a). Conceptualizing AI literacy:
An exploratory review. Computers and Education: Artificial Intelligence, 2, 100041.
Ng, D. T. K., Leung, J. K. L., Chu, K. W. S., & Qiao, M. S. (2021b). AI literacy: Definition, teach-
ing, evaluation and ethical issues. Proceedings of the Association for Information Science and
Technology, 58(1), 504–509.
Okonkwo, C. W., & Ade-Ibijola, A. (2021). Chatbots applications in education: A systematic
review. Computers and Education: Artificial Intelligence, 2, 100033.
Preface. (2021). The ultimate guide for artificial intelligence (AI) for kids. Retrieved from https://
www.preface.ai/blog/kids-­learning/ai-­for-­kids
Prentzas, J. (2013). Artificial intelligence methods in early childhood education. In Artificial
Intelligence, evolutionary computing and metaheuristics (pp. 169–199). Springer.
Reddy, S., Allan, S., Coghlan, S., & Cooper, P. (2020). A governance model for the application of
AI in health care. Journal of the American Medical Informatics Association, 27(3), 491–497.
Robinson, S. C. (2020). Trust, transparency, and openness: How inclusion of cultural values shapes
Nordic national public policy strategies for artificial intelligence (AI). Technology in Society,
63, 101421.
Seldon, A., & Abidoye, O. (2018). The fourth education revolution. Legend Press Ltd..
Shin, S., Koh, M. S., & Yeo, M. H. (2012). A comparative study of the preliminary effects in the
levels of adaptive behaviors: Learning program for the development of children with autism
(LPDCA). Journal of the International Association of Special Education, 13(1), 6.
References 29

Sing, C. C., Teo, T., Huang, F., & Chiu, T. K. (2022). Secondary school students’ intentions to
learn AI: Testing moderation effects of readiness, social good and optimism. Educational
Technology Research and Development, 70(3), 765–782.
Stahl, A. (2021). The rise of artificial intelligence: Will robots actually replace peo-
ple? Retrieved from https://www.forbes.com/sites/ashleystahl/2022/05/03/
the-rise-of-artificial-intelligence-will-robots-actually-replace-people/?sh=3cbae7203299.
Su, J., & Yang, W. (2022). Artificial intelligence in early childhood education: A scoping review
(p. 100049). Artificial Intelligence.
Su, J., & Zhong, Y. (2022). Artificial intelligence (AI) in early childhood education: Curriculum
design and future directions. Computers and Education: Artificial Intelligence, 3, 100072.
Su, J., Zhong, Y., & Ng, D. T. K. (2022). A meta-review of literature on educational approaches
for teaching AI at the K-12 levels in the Asia-Pacific region (p. 100065). Artificial Intelligence.
Surden, H. (2018). Artificial intelligence and law: An overview. Georgia State University Law
Review, 35, 1305.
Tomašev, N., Cornebise, J., Hutter, F., Mohamed, S., Picciariello, A., Connelly, B., ... &
Clopath, C. (2020). AI for social good: unlocking the opportunity for positive impact. Nature
Communications, 11(1), 1–6.
Vartiainen, H., Tedre, M., & Valtonen, T. (2020). Learning machine learning with very young
children: Who is teaching whom? International Journal of Child-Computer Interaction,
25, 100182.
Vincent-Lancrin, S., & van der Vlies, R. (2020). Trustworthy artificial intelligence (AI) in educa-
tion: Promises and challenges. Retrieved from https://www.oecd-­ilibrary.org/content/paper/
a6c90fa9-­en
Vinuesa, R., Azizpour, H., Leite, I., Balaam, M., Dignum, V., Domisch, S., ... & Fuso Nerini,
F. (2020). The role of artificial intelligence in achieving the Sustainable Development Goals.
Nature Communications, 11(1), 1–10.
Williams, R. (2018). PopBots: Leveraging social robots to aid preschool children’s artificial intel-
ligence education (Doctoral dissertation, Massachusetts Institute of Technology).
Williams, R., Park, H. W., & Breazeal, C. (2019a, May). A is for artificial intelligence: The impact
of artificial intelligence activities on young children’s perceptions of robots. In Proceedings of
the 2019 CHI Conference on Human Factors in Computing Systems (pp. 1–11).
Williams, R., Park, H. W., Oh, L., & Breazeal, C. (2019b, July). Popbots: Designing an artificial
intelligence curriculum for early childhood education. Proceedings of the AAAI Conference on
Artificial Intelligence, 33(01), 9729–9736.
Xu, J. J., & Babaian, T. (2021). Artificial intelligence in business curriculum: The pedagogy and
learning outcomes. The International Journal of Management Education, 19(3), 100550.
Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of
research on artificial intelligence applications in higher education–where are the educators?
International Journal of Educational Technology in Higher Education, 16(1), 1–27.
Zhang, D., Mishra, S., Brynjolfsson, E., Etchemendy, J., Ganguli, D., Grosz, B., ... & Perrault,
R. (2021). The AI index 2021 annual report. arXiv preprint arXiv:2103.06312.
Chapter 4
The Landscape of AI Literacy

This chapter introduced various models of AI literacy education, in particular build-

ing on the twenty-first century skills framework to comprise important digital skill
sets in today’s digital world. Such skills are essential given the challenges brought
about by technological advances and changes in the global economic structure to
keep the educational standards across regions. This chapter examines how AI is
adapted from other digital literacies, the global movement of AI literacy education,
and a broad scope of AI literacy from K–16 education.

4.1 AI Literacy as a Twenty-First Century Skill

The definitions of literacy vary. It commonly refers to “the ability to identify, under-
stand, interpret, create, communicate and compute, using printed and written mate-
rials associated with varying contexts” (UNESCO, 2018, p. 2) It involves a
continuum of learning in enabling people to achieve their goals, to develop their
knowledge and potential, and to participate fully in their community and society. In
today’s digital world, every citizen needs to engage positively, critically, and com-
petently to effectively communicate and collaborate with others through responsi-
ble use of technology (Fleaca & Stanciu, 2019). This term has been extended to new
literacies such as media, digital, information, computer, and AI literacy (Kong et al.,
2021). Artificial intelligence is affecting our everyday life to the extent where it is
important for the new generation to develop the necessary literacies of AI so as to
understand and use related technologies (Wong et al., 2020; Ng et al., 2022). To
combine AI and literacy, AI literacy means having the essential abilities that people
need to live, learn, and work in our digital world through AI-driven technologies,
and this should be taught at the K–12 levels (Steinbauer et al., 2021). This provides
the foundation that we map these literacy skills in Table 4.1 to view the concept of
AI literacy as the focal point of the discussion.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 31

D. T. K. Ng et al., AI Literacy in K-16 Classrooms,
https://doi.org/10.1007/978-3-031-18880-0_4
32 4 The Landscape of AI Literacy

Table 4.1 Digital literacy skills

Types of literacy
skills Definitions
Computer The knowledge and ability to use a computer and its software to accomplish
literacy practical tasks (Talja et al., 2005)
Digital literacy It is more than just the technical ability to operate digital devices properly; it
comprises a variety of cognitive skills that are utilized in performing digital
tasks (Van Laar et al., 2017)
Media literacy The ability to access, understand, and create communications in a variety of
contexts. A media-literate person can think critically about what they see,
hear, and read in multiple media (Livingstone, 2004)
Information The ability to know when there is a need for information, to be able to
literacy identify, locate, evaluate, and effectively use that information for the issue or
problem at hand (Bruce, 1997)
Technological The ability to responsibly use appropriate technology in various tasks such as
literacy communication, problem-solving, information access, management, and
creation across disciplines (Gamire & Pearson, 2006)
Adapted from Chu et al. (2021)

As a part of digital literacy, AI literacy education enables students to develop

AI-related technological understanding and the digital competence to use, under-
stand, and create AI solutions. In recent years, scholars have proposed adding “AI
literacy” to a set of twenty-first century technological competencies that everyone
needs to learn to facilitate their learning, working, and living (Long & Magerko,
2020; Ng et al., 2021). Educational frameworks (e.g., the ISTE, DigCompEdu) have
included AI to update the latest educational standards to address digital skill levels
across the globe (Demeshkant et al., 2020; Riina et al., 2022). Governments have
developed AI curricula to align their educational standards with other neighboring
countries (Su et al., 2022). Research examined the effectiveness of their AI curricu-
lum from different perspectives including self-determination theory (Xia et al.,
2022), digital readiness, social good and optimism (Sing et al., 2022), learning
motivation (Lin & Van Brummelen, 2021), behavioral intentions (Chai et al., 2021),
and teacher conceptions (Yau et al., 2022). These studies provide evidence that AI
education has become a necessary component for young learners to equip them-
selves for their future studies and career.

4.2 Emerging Frameworks for AI Literacy Education

A range of international curricula and frameworks have emerged these years to

foster young learners’ AI literacy. According to the UNESCO (2022)’s survey, there
are 14 AI curricula which have been developed by 11 countries. In this section, we
review a number of these frameworks for AI literacy skills that have been developed
in different educational policy environments around the world. We also look at the
education reforms in response to the AI literacy skills frameworks put forward by
various governments. Policymakers and curriculum developers who decide to
4.2 Emerging Frameworks for AI Literacy Education 33

incorporate AI literacy into their curricula could learn from other countries/regions.
There are a number of frameworks to map AI curriculum for K–12 education to
guide K–12 AI literacy education. This section first suggests two renowned frame-
works that were widely discussed and cited in the field of AI literacy education:
Long and Magerko (2020)‘s AI literacy competency framework and Touretzky et al.
(2019)’s five big ideas of AI. Then, we summarize some review papers from various
places to inform its current development for educators, schools, and policymakers
to implement AI literacy education.

4.2.1 Competencies and Design Considerations

Long and Magerko (2020) presented 17 AI literacy competencies (Table 4.2) and 15
design considerations for AI literacy based on a scoping study of existing research
to uncover what AI professionals believe all citizens should know and common
perceptions and misconceptions among learners. The framework focuses on what
educators should involve in their AI curricula and how they design the instruction of
AI literacy education.

4.2.2 The Five “Big Ideas” About AI

Since 2018, the AI4K12 Initiative has been developing national guidelines for AI
literacy education in K–12. Touretzky et al. (2019) divided the learning concepts
into the five “big ideas” of AI to formulate a sound framework on fostering AI lit-
eracy. The guidelines define what every student should know about AI with guide-
lines and activities that help students to build the competencies. The guideline
serves as a framework to build educational standards for curriculum developers and
educators on AI concepts, essential knowledge, and skills across educational levels.
The five big ideas are summarized below: (1) perception (computers perceive the
world using sensors), (2) representation and reasoning (agents maintain representa-
tion of the world and use them for reasoning), (3) learning (computers can learn
from data), (4) natural interaction (intelligent agents require many kinds of knowl-
edge to interact naturally with humans), and (5) societal impact (AI can impact
society in both positive and negative ways). Touretzky et al. (2022) further provides
an in-depth look at how K-12 students should be introduced to AI knowledge and
skills. This review discusses the general format of the guidelines to highlight the
appropriateness of the knowledge and skills. This set of guidelines is informed by
the need for alignment with CSTA’s K-12 Computer Science Standards, Common
Core standards and Next Generation Science Standards. Examples were presented
to present the learning progression across grade levels and concepts for a specific
grade level. Table 4.3 outlined the five big ideas about AI, and the details of the first
four areas have already been available for K–12 educators; however, the fifth big
idea is still under development.
34 4 The Landscape of AI Literacy

Table 4.2 AI literacy competencies (Long & Magerko, 2020)

Competency Descriptions
1. Recognizing AI Distinguish between technological artifacts that use and do not use AI
2. Understanding Critically analyze and discuss features that make an entity “intelligent,”
intelligence including discussing differences between human, animal, and machine
intelligence
3. Interdisciplinarity Recognize that there are many ways to think about and develop
“intelligent” machines. Identify a variety of technologies that use AI,
including technology spanning cognitive systems, robotics, and ML
4. General vs narrow Distinguish between general and narrow AI
AI
5. AI strengths and Identify problem types that AI excels at and problems that are more
weaknesses challenging for AI. Use this information to determine when it is
appropriate to use AI and when to leverage human skills
6. Imagine future AI Imagine possible future applications of AI, and consider the effects of
such applications on the world
7. Representations Understand what a knowledge representation is and describe some
examples of knowledge representations
8. Decision-making Recognize and describe examples of how computers reason and make
decisions
9. ML steps Understand the steps involved in machine learning and the practices and
challenges that each step entails
10. Human role in AI Recognize that humans play an important role in programming, choosing
models, and fine-tuning AI systems
11. Data literacy Understand basic data literacy concepts
12. Learning from Recognize that computers often learn from data (including one’s own
data data)
13. Critically Understand that data needs interpretation. Describe how the training
interpreting data examples provided in an initial dataset can affect the results of an
algorithm
14. Action and Understand how AI systems act in the world that can be directed by
reaction higher-level reasoning or it can be reactive
15. Sensors Understand what sensors are, recognize how computers perceive the
world using sensors, and identify sensors on a variety of devices.
Identify different sensors to support computer representation and
reasoning
16. Ethics Identify and describe key ethical concerns about AI (e.g., risks,
limitations, privacy, bias, misinformation, ethical decision-making,
diversity, transparency, accountability)
17. Programmability Understand that agents are programmable
Table 4.3 Five big ideas about AI (AIK12, 2022)
Big ideas Key concepts Descriptions
Perception Human vs Computers perceive the world using sensors in which
machines perception is the extraction of meaning from sensory
Computer sensors information using knowledge
Digital encoding Such transformation from signal to meaning takes place
Sensing vs in cognitive stages and more abstract features and
perception higher-level knowledge are involved at each stage
Feature extraction
Abstraction via
language and
vision
Types of domain
knowledge
Representation Abstraction Representations are about data structures that support
and reasoning Symbolic varying reasoning methods (i.e., algorithms) to operate
representations on real-world problems
Data structures There are two major types of knowledge representations:
Feature vectors symbols and graphs (e.g., queue, stack, map, graph,
State spaces and game board, decision trees, neural network) and
operators numerical representations and feature vectors (i.e.,
Combinatorial numerical properties of observed phenomena) (e.g.,
search encapsulation, natural language processing,
Types of reasoning polymorphism)
problems Agents are considered intelligent when they employ a
Reasoning nontrivial sense-deliberate-act cycle to make progress to
algorithms achieve their goals. In other words, AI needs to sense,
deliberate (reason), and act that requires computational
sophistication and has significant computing power. For
example, electronic door openers are not intelligent
agents because their sensing, reasoning, and action are
all trivial and simple
Learning Nature of learning Machine learning enables a computer to imitate human
 Humans vs behaviors without people explicitly programming those
machines behaviors
 Finding patterns Learning new behaviors results from changes the
in data learning algorithm makes to the internal representations
 Training a of a reasoning model (e.g., decision tree, neural network)
model Large training datasets are required to narrow down the
 Using and learning algorithm’s choices when the reasoning model
constructing a can enable the AI to behave differently
reasoning model The reasoning model constructed by the ML algorithm
 Adjusting can be applied to new data to solve problems and make
internal decisions
representations
 Learning from
experience
 Neural network
 Structure of a
neural network
 Weight
adjustment
Datasets
 Feature sets
 Handling large
datasets
 Bias
(continued)
36 4 The Landscape of AI Literacy

Table 4.3 (continued)

Big ideas Key concepts Descriptions
Natural Nature language AI can use natural language to communicate using
interaction (structure of information, but it is hard to understand human’s
language and expression (e.g., metaphor, imagery, humor, emotion)
applications) AI can recognize but not experience emotions.
Common sense Appropriate responses to emotion are necessary to be
reasoning programmed by humans
Understanding Current AI systems are narrow reasoners to solve a
emotion well-defined problem. It has not yet achieved a very
Understanding flexible human-like reasoning called broad AI. In other
people/philosophy words, the systems lack consciousness and self-­
of mind awareness. A self-aware computer requires
representations of its own existence and thoughts and
memories of its past experience
Societal impact AI ethics, (The details are still under development by the AIK12)
limitations, risks,
bias
How AI can be
applied in positive
and negative ways

4.2.3 Other Review Papers in AI Literacy Education

Apart from the two renowned frameworks, other researchers from different coun-
tries/regions work rigorously to identify frameworks and timely review AI literacy
education.
In the UK, Lao (2020) presented a rubric to evaluate ML learning programs and
set up the basis for a set of standards in terms of knowledge, skills and attitudes:
• Knowledge: general knowledge, knowledge of methods (e.g., decision trees, neu-
ral networks, ensemble methods), bias in ML systems, and societal implications
• Skills: problem scoping, project planning, creating artifacts, analysis of design
interactions and results, advocacy, independent out-of-class learning
• Attitudes: interest, identity and community, self-efficacy, persistence
In Hong Kong, Su et al. (2022) examined 14 research papers on AI curriculum
for K–12 classrooms in the Asia-Pacific region that were taken from 2018 to 2021
by identifying the content knowledge, tools, platforms, activities, theories and mod-
els, assessment methods, and learning outcomes of the selected studies.
4.3 Rising Publications on AI Literacy Education 37

In Brazil, Gresse von Wangenheim et al. (2021) presented a 10-year systematic

mapping of emerging visual tools to support the teaching of machine learning in
terms of educational characteristics, deployment, and how the tools have been
developed and evaluated (e.g., Scratch, App Inventor, mblock, Google Teachable
Machine, RapidMiner).
In Spain, Rodríguez-García et al. (2020) presented the use of authentic projects
to enhance students’ higher cognition levels (e.g., creativity, critical thinking) using
machine learning models to apply their knowledge and propose solutions to their
problems.
And finally in Finland, Sanusi et al. (2022) identified the four main categories of
resources and tools (conversational agents, programming environment, robotic, and
unplugged activity) to document their uses to teach machine learning in K–12 set-
tings from 38 studies. Also from Finland, Tedre et al. (2021) found that students
could develop advanced concepts such as machine learning and natural language
processing through visualization of concepts using age-appropriate tools. They
reviewed 63 documents in the context of computing education to describe the chal-
lenges and 13 important characteristics that teachers/students faced in AI literacy
courses: (1) new classes of real-world applications for classroom experiments, (2)
shift from rule-driven to data-driven thinking, (3) change in the role of syntax and
semantics, (4) activities well aligned with modern pedagogy, (5) access to bodily
and natural language interaction, (6) a shift away from algorithmic steps, (7) higher
level of abstraction and black-boxed mechanisms, (8) a need for new notional
machines, (9) new models of testing and debugging, (10) new attributes of goodness
of programs, (11) deeper integration with STEAM subjects, (12) ability to explain
many services children use daily, and (13) direct connections to topical issues in AI
ethics. These studies have been conducted to measure different aspects of AI liter-
acy education in terms of pedagogies, tools, and curriculum. These reviews deal
with a number of research focuses: (1) how prepared teachers and schools are in
adopting technologies and conducting AI literacy education with a set of design
principles, (2) what types of technologies, pedagogies, and content knowledge
should be considered to design instruction to foster students’ AI literacy, and (3)
what are the educational standards that should be met by all students.

4.3 Rising Publications on AI Literacy Education

In the past, teaching AI was not possible in the past for K–12 students and noncom-
puter science undergraduates. Educators faced challenges in scaffolding K–12 stu-
dents to understand AI concepts through syntax-based programming (McCarthy,
2007; Wong et al., 2020). In recent years, the emergence of age-appropriate
38 4 The Landscape of AI Literacy

hardwares and softwares enabled educators to improve the learning process for
younger learners. The access to a wide range of technologies in day-to-day life,
such as chatbots and translation apps, presents opportunities for everyone to under-
stand and use AI in everyday life. This enables educators to leverage on the avail-
ability of AI technologies to inculcate AI literacy for young learners. For example,
prior studies discussed the potential to incorporate AI learning in K–12 STEAM
education via playful experience such as gamified and social media tools to prepare
students for future science, technology, engineering, art, and mathematics work-
forces (e.g., Chiu et al., 2021; Ng & Chu, 2021; Ng et al., 2021).
Knowing and using AI for future careers is only one aspect of teaching AI liter-
acy for educators. Any technology as potent as AI would also bring new risks due to
algorithmic bias and malicious uses of AI (Druga et al., 2019; Gong et al., 2020).
The importance of the roles of AI ethics is often underemphasized or even over-
looked, which is considered as extraneous or surplus to technical concerns in work
settings (Hagendorff, 2020). Software developers usually feel a lack of accountabil-
ity and moral significance of their work, especially when economic incentives are
easily overriding commitment to ethical principles and values (Hagendorff, 2020).
As such, educating both citizens and computer scientists AI ethics is essential to
strengthen their social responsibility and consider social inclusion and diversity to
apply AI for societal good (Dignum, 2019).
While the above examples show how substantial AI literacy education can
encompass, the arguments are scattered which make scholars difficult to find well-­
rounded conclusions for development of theoretical frameworks and pragmatic
instructional design principles. As such, we find that a systematic search on the
works of AI literacy education would be beneficial to elucidating the rising demand.
In search of literature on AI literacy, review articles were conducted in two stages.
First, we identified 30 peer-reviewed scholarly articles and conference papers from
K–12 to higher education levels published from 2016 to 2021 through the Web of
Science, Scopus, ProQuest Education Collection, IEEE, and ACM digital library
(Ng et al., 2021). A year later in August, the second stage further focused on each
school level and selected articles using the same databases to understand how to
incorporate AI literacy education across levels.
In Ng et al. (2021)’s review, the Google Scholar search identifies a dramatic
increase in AI literacy publications from 2016 to 2022 (see Fig. 4.1). As AI becomes
more and more important in work settings and everyday life, researchers began to
define AI literacy based on the term “literacy” which has been applied to define skill
sets in varied disciplines (Long & Magerko, 2020). Based on 30 selected studies,
the countries/regions that published 2 or more AI literacy articles include the United
States, China, Hong Kong, Spain, and Austria. Researchers conducted studies and
implemented AI literacy interventions across various educational levels. Most of the
articles focused on primary school and secondary school students that covered
almost half of the reviewed studies. Among these, 14 of the studies are from second-
ary school level. About one-third of the studies were conducted in an informal set-
ting, which included after-school programs, out-of-school activities, and poster
presentations. Seven studies were conducted in regular lessons in a formal setting.
4.3 Rising Publications on AI Literacy Education 39

Fig. 4.1 AI literacy articles from Google Scholar published by year (2016–2022)

One possible reason is that AI literacy is an emerging field, and most researchers
tend to conduct preliminary studies to explore their interventions in an informal set-
ting or merely write opinion papers based on their observation. Regarding the
research method, the empirical studies adopted qualitative methods (12), quantita-
tive methods, (5) and mixed-method approach (8) to evaluate students’ AI literacy
development.

4.3.1 AI Literacy Education for K–12 Students

The UNESCO report listed 11 countries that have developed government-endorsed

AI curricula in K–12 level. The motivations for the development are aimed to
improve graduates’ capacity, to meet the skills demanded by both the job market
and everyday interactions (UNESCO, 2022). They analyzed the curricula and sum-
marized them into nine topic areas of AI K–12 curriculum content and further
grouped them into three main categories: (1) AI foundations, (2) ethics and social
impact, and (3) understanding, using, and developing AI. As illustrated in the report,
AI literacy comprises both data literacy and algorithm literacy, and it serves as an
orientation of “knowledge, understanding, skills, and value of AI” (UNESCO,
2022, p. 11).
Other than the examples captured in the UNESCO report, educators have also
reported cross-sector projects on the development of AI curriculum. In Hong Kong,
Chiu et al. (2021) collaborated with the local government, secondary schools, and
industrial partners to create an AI curriculum at the secondary level named AI for
the Future (AI4Future). They proposed a curriculum framework that maps their
teaching units into five modules: (1) awareness, (2) knowledge, (3) interaction, (4)
empowerment, and (5) ethics (AKIEE) (Chiu et al., 2021). Xia et al. (2022) further
added inclusion and diversity of education into this AI initiative using
40 4 The Landscape of AI Literacy

self-determination theory to explain student engagement from the needs satisfaction

perspective. Likewise, Ng et al. (2021) proposed an AI curriculum called AI for All
in primary and secondary schools in Hong Kong to foster students’ AI literacy. Pilot
studies have been conducted to show the effectiveness of incorporating the curricu-
lum in informal settings through social networking sites (Ng & Chu, 2021), and
digital storytelling approach (Ng et al., 2022).
In Europe, Kandlhofer et al. (2016, p. 2) defined AI literacy as the understanding
of “the techniques and concepts behind AI products and services instead of just
learning how to use certain technologies or current applications”. They proposed
seven teaching topics to supplement this definition, including: automata, intelligent
agents, graphs and data structures, sorting, problem solving by search, classic plan-
ning, and machine learning. In the US, MIT’s AI experts, researchers, curriculum
developers and educators work to set the standard for how middle schoolers learn
about AI. They proposed an AI initiative called Responsible AI for Social
Empowerment and Education (RAISE) to establish a project-centered curriculum.
In the learning programs, students will produce authentic learning artifacts through
design thinking projects that are driven by exploring and doing. Along the way stu-
dents will deepen their understanding of AI literacy, computational action, and ethi-
cal thinking (MIT, 2021).
These curricula and frameworks are evidence to show that AI education has
moved from a subject specialized by computer science students to a direction where
it reaches all education levels and that it is diversified. However, it is noteworthy
that the above examples focus on K–12 level while little has been mentioned for
postsecondary-level students who have few chances to learn about AI.

4.3.2 AI Literacy Education for Noncomputer Science

University Students

To realize the motivations for the growing attention in AI education for noncom-
puter science undergraduates, Ashok Goel posited three converging factors that
would shape AI education to fit for twenty-first century workplaces – (1) the demand
created by increasing integration of AI technologies in industries, (2) the rapid
diversification of AI education that used to be offered in graduate studies and now
spread widely across disciplines and educational levels, and (3) the need for all citi-
zens to be literate in AI such that they aware of the use of AI technologies in their
daily lives and be informed of the principles and values behind (Goel, 2017).
The evolution of AI in the past two decades unfolds a broad spectrum of techno-
logical applications that affect our daily lives, in particular the rapid development of
artificial neural networks (ANN) and deep learning (DL) (Chan & Zary, 2019).
Implications are profound, from improving product efficiency, using big data for
business strategies and analyses, to individualizing user experiences, etc. For
instance, personal mobile devices are a typical example of applying AI features like
4.4 New Education Policies on AI Literacy Across the Globe 41

face recognition for unlocking screens and online payments. The game industry cre-
ates AI characters, or non-player characters (NPCs), to make virtual reality experi-
ences more realistic. Educators use generative adversarial networks (GAN) to create
new images, videos, or styles for drawing learning. The list goes on. The demand of
AI is no longer a privilege to advanced tech companies but an indispensable part of
existing and uprising industries. Corporates are looking for future-ready employees
who possess high command in tech and AI literacies (Microsoft News Center India,
2022). UNESCO released a report on the global status of AI curricula and described
“AI as the basic grammar of our century” (UNESCO, 2022, para. 5). Hence, the
need to push forward AI education is not only important for the growth of society
but also a matter of urgency.
The report concluded that “all citizens need to be equipped with some level of AI
literacy covering the values, knowledge and skills relating to AI” (UNESCO, 2022,
p. 61). This notion is agreed by other scholars (e.g., Kong et al., 2021; Long &
Magerko, 2020), but they also noted that limited efforts have been made to promote
AI literacy for citizens. To generate a structure to classify the data that may emerge
from this review and for it to be meaningful, we may address this by looking through
the lens of AI literacy. Recent studies have conceptualized the term “AI literacy” by
defining what it means, what it comprises, and how it should be assessed; however,
there is no universal consensus to a single definition to date. Therefore, Sect. 4.4
summarizes education policies of AI literacy education across the countries/regions,
and Sect. 4.5 further categorizes the existing evidence and proposes frameworks on
AI literacy, based on three existing consensus educational models.

4.4 New Education Policies on AI Literacy Across the Globe

Learning from the existing frameworks and rising demands of AI literacy men-
tioned by scholars and practitioners, we see a paradigm shift on the significance of
this matter. Not only AI literacy education should be advocated at school and insti-
tutional level but also influenced by government or even global policies. This sec-
tion captures the AI literacy education policy in countries with more than two AI
literacy publications in secondary level (i.e., the United States, China, Hong Kong,
Spain) and neighboring countries (e.g., Japan, Singapore, Korea) in Ng et al.
(2021)’s reviews.
Both the United States and China have made great progress in incorporating AI
education into their workforce development and K–12 education systems. However,
they are approaching education goals in different ways. The United States is devel-
oping AI curricula through industry and university collaboration, whereas China is
using its centralized authority to mandate AI curricula in its K–12 education with
the support from AI companies that partnered with schools and universities to train
students (CSET, 2021). The United States had long led in tech innovation despite
strong global competition. Many tech companies (e.g., Microsoft, Apple, Amazon)
and organizations (e.g., AIK12) help promote AI literacy education which makes 28
42 4 The Landscape of AI Literacy

states in the country adopt policies to support K–12 computer science education in
2021. In China, the Education Ministry has introduced AI into the K–12 school cur-
riculum with the first AI textbook supported by SenseTime to learn the basics of
image recognition, sound recognition, text recognition, and deep learning across 40
pilot schools in 2018.
In Europe, Spain is also a country with rich research publications in AI literacy
education. In 2019, the European Commission has developed the scope of an
Erasmus+ project called AI+, which aims to develop an AI curriculum for high
school students. The project also supported some neighboring European countries
(e.g., Italy, Slovenia, Lithuania, Finland). A year later, another scheme called
LearningML was presented to bring the fundamentals of AI to students and people
who are interested in acquiring related knowledge. Both initiatives are driven from
universities that encourage learners to learn AI literacy at a younger age.
Although other Asian countries did not have many AI literacy publications, their
practices and implementation are worthy for policymakers to think behind for
referencing.
• Japan: In 2020, a national curriculum mandated CS education is designed to
prepare the K–12 students capable of understanding and using AI technologies in
their future to maintain the competitiveness of Japan as one of the leading coun-
tries in the AI-driven world.
• Singapore: In 2018, the government announced the “AI Singapore” project to
develop students’ AI capabilities. The project had brought together the research
institutions including AI startups and companies to develop AI learning products
to grow K–12 students’ AI knowledge and develop talents to power Singapore’s
AI efforts. At the same time, two AI research programs, “AI for Students” and
“AI for Kids” had also started.
• Korea: In 2019, the Korean government announced the “National AI Strategy” to
enhance the country’s AI competitiveness by 2030. The government had made
plans to introduce AI to all high school students in 2021 and further extended AI
education to kindergartens and primary and middle students by 2025.
• Taiwan: The Ministry of Education (MOE) is promoting AI learning to reshape
the school environment to make AI an integral part of students’ lives. In 2019,
the MOE announced the compulsory integration of AI-related educational mate-
rials in public school curriculums from elementary to high school.
Hong Kong has also incorporated AI curricula in K–12 education for a few years.
In 2020, an AI curriculum called “AI for Future” was implemented into pre-­
university education to cultivate the competitiveness of the young generation among
18 schools. A year later, the Hong Kong Education City announced the “Go AI
Scheme” which aims to promote AI Education in Hong Kong through introducing
self-paced learning platforms and enhancing teachers’ and students’ AI knowledge
and future-ready skills in the twenty-first century. Local researchers (e.g., Chai
et al., 2021; Chiu et al., 2021; Ng et al., 2021; Wang & Cheng, 2021) have worked
vigorously to promote AI literacy education across educational levels from different
perspectives.
4.5 Our Three Proposals of AI Literacy Educational Frameworks (Bloom’s, TPACK,… 43

4.5 Our Three Proposals of AI Literacy Educational

Frameworks (Bloom’s, TPACK, P21)

This section builds on a previously developed ideas based on our three recent pub-
lished systematic reviews (i.e., Ng et al., 2021) reflecting on three classic educa-
tional theories (i.e., Bloom’s taxonomy, twenty-first century literacy skills,
Technological, Pedagogical and Content Knowledge framework). We hope to
extend the existing theories to underlie the theoretical contributions and significant
advances in AI literacy education studies. First, we applied the Bloom’s taxonomy
to conceptualize varied cognitive levels across the spectra of AI learning inquiry
with four perspectives, namely, know and understand, use and apply, create and
evaluate, as well as ethical issues. Then, the Technological, Pedagogical and Content
Knowledge (TPACK) framework was also explored in terms of learning artifacts,
pedagogical approaches, and subject matters to offer effective means to integrate AI
literacy into school curricula and how educators help students develop AI literacy
understanding. Lastly, we discuss the need to add AI to twenty-first century literacy
in work settings and everyday life and as a fundamental skill for everyone, not just
for computer scientists.

4.5.1 AI Literacy and Bloom’s Taxonomy

A definition for AI literacy learning is presented in four cognitive domains. Ng et al.

(2021) proposed a set of the abilities and skills that has potential to be mapped to the
cognitive domains in Bloom’s taxonomy, an approach to categorize the levels of
reasoning skills and ordered thinking required across different learning contexts.
There are six levels in the taxonomy, each requiring a higher level of complexity and
ordered thinking from the students. The levels are understood to be successive, so
that one level must be mastered before the next level can be reached (Huitt, 2011).
The reason why we adopted the Bloom architecture is that AI literacy is novice to
educators and a classification of levels of cognitive processes has not yet been
developed in the context of AI learning. However, this model is a classic pedagogi-
cal theory that establishes the core foundation of AI taught to young learners. In this
review, it is proposed to assign these three aspects (i.e., know and understand, use,
and evaluate and create AI) into the cognitive levels of Bloom’s taxonomy. “Know
and understand AI” is assigned to the bottom two levels, “use AI” in applying con-
cepts and applications is assigned to the apply level, and “evaluate and create AI”
are assigned to the top three levels to analyze, evaluate, and create AI (see Fig. 4.2).
The AI literacy studies discussed how to foster learners’ AI literacy in knowing
and understanding AI, as well as how to use AI applications in everyday life and
apply its underlying concepts in different contexts. However, according to Ng et al.
(2021)’s review, not many studies mentioned how to enhance students to analyze,
evaluate, and create AI applications through higher-order thinking activities. A
44 4 The Landscape of AI Literacy

Fig. 4.2 AI literacy and Bloom’s taxonomy

Table 4.4 Framework of AI literacy (Ng et al., 2021)

AI literacy Definitions Sample studies
Know and Know the basic functions of AI and how to Kandlhofer et al. (2016);
understand AI use AI applications Robinson et al. (2020)
Use AI Applying AI knowledge, concepts, and Druga et al. (2019); Julie et al.
applications in different scenarios (2020); Vazhayil et al. (2019)
Evaluate and Higher-order thinking skills (e.g., evaluate, Druga et al. (2019); How and
create AI appraise, predict, design) with AI Hung (2019)
applications
AI ethics Human-centered considerations (e.g., Chai et al. (2020); Druga et al.
fairness, accountability, transparency, ethics, (2019)
safety)

possible reason that existing AI literacy studies focused more on general skills and
knowledge about AI is that AI literacy is a set of fundamental skills and abilities in
helping everyone, including students and citizens, to acquire, construct, and apply
knowledge. They may not necessarily handle how to abstract and decompose AI
problems nor build AI applications in everyday lives; instead, they need to know the
basic concepts and use AI ethically. As such, it is noticed that prior AI literacy stud-
ies put more emphasis on engaging learners in lower-level thinking activities.
However, when students are promoted to secondary schools and universities, they
become knowledgeable to apply their prior knowledge to create their own artifacts
and justify decisions with AI applications and algorithms for their future career. In
summary, four aspects of fostering AI literacy were identified from the review (see
Table 4.4).
Know and Understand AI Prior studies conceptualize AI literacy as educating
learners about acquiring fundamental concepts, skills, knowledge, and attitudes that
4.5 Our Three Proposals of AI Literacy Educational Frameworks (Bloom’s, TPACK,… 45

require no prior knowledge. On top of being the end users of AI applications, learn-
ers should understand the technologies behind. Burgsteiner et al. (2016) and
Kandlhofer et al. (2016) defined AI literacy as the ability to understand the basic
techniques and concepts behind AI in different products and services. Moreover,
some researchers associate AI literacy with perceived abilities, confidence, and
readiness in learning AI. In K–12 education, Druga et al. (2019) and Lee et al.
(2021) designed learning curriculums and activities that foster AI literacy that
focuses on how learners gain AI concepts.

Use and Apply AI Prior research emphasized the importance of educating learners
to know how to apply AI concepts in different contexts and applications in everyday
life. For example, Rodríguez-García et al. (2020) evaluated LearningML, a machine
learning model builder, to educate citizens to understand AI applications and how it
can affect our lives, as well as knowing the ethical issues regarding AI technologies.
In addition, it is identified that half of the studies in Ng et al. (2021)’s review dis-
cussed the human-centered and ethical considerations and focused on using AI con-
cepts and application ethically. Eight out of 30 articles borrowed the ideas of
computational thinking to interplay AI literacy and AI thinking (Ng et al., 2021)
(see Table 4.5). AI thinking refers to the construction of logic and algorithms in
order to support students’ understanding of how to use knowledge bases for
problem-­solving, processing semantics and handling unstructured data (Vazhayil
et al., 2019). For example, How and Hung (2019) leveraged AI thinking through
conducting data analytics with computing and interpreted new findings from the
machine-learned discovery of hidden patterns in data.

Evaluate and Create AI AI augments human intelligence with digital automation

and articles alluded to AI literacy to engage learners in higher-order thinking activi-
ties. Two-thirds of AI literacy studies in Ng et al. (2021)’s review had extended AI
literacy to higher-order thinking abilities that enabled individuals to critically evalu-

Table 4.5 AI thinking elements

Elements Descriptions Examples
AI concepts Technical and conceptual Understand the basic AI concepts and their
understanding of the basic AI origins such as machine learning, deep learning
concepts and neural network
AI practices The techniques and strategies Appreciate the real-world applications of AI
used when applying AI concepts such as speech recognition, robotics
Training, validation, and testing
Remixing or reusing code
AI Attitudes and dispositions Collaborating to solve problems, understanding
perspectives adopted while solving of technology as a problem-solving tool
problems Consider the ethical and safety concerns when
applying AI technologies in real-world
applications
46 4 The Landscape of AI Literacy

ate AI technologies, communicate, and collaborate effectively with AI (e.g., Long &
Magerko, 2020). For example, Han et al. (2019) enhanced students’ scientific and
technological knowledge which then was applied in scientific research-based learn-
ing to solve practical problems. Long and Magerko (2020) engaged citizens in co-­
creating AI amenities in public spaces to broaden their public AI literacy and
experiences. Participants could engage with public interactive artworks’ progress
sequentially from being initially attracted to an AI-enabled installation to relate
their interaction with the installation and other people.
Overall, although there are slight variations on the definition of AI literacy, prior
studies support the notion that everyone, especially K–12 students, should acquire
basic AI knowledge and abilities, enhance motivation for future career, as well as
use AI-enabled technology ethically (Chai et al., 2021). In terms of cognitive
domain, AI literacy serves as a set of competencies that enables individuals to know
and use AI ethically, critically evaluate AI technologies, communicate, and collabo-
rate effectively with AI (Long & Magerko, 2020).

4.5.2 AI Literacy and TPACK Framework

Inspired from the Technological, Pedagogical and Content Knowledge (TPACK)

framework, this section evaluates how educators help learners develop AI literacy
skills in terms of pedagogy, content, and pedagogy (see Fig. 4.3). This model
enriches the effective means to integrate AI literacy into school curricula and how
educators help learners develop AI literacy. The reason that we adopt the TPACK
model is that it is widely used across studies to identify how teachers can incorpo-
rate technologies into their pedagogical methods and content knowledge and con-
ceptualizes their capacity and knowledge that is needed to integrate relevant
technologies in AI literacy education (Koehler et al., 2013). It provides a map for
understanding how to integrate AI literacy into classrooms effectively (e.g., Celik,
2022; Kim et al., 2021). For example, Kim et al. (2021) based on AI learning
resources to conceptualize TPACK to improve teaching for K–12 AI education,
which offers core foundations of AI taught to young learners. Among the three
knowledge, technological knowledge involves the affordances and use of domain-­
specific learning tools such as hardware and software in AI literacy education,
AI-enabled tools (e.g., intelligent agents), and unplugged learning tools (e.g., role-­
playing). Second, pedagogical knowledge relates to teaching methods and their
application to promote student AI literacy learning, which entails teaching strate-
gies and scaffolding, feedbacking students’ learning processes (Janssen et al.,
2019). Third, content knowledge concerns knowledge about the AI literacy subject
matter that specific subjects should be covered in the curriculum.
Technological Knowledge Given the complexity of AI, age-appropriate learning
artifacts were important to scaffold students’ AI conceptual understandings and
stimulate their motivation and interest in learning AI. Table 4.6 provides an over-
4.5 Our Three Proposals of AI Literacy Educational Frameworks (Bloom’s, TPACK,… 47

Fig. 4.3 AI literacy TPACK framework

view of the types of AI learning artifacts ranging from hardware to software-focused

artifacts, intelligent agents, and unplugged learning tools. In recent years, there has
been an increase in hardware and software that enhance AI concepts accessible to
younger learners. Hardware- and software-focused artifacts engaged students to use
physical and digital artifacts to make and create AI-driven applications. For exam-
ple, Chiu et al. (2021) designed a robotic car called CUHKiCar with built-in AI
functions to offer interactive learning experiences for students to do face-tracking
and line following tasks. It is found that students obtained a significant improve-
ment in perceived knowledge gain, confidence, motivation, and AI readiness.
Another study conducted by Chai et al. (2021) who used a platform in which stu-
dents could use Alpha dog robots and design algorithms to do recognitions of physi-
cal characteristics such as temperature, voice, face, and images. It enables students
to use AI and mathematics knowledge (e.g., calculus, statistics) to build their AI
solutions.
The popularity of current AI technologies encourages students to make intelli-
gent agents and machine learning models without needing to program such as
ML-for-kids and Teachable Machine (Long & Magerko, 2020). In this context,
there is an opportunity for educators to provide students access to AI literacy and
reinforce the AI concepts through these emerging tools. In addition, AI-driven tools
such as chatbot, writing assistants, and web mapping encourage students to experi-
ence the technological affordances of AI applications across subject disciplines.
This enables students to express knowledge using AI and access the more advanced
concepts which was not possible in the past. Alternatively, unplugged learning
48 4 The Landscape of AI Literacy

Table 4.6 Learning artifacts

Definition Learning artifacts examples Sample studies
Hardware-­ Use physical artifacts to Bee-bots, LEGO Mindstorms Kandlhofer et al.
focused learn AI such as robotics, NXT, Cubelets, alpha dog (2016); Chai et al.
artifacts sensors and Arduino robot, Kinect LuminAI, VR (2021); Long and
devices Robot Improv Circus, Sound Magerko (2020);
Happening, Shape of Story Druga et al. (2019);
AI home assistants: Jibo robot, Burgsteiner et al.
Anki’s Cozmo robot, and (2016)
Amazon’s Alexa
Lego Mindstorms NXT
Software-­ Use digital artifacts to Google maps, Golog, YAGI, Kandlhofer et al.
focused learn AI such as block-/ ASRAEL (2016); Wan et al.
artifacts syntax-based SmileyCluster, A* algorithm in (2020); Burgsteiner
programming and C# et al. (2016)
simulation
AI-related Use intelligent agents Scratch, Google’s Teachable Lee et al. (2021);
agents such as expert systems, Machine, Generative Vazhayil et al.
machine learning trainers, Adversarial Networks (2019); How and
and chatbots to build their (GANS), Watson AI services, Hung (2019); Druga
custom machine learning BayesiaLab, AI home et al. (2019)
models without coding assistants: Jibo robot, Anki’s
Cozmo robot, and Amazon’s
Alexa
Unplugged Use learning activities to Lectures, career talk, textbook, Lee et al. (2021);
learn AI without a case study, webinar, role-­ Rodríguez-García
computer such as lecture, playing, storytelling et al. (2020); Julie
case study, role-playing, et al. (2020)
and storytelling

activities were designed to foster students’ AI literacy without using a computer

through engaging approaches such as case study, role-playing, and storytelling
(e.g., Julie et al., 2020; Rodríguez-García et al., 2020).
Pedagogical Knowledge The pedagogies including teaching methods and strate-
gies are classified according to the levels of education. One of the aims of AI liter-
acy education for secondary schools is to familiarize students with the fundamental
concepts of AI/computer science and encourage them to discover the connection
between AI applications and the underlying concepts. For example, researchers
introduced students to AI concepts in playful and inquiry approaches through robot-
ics making (Kandlhofer et al., 2016), performing Turing test with intelligent agents,
creating chatbot and inference algorithms (Wong et al., 2020), and building applica-
tions through blockly-based programming (Gong et al., 2020). In addition to under-
standing the connection between those AI techniques and common AI applications,
secondary students should have the abilities to apply prior AI knowledge in practi-
cal group projects to analyze and solve problems independently (Kandlhofer et al.,
2016). Thus, educators could design real-world, collaborative projects based on the
principles of constructionism and instructionism (Kandlhofer et al., 2016).
4.5 Our Three Proposals of AI Literacy Educational Frameworks (Bloom’s, TPACK,… 49

Researchers suggest various hands-on activities such as robot constructions

(Williams et al., 2019), data and comparative visualization (Wan et al., 2020), as
well as training AI models (Vazhayil et al., 2019) as possible means to promote AI
literacy in secondary school levels.
This section identifies the major three pedagogical methods and strategies that
are used in this dissertation: project-/problem-based learning, collaborative learn-
ing, and experiential learning via playful games and simulations (see Table 4.7).
First, project-/problem-based learning is the learning approach that engages stu-
dents to gain knowledge and skills by working to investigate authentic questions,
problems, or challenges (Kokotsaki et al., 2016). For example, Vachovsky et al.
(2016) engaged 24 girls in projects to learn computer visions, robotics, and natural
language processing in a summer camp. In the survey, 95.8% of students believed
that the projects that students built can help the society. Moreover, students believed
that the course was interesting (83.3%) and enhanced their confidence in using AI
(75%). Rodríguez-García et al. (2020) presented the LearningML projects (a low
floor high ceiling platform to learn machine learning through doing) to bring the
fundamentals of machine learning (ML) to students to acquire its knowledge and
educate them to become critical thinking citizens. Project-/problem-based learning
could provide authentic settings such as building modals of athletic moves
(Zimmermann-Niefield et al., 2019), meaningful scientific and STEM contexts that
could motivate them to learn AI through a sense of authenticity and real-world
applicability (Wan et al., 2020).
Collaborative learning allows students to learn how to communicate and work
with classmates to gain AI knowledge and manipulate with smart devices (Roll &
Wylie, 2016). For example, Gong et al. (2020) engaged students to take different
roles such as project managers, software designers, hardware designers, and art
designers to build smart vehicles in authentic settings. Another study conducted by
also asked students to be buyers and sellers to find out problems and shortcomings
about the intelligent functions in the smart home systems. Kaspersen et al. (2021a)
asked students to formulate three to four students of a group to design ML models
to predict whether a person will vote for a particular political party. A combination

Table 4.7 Pedagogies of AI literacy education

Pedagogies Descriptions Sample studies
Project-/ Learning approaches to engage students to gain Rodríguez-García et al.
problem-based knowledge and skills by working to investigate (2020); Sakulkueakulsuk
learning authentic questions, problems, or challenges et al. (2018)
Collaborative This pedagogy allows students to learn how to Deng et al. (2021); Gong
learning communicate and work with classmates to gain et al. (2020)
AI knowledge and manipulate with smart devices
Experiential Process of learning by doing through hands-on Chiu et al. (2021); Morris
learning experiences and reflection; students could gain (2020); Tamborg et al.
better understanding to connect theories and (2022)
knowledge learned in the classroom to real-world
situations
50 4 The Landscape of AI Literacy

of collaborative learning and project-/problem-based learning could effectively

enhance students’ high-order thinking skills such as problem-­solving, leadership
skills, project management, and creativity (e.g., Deng et al., 2021; Gong et al.,
2020). The third common method is to engage students in experiential learning.
Experiential learning is the process of learning by doing through hands-on experi-
ences and reflection; students could gain better understanding to connect theories
and knowledge learned in the classroom to real-world situations (Morris, 2020).
This allows students to explore what ML is via Teachable Machine (Chiu et al.,
2021; Tamborg et al., 2022), Code.org games (Ng & Chu, 2021), intelligent agents,
chatbot, syntax-based programs (e.g., Python) (Gong et al., 2020; Gunasilan, 2021),
and Blockly-based programs (e.g., Scratch, Snap!) (Estevez et al., 2019; Kahn et al.,
2018). These activities provide students hands-on experience to explore what AI is,
visualize advanced concepts, and build ML models (Reyes et al., 2020). Although
these activities offer hands-on experience to scaffold their AI understandings, most
studies further applied minds-on collaborative projects to encourage students to fur-
ther build knowledge through making digital and tangible artifacts in constructionist
approaches. In this way, educators could enable students to reach higher cognition
levels and apply AI skills and knowledge to solve real-world problems for future
learning and career challenges (Chai et al., 2021).
Content Knowledge In recent years, governments and universities have worked
vigorously to design meaningful K–12 learning curricula and activities that focus
on different AI concepts and how they apply AI to contexts of their interests (Druga
et al., 2019; Lee et al., 2021; UNESCO, 2022). For example, Long and Magerko
(2020) proposed 16 competencies that students should learn: recognizing AI, under-
standing AI, interdisciplinary skills, distinguishing general and narrow AI, identify-
ing strengths and weaknesses of AI, imagining future applications of AI and their
societal impacts, knowledge representations, decision-making, understanding
machine learning, recognizing the roles of human in AI, data literacy, learning from
data, critically interpreting data, higher-level reasoning of AI, sensors, and ethical
concerns behind. Touretzky et al. (2019)’s five “big ideas” of AI have set a sound
framework for future research on fostering AI literacy:
• Perceptions: Computers perceive the world using sensors.
• Representation and reasoning: Agents maintain representation of the world and
use them for reasoning.
• Learning: Computers can learn from large amounts of data.
• Natural interaction: Intelligent agents require many kinds of knowledge to inter-
act naturally with humans.
• Societal impact: AI can impact society in both positive and negative ways.
Wong et al. (2020) categorized AI literacy in K–12 into three dimensions: AI
concepts, applications, and ethics. In another study, Chiu et al. (2021) further pro-
posed five modules of AI learning: (1) awareness, (2) knowledge, (3) interaction, (4)
empowerment, and (5) ethics. These modules can be categorized into the beginner,
intermediate, advanced levels, and caters for capacity building by offering a clear
4.5 Our Three Proposals of AI Literacy Educational Frameworks (Bloom’s, TPACK,… 51

path to the development of student AI techniques and skills. UNESCO (2022) ana-
lyzed the government-endorsed curricula in 11 member states and pointed out 3
major categories that AI curricula should have: (1) AI foundations including algo-
rithms and programming, data literacy, and contextual problem-solving; (2) AI eth-
ics, societal implications, and the applications of AI to other domains; and (3)
understanding and using AI techniques, understanding and using AI technologies,
and developing AI.
However, very few prior studies do not categorize the topic areas according to
grade levels. UNESCO (2022) tried to map the learning outcomes of AI curricula
from the member states for each education level. It is important for educators to
understand the cognitive development of each developmental stage, and we pro-
posed the use of Bloom’s taxonomy to categorize the learning contents. To under-
stand what should be taught at the secondary level, prior studies documented that
junior secondary students should focus on preliminary and simple AI concepts such
as machine learning, natural language processing, and Turing tests (AIK12, 2022;
Chiu et al., 2021). Educators should design experiential learning activities for stu-
dents to have a hands-on experience to taste and use the related AI applications and
discuss their benefits, challenges, ethical concerns, and shortcomings of these tools
(Sabuncuoglu, 2020). In this way, students needed to apply these knowledge and
skills to solve problems using well-defined hardware, software, and intelligent
agents. Table 4.8 illustrates the learning contents that were taught in junior and sec-
ondary school levels.

4.5.3 AI Literacy and P21’s Framework for the 21st

Century Learning

Most of the studies focused on students’ technological knowledge and skills; how-
ever, few of them identify that AI literacy should extend to broader digital compe-
tencies that support students to use AI technologies to facilitate their learning. This
section highlights the notion that AI literacy should be viewed as an important
twenty-first century skill set (Ng et al., 2021). Twenty-first century skills and AI
digital competencies are both concepts that focus on a broad spectrum of knowl-
edge, skills, and attitudes that were viewed as essential components of digital liter-
acy in our digital world. Such digital aspects should go beyond technical use and
focus more on a more holistic understanding that recognizes other contextual, criti-
cal, and complex literacy. In other words, although the term “digital literacy” con-
sists of “digital”; the digital aspect is often seen as a discrete skill, implying that the
twenty-first-century skills are not necessarily underpinned by ICT. The focus should
be more on knowledge- or content-related skills, instead of technical skills required
for the workforce, such as life and career skills, multidisciplinary skills, and learn-
ing and innovation skills (National Research Council, 2012; Van Laar et al., 2017).
52 4 The Landscape of AI Literacy

Table 4.8 Learning contents in primary, junior, and senior secondary school levels
Levels Learning contents samples Sample studies
Primary Experiencing AI: Interacting with AI machines such as Heinze et al. (2010);
school driving an AI toy car Narahara &
AI foundations: Knowing and understanding basic Kobayashi (2018)
concepts, vocabulary, history of AI, facial recognition and
machine learning, hands-on experience, using applications
(e.g., Google’s Teachable Machine), and programming
tools to solve problems
Societal impacts and AI ethics: AI ethics, societal impacts
of AI, algorithmic bias
Junior Experiencing AI: Using AI applications, benefits, and Chiu et al. (2021);
secondary disadvantages of using AI, machine learning, face Fernández-Martínez
recognition, image stylizer, machine generation of creative et al. (2021); Ng and
content, experimentation of using AI technologies Chu (2021)
AI foundations: History/introduction/recent development
of AI and its subareas, differences between humans and
machines, natural language processing, scratch activities
on machine learning and image recognition, machine
learning
Societal impacts and AI ethics: AI ethics, societal impacts
of AI, machine reasoning and its bias
Senior Complex AI topics: Natural language processing, computer Kaspersen et al.
secondary vision, cognition, biomedical informatics, robotics, (2021b); Kahn et al.
information networks, human-robot interactions, (2018); Zhang & Du
computational sustainability (2008)
AI technical components: Fisher’s exact test, inductive
reasoning, nearest neighbor algorithm, correlation, graph
search algorithms, computational game theory,
optimization, agent-based modeling, probabilistic
reasoning
AI literacy: Understanding how ML works, the process
behind creating ML models, and the ability to reflect on its
personal and societal implications

Based on the P21’s Framework for the 21st Century Learning, this section dis-
cusses the potential of adding AI literacy to map different components of the frame-
work (see Fig. 4.4). The P21’s Framework lists three types of skills: learning and
innovation skills (creativity, critical thinking and problem-solving, communication
and collaboration), literacy skills (information, media, and ICT literacy), and life
skills (flexibility and adaptability, initiative and self-direction, social and cross-­
cultural skills, productivity and accountability, and leadership and responsibility)
that are required in core subjects and twenty-first century themes.
Information, Media, and Technology Skills The twenty-first century skills cover
digital literacies, and AI literacy belongs to one of the digital literacy skills. There
are various information and technological skills involved when learning AI. First,
students can accomplish learning tasks in an AI-driven environment. Students learn
the characteristics of AI-empowered devices and applications to make use of them
in their day-to-day activities. Moreover, various AI systems such as recommenda-
tion systems, intelligent agents, and advanced AI algorithms can help people facili-
4.5 Our Three Proposals of AI Literacy Educational Frameworks (Bloom’s, TPACK,… 53

Fig. 4.4 AI literacy as a twenty-first century skill

tate information search and retrieval. Students can also critically judge the usefulness
and sufficiency of AI-generated advice and information. Furthermore, various social
media platforms use AI to analyze key moments in photos and live videos which
make users more personalized and customized resulting in higher engagement (Vale
& Fernandes, 2018). AI can also help automate video creation and adjust their video
contents and elements seamlessly (e.g., colors, audio tracks, video) and speed up
their editing process using mobile devices. AI tools can help interact with users and
customers to enhance their user experience.

Learning and Innovation Skills The skill sets (or the 4 Cs) of twenty-first century
learning include (1) critical thinking and problem-­solving, (2) creativity and inno-
vation, (3) communication, and (4) collaboration. First, AI literates can use appro-
priate AI applications to solve authentic problems and make critical decisions. In
Yoder et al. (2020)’s study, students were asked to utilize particular AI algorithms
and leverage-related techniques to solve real-world problems such as building a
contact tracing application for the pandemic. Another study conducted by Long and
Magerko (2020) encouraged students to collaboratively create music together using
AI-generated music machines through changing parameters.
54 4 The Landscape of AI Literacy

Communication requires students to transmit ideas, information, and knowledge

with others effectively using a variety of AI tools (e.g., chat bots, diagnostic evalu-
ation). Huo (2019) proposed an online diagnosis evaluation program to assist sec-
ondary students to improve their English presentations. A diagnostic evaluation of
presentation skills can help students examine their strengths and weaknesses, pre-
sentation content, style, and length to optimize their communication. On a higher-­
level application of AI, some studies have tried to incorporate these elements as
important curriculum standards to cultivate AI and computational thinking (Deng
et al., 2021).
Life and Career Skills AI literacy is more than technological understanding which
should involve a range of critical, reflective, and social perspectives (Gut, 2011).
Therefore, skill sets such as ethical awareness, flexibility and adaptability, initiative
and self-direction, cultural awareness, lifelong learning, productivity and account-
ability, and leadership and responsibility are important to enhance students’ com-
petitiveness and capacity in our global knowledge economy. First, students should
learn AI-related awareness and knowledge about legal and ethical aspects. Several
key components (e.g., fairness, responsibility, transparency, trust) were identified in
prior studies to learn about the legal, ethical, and cultural concerns of personally and
socially responsible use of AI. Students should understand the potential risks and
evaluate the impact of AI in social, economic, and cultural contexts (Hagendorff,
2020). Moreover, since AI literacy is important for today’s digital era, students have
to adapt to such AI digital transformation and get ready for the future workforce.
Other contextual twenty-first century skills (e.g., self-direction, flexibility, lifelong
learning) are also vital for students to adapt to their thinking and attitude to the fast-
changing society, managing learning progression toward these goals, and exploring
new opportunities when using AI (Goralski & Tan, 2020). In other words, students
are not only end users of AI technologies who know the basic concepts behind. They
will grow up to become responsible citizens who know how to use AI and work with
AI to make people and our society a better place to live in.

Core Subjects: 3Rs and Twenty-First Century Themes To prepare students to

be successful in the future, students are required to master the core subjects and
twenty-first century themes. In fact, AI has been incorporated across subject knowl-
edge (e.g., language, arts, mathematics, economics, science) and other twenty-first
century themes (i.e., global awareness, environmental, financial, civic, and health
literacy). Educators can discuss how AI influences our everyday life and industries.
For example, in language education, Zhang (2018) developed young learners’ AI
literacy through interacting with an AI-empowered robot to practice English speak-
ing. Moreover, students can learn the potential of using machine learning to help
disease prediction and diagnosis and facilitate treatment effectiveness and drug dis-
covery (Noorbakhsh-Sabet et al., 2019). It is meaningful to engage students in
expressing knowledge from other disciplines and equip them with insights into
AI-empowered modeling, data analysis, and problem-­solving such as predicting the
quality of mangoes in STEM education (Sakulkueakulsuk et al., 2018), using drones
4.6 Conclusion 55

to help traffic management (Lundberg et al., 2018) and practice AI and computa-
tional thinking in mathematics education (Tamborg et al., 2022).
To summarize, through integrating three essential AI digital competencies (i.e., (1)
information, media and technology skills; (2) learning and innovation skills; and (3)
life and career skills), into different core subjects, this section helps establish a con-
ceptual twenty-first century skills framework and propose different digital skill
dimensions by evaluating research articles that define the skill sets. It has resulted in
a framework of core skills including technical, information skills, communication,
collaboration, creativity, critical thinking and problem-solving, as well as other con-
textual skills such as ethical awareness, cultural awareness, flexibility, self-direc-
tion, and lifelong learning. This framework informs policymakers to meet the
educational standards in their countries/regions and provides related professional
development for educators to move forward. The framework also serves as a basis
of educational reform and digital transformation across educational institutions. In
a fast-changing knowledge economy, these skills bring learning opportunities for
students to enhance their competitiveness and capacity to drive innovation, develop
twenty-first century literacy skills to cope in this changing society, and fulfill the job
demands. AI literacy education is still in its infancy stage, and the existing research
papers are not mature enough to uncover all necessary skills at our ever-changing
times. As such, this section is suggestive that AI literacy plays an increasingly
important role as part of the twenty-first-century digital skills. A more detailed
structure and operational instructions on incorporating this notion require further
exploration.

4.6 Conclusion

In recent 5 years, countries/regions and K–16 educational institutions have started

to design related policies, curricula, and learning programs to promote AI literacy
for students. We start to see educational frameworks that emerge in supporting the
habitat for AI literacy education to blossom (e.g., Long & Magerko, 2020; AIK12,
2022). Other scholars also wrote review articles to summarize the important find-
ings. They identified certain technologies, learning content, pedagogy, and educa-
tional standards to be appropriate for AI literacy education. By our research team, a
scoping review was also conducted to give an overview of AI literacy education
implemented from 2016 to 2021.
This chapter examined how AI literacy interacts with twenty-first century skills
by looking through the lens of three proposed educational models (i.e., Bloom,
TPACK, and P21). The results demonstrated the use of Bloom’s taxonomy to view
cognition levels of AI literacy education, TPACK model to understand how educa-
tors should learn to equip their students with AI literacy, and the P21’s Framework
to identify the four types of digital competences in the twenty-first century. As sug-
gested by the UNESCO (2022, p. 61), “all citizens need to be equipped with some
56 4 The Landscape of AI Literacy

level of AI literacy covering the values, knowledge and skills relating to AI.”
However, there is a lack of studies investigating what and how to teach AI at a spe-
cific educational level. The results in this chapter can formulate a theoretical basis
for us to further investigate how AI literacy can and should be conducted at each
educational level. Our proposal serves as guidelines to define what K–16 students
should know about AI and assist different stakeholders (e.g., curriculum designers,
AI developers, policymakers) to learn the standards of essential AI knowledge,
skills, attitudes, values, and ethics across grade bands.

References

AIK12. (2022) Grade band progression chats. Retrieved from https://ai4k12.org/

gradeband-­progression-­charts/
Bruce, C. (1997). The seven faces of information literacy. Auslib Press, Adelaide, South Australia.
Retrieved from https://eprints.qut.edu.au/139611/.
Burgsteiner, H., Kandlhofer, M., & Steinbauer, G. (2016, March). Irobot: Teaching the basics
of artificial intelligence in high schools. In Proceedings of the AAAI Conference on Artificial
Intelligence (Vol. 30, No. 1).
Celik, I. (2022). Towards intelligent-TPACK: An empirical study on teachers’ professional knowl-
edge to ethically integrate artificial intelligence (AI)-based tools into education. Computers in
Human Behavior, 138, 107468.
Chai, C. S., Lin, P. Y., Jong, M. S. Y., Dai, Y., Chiu, T. K., & Qin, J. (2021). Perceptions of and
behavioral intentions towards learning artificial intelligence in primary school students.
Educational Technology & Society, 24(3), 89–101.
Chan, K. S., & Zary, N. (2019). Applications and challenges of implementing artificial intelligence
in medical education: Integrative review. JMIR Medical Education, 5(1), e13930.
Chiu, T. K., Meng, H., Chai, C. S., King, I., Wong, S., & Yam, Y. (2021). Creation and evalua-
tion of a pretertiary artificial intelligence (AI) curriculum. IEEE Transactions on Education,
65(1), 30–39.
Chu, S. K. W., Reynolds, R. B., Tavares, N. J., Notari, M., & Lee, C. W. Y. (2021). 21st cen-
tury skills development through inquiry-based learning from theory to practice. Springer
International Publishing.
CSET. (2021). AI education in China and the United States. Retrieved from https://cset.george-
town.edu/wpcontent/uploads/CSET-AI-Education-in-China-and-the-United-States-1.pdf
Demeshkant, N., Potyrala, K., & Tomczyk, L. (2020). Levels of academic teachers digital com-
petence: Polish case-study. In 2020 Proceedings of the 28th International Conference on
Computers in Education (pp. 591–601).
Deng, W., Huang, X., Liu, Q., & Wang, Z. (2021). Curriculum design of artificial intelligence in
middle school – Taking posture recognition as an example. In 2021 tenth international confer-
ence of educational innovation through technology (pp. 310–315).
Dignum, V. (2019). Responsible artificial intelligence: How to develop and use AI in a responsible
way. Springer.
Druga, S., Vu, S. T., Likhith, E., & Qiu, T. (2019). Inclusive AI literacy for kids around the world.
In Proceedings of FabLearn 2019 (pp. 104–111).
Estevez, J., Garate, G., & Graña, M. (2019). Gentle introduction to artificial intelligence for high-
school students using scratch. IEEE Access, 7, 179027–179036.
Fernández-Martínez, C., Hernán-Losada, I., & Fernández, A. (2021). Early introduction of AI
in Spanish Middle Schools. A motivational study. KI-Künstliche Intelligenz, 35(2), 163–170.
References 57

Fleaca, E., & Stanciu, R. D. (2019). Digital-age learning and business engineering education–A
pilot study on students’ E-skills. Procedia Manufacturing, 32, 1051–1057.
Gamire, E., & Pearson, G. (Eds.). (2006). Tech tally: Approaches to assessing technological lit-
eracy. Island Press.
Goel, A. (2017). AI education for the world. AI Magazine, 38(2), 3–4.
Gong, X., Tang, Y., Liu, X., Jing, S., Cui, W., Liang, J., & Wang, F. Y. (2020, October). K-9 arti-
ficial intelligence education in Qingdao: Issues, challenges and suggestions. In 2020 IEEE
International Conference on Networking, Sensing and Control (ICNSC) (pp. 1–6). IEEE.
Goralski, M. A., & Tan, T. K. (2020). Artificial intelligence and sustainable development. The
International Journal of Management Education, 18(1), 100330.
Gresse von Wangenheim, C., Hauck, J. C., Pacheco, F. S., & Bertonceli Bueno, M. F. (2021).
Visual tools for teaching machine learning in K-12: A ten-year systematic mapping. Education
and Information Technologies, 26(5), 5733–5778.
Gunasilan, U. (2021). Debate as a learning activity for teaching programming: A case in the sub-
ject of machine learning. Higher Education, Skills and Work-Based Learning, 12(4), 705–718.
Gut, D. M. (2011). Integrating 21st century skills into the curriculum. In Bringing schools into the
21st century (pp. 137–157). Springer.
Hagendorff, T. (2020). The ethics of AI ethics: An evaluation of guidelines. Minds and Machines,
30(1), 99–120.
Han, E. R., Yeo, S., Kim, M. J., Lee, Y. H., Park, K. H., & Roh, H. (2019). Medical education trends
for future physicians in the era of advanced technology and artificial intelligence: An integra-
tive review. BMC Medical Education, 19(1), 1–15.
Heinze, C. A., Haase, J., & Higgins, H. (2010, July). An action research report from a multi-­
year approach to teaching artificial intelligence at the k-6 level. First AAAI Symposium on
Educational Advances in Artificial Intelligence, 24, 1890–1895.
How, M. L., & Hung, W. L. D. (2019). Educing AI-thinking in science, technology, engineering,
arts, and mathematics (STEAM) education. Education Sciences, 9(3), 184.
Huitt, W. (2011). Bloom et al.’s taxonomy of the cognitive domain. Educational Psychology
Interactive, 22, 1–4.
Huo, Y. (2019). Analysis of intelligent evaluation algorithm based on English diagnostic system.
Cluster Computing, 22(6), 13821–13826.
Janssen, C., Segers, E., McQueen, J. M., & Verhoeven, L. (2019). Comparing effects of instruction
on word meaning and word form on early literacy abilities in kindergarten. Early Education
and Development, 30(3), 375–399.
Julie, H., Alyson, H., & Anne-Sophie, C. (2020, October). Designing digital literacy activities: An
interdisciplinary and collaborative approach. In 2020 IEEE Frontiers in Education Conference
(FIE) (pp. 1–5). IEEE.
Kahn, K. M., Megasari, R., Piantari, E., & Junaeti, E. (2018). AI programming by children using
snap! Block programming in a developing country. Retrieved from https://ora.ox.ac.uk/objects/
uuid:9a82b522-6f9f-4c67-b20dbe6c53019b3b
Kandlhofer, M., Steinbauer, G., Hirschmugl-Gaisch, S., & Huber, P. (2016, October). Artificial
intelligence and computer science in education: From kindergarten to university. In 2016 IEEE
Frontiers in Education Conference (FIE) (pp. 1–9). IEEE.
Kaspersen, M. H., Bilstrup, K. E. K., & Petersen, M. G. (2021a, February). The machine learning
machine: A tangible user interface for teaching machine learning. In Proceedings of the fif-
teenth international conference on Tangible, Embedded, and Embodied Interaction (pp. 1–12).
Kaspersen, M. H., Bilstrup, K. E. K., Van Mechelen, M., Hjorth, A., Bouvin, N. O., & Petersen,
M. G. (2021b, June). VotestratesML: A high school learning tool for exploring machine learn-
ing and its societal implications. In FabLearn Europe/MakeEd 2021-An international confer-
ence on computing, design and making in education (pp.1–10).
Kim, S., Jang, Y., Kim, W., Choi, S., Jung, H., Kim, S., & Kim, H. (2021, May). Why and what
to teach: AI curriculum for elementary school. In Proceedings of the AAAI Conference on
Artificial Intelligence (Vol. 35, No. 17, pp. 15569–15576).
58 4 The Landscape of AI Literacy

Koehler, M. J., Mishra, P., & Cain, W. (2013). What is technological pedagogical content knowl-
edge (TPACK)? Journal of Education, 193(3), 13–19.
Kong, S. C., Cheung, W. M. Y., & Zhang, G. (2021). Evaluation of an artificial intelligence literacy
course for university students with diverse study backgrounds. Computers and Education:
Artificial Intelligence, 2, 100026.
Kokotsaki, D., Menzies, V., & Wiggins, A. (2016). Project-based learning: A review of the litera-
ture. Improving Schools, 19(3), 267–277.
Lao, N. (2020). Reorienting machine learning education towards tinkerers and ML-engaged citi-
zens. Massachusetts Institute of Technology.
Lee, I., Ali, S., Zhang, H., DiPaola, D., & Breazeal, C. (2021, March). Developing middle school
Students’ AI literacy. In Proceedings of the 52nd ACM Technical Symposium on Computer
Science Education (pp. 191–197).
Lin, P., & Van Brummelen, J. (2021, May). Engaging teachers to co-design integrated AI curricu-
lum for K-12 classrooms. In Proceedings of the 2021 CHI Conference on Human Factors in
Computing Systems (pp. 1–12).
Livingstone, S. (2004). Media literacy and the challenge of new information and communication
technologies. The Communication Review, 7(1), 3–14.
Long, D., & Magerko, B. (2020, April). What is AI literacy? Competencies and design consider-
ations. In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems
(pp. 1–16).
Lundberg, J., Arvola, M., Westin, C., Holmlid, S., Nordvall, M., & Josefsson, B. (2018). Cognitive
work analysis in the conceptual design of first-of-a-kind systems–designing urban air traffic
management. Behaviour & Information Technology, 37(9), 904–925.
McCarthy, J. (2007). From here to human-level AI. Artificial Intelligence, 171(18), 1174–1182.
Microsoft News Center India. (2022). Empowering India to be future ready. Retrieved from https://
news.microsoft.com/en-in/
MIT (2021). RAISE- Responsible AI for Social Empowerment and Education. Retrieved from
https://raise.mit.edu/
Morris, T. H. (2020). Experiential learning–a systematic review and revision of Kolb’s model.
Interactive Learning Environments, 28(8), 1064–1077.
Narahara, T., & Kobayashi, Y. (2018). Personalizing homemade bots with plug & play AI for
STEAM education. In SIGGRAPH Asia 2018 technical briefs (pp. 1–4).
National Research Council. (2012). Education for life and work: Developing transferable knowl-
edge and skills in the 21st century. National Academies Press.
Ng, D. T. K., & Chu, S. K. W. (2021). Motivating students to learn AI through social networking
sites: A case study in Hong Kong. Online Learning, 25(1), 195–208.
Ng, D.T.K., Lai, L.F.W., & Cheung T.W.T. (2021). Artificial Intelligence for All. eFunLearning
Ltd., Hong Kong Science and Technology Parks, Hong Kong.
Ng, D. T. K., Leung, J. K. L., Chu, K. W. S., & Qiao, M. S. (2021). AI literacy: Definition, teach-
ing, evaluation and ethical issues. Proceedings of the Association for Information Science and
Technology, 58(1), 504–509.
Ng, D. T. K., Luo, W., Chan, H. M. Y., & Chu, S. K. W. (2022). Using digital story writing as a
pedagogy to develop AI literacy among primary students. Computers and Education: Artificial
Intelligence, 3, 100054.
Noorbakhsh-Sabet, N., Zand, R., Zhang, Y., & Abedi, V. (2019). Artificial intelligence transforms
the future of health care. The American Journal of Medicine, 132(7), 795–801.
Reyes, M., Meier, R., Pereira, S., Silva, C. A., Dahlweid, F. M., Tengg-Kobligk, H. V., et al. (2020).
On the interpretability of artificial intelligence in radiology: Challenges and opportunities.
Radiology: Artificial Intelligence, 2(3), e190043.
Riina, V., Stefano, K., & Yves, P. (2022). DigComp 2.2: The digital competence framework for
citizens – With new examples of knowledge, skills and attitudes. Retrieved from https://publica-
tions.jrc.ec.europa.eu/repository/handle/JRC128415
References 59

Robinson, S., Orsingher, C., Alkire, L., De Keyser, A., Giebelhausen, M., Papamichail, K. N., et al.
(2020). Frontline encounters of the AI kind: An evolved service encounter framework. Journal
of Business Research, 116, 366–376.
Rodríguez-García, J. D., Moreno-León, J., Román-González, M., & Robles, G. (2020, October).
Introducing artificial intelligence fundamentals with LearningML: Artificial intelligence
made easy. In Eighth International Conference on Technological Ecosystems for Enhancing
Multiculturality (pp. 18–20).
Roll, I., & Wylie, R. (2016). Evolution and revolution in artificial intelligence in education.
International Journal of Artificial Intelligence in Education, 26(2), 582–599.
Sabuncuoglu, A. (2020, June). Designing one year curriculum to teach artificial intelligence for
middle school. In Proceedings of the 2020 ACM Conference on Innovation and Technology in
Computer Science Education (pp. 96–102).
Sakulkueakulsuk, B., Witoon, S., Ngarmkajornwiwat, P., Pataranutaporn, P., Surareungchai,
W., Pataranutaporn, P., & Subsoontorn, P. (2018, December). Kids making AI: Integrating
machine learning, gamification, and social context in STEM education. In 2018 IEEE
International Conference on Teaching, Assessment, and Learning for Engineering (TALE)
(pp. 1005–1010). IEEE.
Sanusi, I. T., Olaleye, S. A., Agbo, F. J., & Chiu, T. K. (2022). The role of learners’ competencies in
artificial intelligence education. Computers and Education: Artificial Intelligence, 3, 100098.
Sing, C. C., Teo, T., Huang, F., & Chiu, T. K. (2022). Secondary school students’ intentions to
learn AI: Testing moderation effects of readiness, social good and optimism. Educational
Technology Research and Development, 70(3), 765–782.
Steinbauer, G., Kandlhofer, M., Chklovski, T., Heintz, F., & Koenig, S. (2021). A differentiated
discussion about AI education K-12. KI-Künstliche Intelligenz, 35(2), 131–137.
Su, J., Zhong, Y., & Ng, D. T. K. (2022). A meta-review of literature on educational approaches
for teaching AI at the K-12 levels in the Asia-Pacific region (p. 100065). Artificial Intelligence.
Talja, S., Tuominen, K., & Savolainen, R. (2005). “Isms” in information science: Constructivism,
collectivism and constructionism. Journal of Documentation (Vol. 61, pp. 79–101).
Tamborg, A. L., Jankvist, U. T., & Misfeldt, M. (2022). Comparing programing and computa-
tional thinking with mathematical digital competencies form an implementation perspective.
In Proceedings of the 15th International Conference on Technology in Mathematics Teaching
(ICTMT 15) (p. 298). Danish School of Education, Aarhus University.
Tedre, M., Toivonen, T., Kahila, J., Vartiainen, H., Valtonen, T., Jormanainen, I., & Pears, A. (2021).
Teaching machine learning in K–12 classroom: Pedagogical and technological trajectories for
artificial intelligence education. IEEE Access, 9, 110558–110572.
Touretzky, D., Gardner-McCune, C., Martin, F., & Seehorn, D. (2019). Envisioning AI for K-12:
What should every child know about AI? Proceedings of the AAAI Conference on Artificial
Intelligence, 33(1), 9795–9799.
Touretzky, D., Gardner-McCune, C., & Seehorn, D. (2022). Machine Learning and the Five Big
Ideas in AI. International Journal of Artificial Intelligence in Education, 1–34.
UNESCO. (2018). A global framework of reference on digital literacy skills for indicator 4.4.2.
Retrieved from https://uis.unesco.org/sites/default/files/documents/ip51-global-framework-
reference-digital-literacy-skills-2018-en.pdf
UNESCO. (2022). K-12 AI curricula A mapping of government-endorsed AI curricula. Retrieved
from https://unesdoc.unesco.org/ark:/48223/pf0000380602.
Vachovsky, M. E., Wu, G., Chaturapruek, S., Russakovsky, O., Sommer, R., & Fei-Fei, L. (2016,
February). Toward more gender diversity in CS through an artificial intelligence summer pro-
gram for high school girls. In Proceedings of the 47th ACM technical symposium on Computing
Science Education (pp. 303–308).
Vale, L., & Fernandes, T. (2018). Social media and sports: Driving fan engagement with football
clubs on Facebook. Journal of Strategic Marketing, 26(1), 37–55.
60 4 The Landscape of AI Literacy

Van Laar, E., Van Deursen, A. J., Van Dijk, J. A., & De Haan, J. (2017). The relation between 21st-­
century skills and digital skills: A systematic literature review. Computers in Human Behavior,
72, 577–588.
Vazhayil, A., Shetty, R., Bhavani, R. R., & Akshay, N. (2019, December). Focusing on teacher
education to introduce AI in schools: Perspectives and illustrative findings. In 2019 IEEE Tenth
International Conference on Technology for Education (T4E) (pp. 71–77). IEEE.
Wan, X., Zhou, X., Ye, Z., Mortensen, C. K., & Bai, Z. (2020, June). SmileyCluster: Supporting
accessible machine learning in K-12 scientific discovery. In Proceedings of the Interaction
Design and Children Conference (pp. 23–35).
Wang, T., & Cheng, E. C. K. (2021). An investigation of barriers to Hong Kong K-12 schools incor-
porating artificial intelligence in education. Computers and Education: Artificial Intelligence,
2, 100031.
Williams, R., Park, H. W., Oh, L., & Breazeal, C. (2019). Popbots: Designing an artificial intel-
ligence curriculum for early childhood education. Proceedings of the AAAI Conference on
Artificial Intelligence, 33(01), 9729–9736.
Wong, G. K., Ma, X., Dillenbourg, P., & Huan, J. (2020). Broadening artificial intelligence educa-
tion in K-12: Where to start? ACM Inroads, 11(1), 20–29.
Xia, Q., Chiu, T. K., Lee, M., Sanusi, I. T., Dai, Y., & Chai, C. S. (2022). A self-determination
theory (SDT) design approach for inclusive and diverse artificial intelligence (AI) education.
Computers & Education, 189, 104582.
Yau, K. W., Chai, C. S., Chiu, T. K., Meng, H., King, I., & Yam, Y. (2022). A phenomenographic
approach on teacher conceptions of teaching artificial intelligence (AI) in K-12 schools.
Education and Information Technologies, 1–24.
Yoder, S., Tatar, C., Aderemi, I., Boorugu, S., Jiang, S., & Akram, B. (2020). Gaining insight into
effective teaching of AI problem-solving through CSEDM: A case study. In 5th Workshop on
Computer Science Educational Data Mining.
Zhang, Z. (2018, February). Develop the AI literacy of infants by deeply integrated English learn-
ing and robot education. In 2018 International Conference on Computer Science, Electronics
and Communication Engineering (CSECE 2018) (pp. 154–156). Atlantis Press.
Zhang, J., & Du, H. (2008, December). The PBL’s application research on Prolog Language’s
Instruction. In 2008 international workshop on education technology and training & 2008
international workshop on geoscience and remote sensing (Vol. 1, pp. 112–114). IEEE.
Zimmermann-Niefield, A., Turner, M., Murphy, B., Kane, S. K., & Shapiro, R. B. (2019, June).
Youth learning machine learning through building models of athletic moves. In Proceedings
of the 18th ACM international conference on interaction design and children (pp. 121–132).
 Part II
K-16 AI Literacy Education
Chapter 5
AI Literacy Education in Early Childhood
Education

Part I of this book gave us basic ideas about what AI literacy is, why it is important
for all K–16 learners, as well as the theoretical frameworks and important theories
involved in AI literacy education. However, students at each educational level have
varied needs and intended learning outcomes. Therefore, the way of how AI literacy
education is being implemented across levels of education should not be the same.
Identifying specific age-appropriate approaches on different educational levels
could inform schools, policymakers, educators, and parents on the design and devel-
opment of adequate environment, pedagogy, learning content, technology, and
assessment tools that best meet the needs of their students.
Part II of this book further gives us an outline of AI literacy education across
educational levels. Several models of AI literacy education, in particular Bloom’s
taxonomy and the TPACK model, comprise key digital competencies to inform
K–16 educators what knowledge, skills, and attitudes students should equip with.
The following three chapters will first explain the research method including litera-
ture search and data analysis. After that, we conduct a systematic scoping review on
four research questions and discuss the findings based on the TPACK model for
each educational level:
• What pedagogical strategies were commonly used in AI literacy studies?
• What learning content is appropriate for students in AI literacy studies?
• What learning tools have been used in AI literacy studies?
• What assessment methods have been used in AI literacy studies?

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 63

D. T. K. Ng et al., AI Literacy in K-16 Classrooms,
https://doi.org/10.1007/978-3-031-18880-0_5
64 5 AI Literacy Education in Early Childhood Education

5.1 Introduction

Most AI literacy research was conducted in primary, secondary, and higher educa-
tion settings (Eguchi et al., 2021; Su et al., 2022). However, much less attention has
been paid to how AI education for young children aged 3–8 years relative to other
age groups. Early childhood education refers to the education and care of children
from birth up to 8 years of age. Although previous studies have brought AI learning
tools into early childhood education classrooms and shown their promising effects
(e.g., Williams et al., 2019a; Lin et al., 2020; Tseng et al., 2021), very little has been
known about the AI literacy for early childhood education.
Artificial intelligence literacy education can be started as young as kindergarten-
ers. As suggested in Chap. 3, not only AI literacy helps young children to develop
adequate technological literacy that positively influence their living and future
study; it effectively develops many other skills for their cognitive development,
such as computational thinking, theory of mind skills, inquiry skills, emotional lit-
eracy, and collaboration (Su & Yang, 2022; Williams et al., 2019a, b; Kewalramani
et al., 2021). For example, Kewalramani et al. (2021) designed a set of AI learning
activities for 4 to 5-year-old children to play and interact with AI toys to enhance
their three types of literacies including creating inquiry literacy, emotional inquiry
literacy, and collaborative inquiry literacy. Another study conducted by Druga et al.
(2019) designed an AI-interfaced robot for children in order to teach them about the
capabilities of AI agents. Therefore, it is meaningful to systematically analyze and
discuss existing work focusing on AI literacy in early childhood education (ECE)
development. Four research questions (RQ) formed the basis of this section:
RQ1: What are the pedagogical strategies used in AI literacy studies at the ECE level?
RQ2: What learning content is appropriate in AI literacy studies at the ECE level?
RQ3: What learning tools have been used in AI literacy studies at the ECE level?
RQ4: What assessment methods have been used in AI literacy studies at the
ECE level?

5.2 Methods

This chapter followed the procedures for the Preferred Reporting Items for
Systematic Reviews and Meta-Analyses the PRISMA statement (Moher et al., 2009).
In order to facilitate database search, this study surveyed peer-reviewed academic
articles published in all years. All articles were accessed in June 2022. Databases
searched were Web of Science, IEEE, and Scopus. We formulated a search string
based on our understanding of and knowledge in the AI education domain and also
by referring to related AI education search strings used in other studies such as Su
and Yang (2022). The search string used for this study was “artificial intelligence”
OR “AI” OR “machine learning” OR “ML” AND “early childhood” OR “young
child*” OR “preschool*” OR “kindergarten*” OR “pre-k*” OR “childcare” OR
5.2 Methods 65

“child care” OR “day-care” OR “children” AND “curricula*” OR “learning” OR

“curriculum design.” As a result, a total of 3474 articles were retrieved, among
which 142 were from Web of Science, 3042 from Scopus, and 299 from IEEE.
First, by reading through their title and abstract, 3389 results were excluded due
to their irrelevance to the research topic. Second, 18 results were excluded that
reported duplicate studies. Then, by closely examining their full text, we excluded
the articles (a) in which participants were not 3–8 years old (15), (b) that focus was
not AI or machine learning (ML) (11), and (c) that did not discuss curriculum design
or learning programs/activities (26). Eligible studies included in this review were
restricted to articles written in the English language. After considering the articles
based on the criteria, 15 studies remained. Figure 5.1 below provides an overview
of our search protocol.

Fig. 5.1 PRISMA diagram of included articles

66 5 AI Literacy Education in Early Childhood Education

5.3 Results and Discussion

Fifteen articles that focused on AI literacy in early childhood education were thor-
oughly examined in this review (Table 5.1). Different types of literature are included
in this review, such as research articles and conference papers. Fifteen articles that
focused on AI literacy in early childhood education were thoroughly reviewed
(2016, 1 article; 2018, 1 article; 2019, 3 articles; 2020, 2 articles; 2021, 4 articles,
and 2022, 4 articles). To facilitate the database search, this study identifies all peer-­
reviewed academic articles published from 2018 to 2021 since the first article was
found in 2016 (Kandlhofer et al., 2016). This review shows that all studies were
conducted in developed countries/regions (e.g., the United States, Austria, Germany,
Denmark, Sweden, Japan, Hong Kong, and Australia).
RQ1: What Were the Pedagogical Strategies Commonly Used in AI Early
Childhood Education Studies?
In terms of the pedagogical strategies design used, most studies were very success-
ful. To begin with, the pedagogical strategies include learning activities, learning
methods, and AI learning tools. Most researchers in the AI in early childhood edu-
cation field have developed learning programs designed to improve children’s
AI-related knowledge, such as AI concepts, knowledge-based systems, supervised
machine learning, generative AI, machine learning and data science, and AI and
ethics (e.g., Kandlhofer et al., 2016; Williams et al., 2019a, b; Lin et al., 2020;
Tseng et al., 2021). Second, one study used methods of discovery- and inquiry-­
based learning, storytelling, and educational robotics to teach different AI and com-
puter science topics to children (Kandlhofer et al., 2016). Third, three popular AI
tools, such as PopBots (Williams, 2018; Williams et al., 2019a, b), Zhorai (Lin
et al., 2020), and PlushPal (Tseng et al., 2021) designed an AI and machine learning
curriculum for children, helping children in better understanding the concept or
knowledge of AI or machine learning (Lin et al., 2020; Williams et al., 2019a, b;
Tseng et al., 2021). Lastly, six children explored machine learning based technol-
ogy in nonschool settings (Vartiainen et al., 2020).
RQ2: What Learning Contents Are Appropriate for Students at the
ECE Level?
Different AI learning contents are included in ECE level, such as rule-based sys-
tems, supervised learning, generative AI, and machine learning (e.g., Lin et al.,
2020; Su & Zhong, 2022; Williams, 2018). For example, Su and Zhong (2022)
designed an AI curriculum framework in ECE settings; the learning contents include
definition of AI and examples of AI, the five big ideas in AI, machine learning,
applications, and AI ethics (Table 5.2). This study also designed some learning
activities to enhance children’s AI tools skill (Su & Zhong, 2022). For example, Su
and Zhong (2022) designed Google’s Quick, Draw! activities to enhance children’s
use of AI tools skill.
5.3 Results and Discussion 67

Table 5.1 Descriptive information of the included studies

Research Pedagogical Learning Learning Assessment
Author/year methods strategies content tools methods
Lin et al. Quantitative / What does Zhorai Knowledge
(2020) Zhorai know assessments
Teaching (pre- and
Zhorai post-­
Witnessing assessments)
machine
learning
AI and ethics
Kandlhofer Qualitative Discovery- and AI/computer / Video data,
et al. (2016) inquiry-based science topic pictures and
learning, storytelling, observations
and educational (field notes)
robotics
Williams Quantitative Creative design Rule-based PopBots Theory of
(2018) activities: draw the systems mind
AI robots Supervised assessment,
Real-world AI machine rock paper
activities: Google learning, and scissors
quick draw generative AI assessment
performance,
supervised
learning
assessment
performance,
generative
assessment
performance,
pre- and
posttests of
children’s
perception,
attitudes
assessment
Williams Quantitative AI activities Knowledge-­ PopBots PopBots
et al. (2019a) Introduction to based assessment,
programming with systems, theory of
the supervised mind
PopBotsKnowledge-­ machine assessment,
based systems with learning, and and
rock-paper-scissors; generative AI perception of
supervised machine robots
learning with food questionnaire
classification;
generative AI with
music remix
(continued)
68 5 AI Literacy Education in Early Childhood Education

Table 5.1 (continued)

Research Pedagogical Learning Learning Assessment
Author/year methods strategies content tools methods
Williams Quantitative Knowledge-based Knowledge-­ PopBots Pre- and
et al. (2019b) systems with based post-­
rock-paper-scissors systems, assessments
Supervised machine supervised (knowledge)
learning with food; machine
generative AI with learning, and
music remix generative
music AI
Druga et al. Quantitative Interact with AI AI perception Jibo robot, AI perception
(2019) agents; AI perception and Anki’s questionnaire
monster game expectations Cozmo
robot and
Amazon’s
Alexa
Dwivedi Mixed The design activity: Machine Teachable Machine
(2021) Teachable Machine learning Machine learning
concepts metric and
training data
Druga and Mixed Learning activities: Cognimates Cognimates Pre-/
Ko (2021) “Make me happy AI platform AI platform post-­
program”, “rock (train, code perception
paper scissors and test a game
program”, “smart series of responses and
home program” smart observations
programs)
Tseng et al. Quantitative Created a project The PlushPal Surveys
(2021) using three common fundamentals
gestures, such as of machine
none, jump, and learning and
running data science
Kewalramani Qualitative Playing with / Cozmo Interviews
et al. (2021) battery-operated Blue Bot and
robots (Botley and Coji by observations
Beebots) and using Wowee,
block play to build Qobo the
ramps and roadways snail, and
for their robots to vernie-
travel Lego boost
bot
(continued)
5.3 Results and Discussion 69

Table 5.1 (continued)

Research Pedagogical Learning Learning Assessment
Author/year methods strategies content tools methods
Druga et al. Qualitative Activity: Image Teachable Video
(2022a) Classification game; classification; Machine recording:
anchor game; Object transcribed
reflection; object recognition the videos and
recognition; train AI; Voice noted
prediction game; assistants comments
compare with voice Unplugged
assistant; draw what AI games and
is inside; AI bingo co-design
game; analyze AI; Reflection on
design AI study and
learning
activities
Yang (2022) / AI + ocean Machine Popbo /
protection learning Clearbot
Vartiainen Qualitative / Explore the Google’s Video data
et al. (2020) input-output Teachable
relationships Machine
with GTM
Su and / Problem-based Module: PictoBlox /
Zhong learning: AI farming Introduction
(2022) to AI
Module 2:
Machine
learning
Module 3:
Speech
recognition
Module 4:
Flaws and
biases of AI
Tazume et al. / Play with AI robots / RoBoHoN /
(2021)

Table 5.2 AI curriculum for ECE

AI knowledge AI skills AI attitudes
K1: Definition of AI and S1: Using AI tools A1: Social impact
examples of AI
K2: The five big ideas in AI S2: Computational thinking and A2: Collaborate with
programming AI
K3: Machine learning S3: Critical thinking
K4: Applications S4: Problem-solving
K5: AI ethics
Adopted from Su and Zhong (2022)
70 5 AI Literacy Education in Early Childhood Education

RQ3: What Were the Learning Tools Used in AI Early Childhood Education?
As shown in Table 5.3, three studies have used PopBots and Teachable Machines as
the learning tools for supporting children’s engagement in AI. Other AI learning
tools reported in the studies include Zhorai, Jibo robot, Anki’s Cozmo robot,
Amazon’s Alexa, Cognimates AI platform, PlushPal, Popbo, Clearbot, RoBoHoN,
and PictoBlox. These learning tools seem to enhance children’s learning AI con-
cepts. For example, children understand three AI concepts (knowledge-based sys-
tems, supervised machine learning, and generative music) using AI learning tools
(i.e., PopBots) (Williams, 2018; Williams et al., 2019a, b).
RQ4: What Assessment Methods Have Been Used in Researching AI in Early
Childhood Education?
Most studies used a quantitative design and mixed-method design, followed by the
qualitative design. Of note is that three articles (Su & Zhong, 2022; Yang, 2022;
Tazume et al., 2021) only introduced and described their AI curricula in early child-
hood education, without implementing them in practice, and thus no data was col-
lected. More details are shown in Fig. 5.2.
In terms of assessment methods, knowledge assessments/tests (4) are the most
commonly used, followed by robot-based assessment and video analysis (3), as
shown in Table 5.4. Four studies used knowledge assessments to assess children’s
AI understandings (Lin et al., 2020; Williams, 2018; Williams et al., 2019a, b) (see
Table 5.1). For example, some scholars used knowledge tests to examine children’s
machine learning knowledge using Zhorai (Lin et al., 2020). According to the
results, Zhorai can help children understand machine learning concepts more easily.

Table 5.3 Learning tools

Learning tools Website

PopBots https://www.media.mit.edu/projects/pop-­kit/
overview/
Teachable Machine https://teachablemachine.withgoogle.com/
Zhorai http://zhorai.csail.mit.edu/
Jobo robot https://jibo.com/
Anki’s Cozmo robot https://www.digitaldreamlabs.com/pages/
cozmo
Amazon’s Alexa https://developer.amazon.com/en-­US/alexa
Cognimates AI platform http://cognimates.me/home/
PlushPal https://www.plushpal.app/
RoBoHoN https://robohon.com/global/
PictoBlox https://pictoblox.ai/
5.4 Conclusion 71

Fig. 5.2 Research Research Methods

methods

2 Quantave

Qualitave
6
Mixed
4

5.4 Conclusion

This chapter provides an overview of AI literary studies in ECE, focusing on

research methods, pedagogical strategies, learning content, AI tools, and assess-
ment methods. Although there were few studies on AI literacy for early childhood
education, the existing references did provide new insights into various aspects of
AI literacy for children. This chapter would provide valuable directions for AI edu-
cation in early childhood education and serve as a reference for future AI literacy
research in the digital society.
Researchers should design more interesting AI activities for children, such as
interacting or playing with AI robots, which could change their existing attitudes
toward AI. For example, Tazume et al. (2020) designed play with AI robots (i.e.,
RoBoHoN) activities for children. Results show that “children were strongly moti-
vated to engage with human-type AI media” (Tazume et al., 2020, p. 328). Moreover,
future researchers will develop quantitative and qualitative methods to evaluate chil-
dren’s learning outcome (i.e., knowledge, attitudes, and motivation) through pre and
post-knowledge tests, surveys, questionnaires, and observations. Furthermore,
future studies should explore what types of learning contents and teaching methods
are more suitable for early childhood education and what kind of teacher training is
necessary for kindergarteners. Lastly, most studies were conducted in developed
countries/regions, such as the United States, Austria, Germany, Denmark, Sweden,
Japan, Hong Kong, and Australia. Recommend that future research needs to inves-
tigate how AI literacy in early childhood can be applied in developing countries.
72 5 AI Literacy Education in Early Childhood Education

Table 5.4 Assessment methods

Studies Details of assessments
Knowledge assessments/tests
Williams Rule-based systems assessment:
(2018); 1. Control: Which of these is rock? Rock, paper, or scissors?
Williams et al. 2. We teach the robot the normal rules. Then, Sally plays rock and the robot
(2019a, b) plays paper, who does the robot think has won? Sally or the robot?
3. Sally plays paper five times.
What does the robot think she will play next? Rock, paper, or scissors?
4. The robot thinks that Sally will play paper next. What will the robot play so
that it can beat Sally? Rock, paper, or scissors?
5. We changed the rules so that they are all opposite rules (paper beats
scissors). Sally plays scissors and the robot plays paper.
Who does the robot think has won? Sally or the robot?
Supervised machine learning:
1. Control: Which one of these foods is bad for your teeth? Strawberry, ice
cream, or corn?
2. You start the robot and put strawberries and tomatoes into the good group.
Which group will the robot think chocolate goes in? The good group or the bad
group?
3. What food does the robot think is most like a tomato? Strawberry, banana, or
milk?
4. You put ice cream in the good category and bananas in the bad category.
What category will the robot put corn in? The good category or the bad
category?
Generative AI assessment
1. Control: Which one of these notes will make the robot’s eyes go orange?
Purple note, orange note, or green note?
2. Priya asks the robot to play back with the bars in the middle. Does the robot
play the same song or a different song?
3. Priya asks the robot to play back with the bars to the right. Does the robot
play the same song or a different song?
4. Does the robot’s song have to have the same notes as the input?
Lin et al. Zhorai assessment (machine learning)
(2020) 1. Which sentences could you say to Zhorai to create the following mind map?
(this assesses their understanding of how knowledge is represented.)
2. What could you tell Zhorai about monkeys so that it could correctly guess
that monkeys live in rainforests? (this assesses their understanding of how
Zhorai learns.)
3. The following histogram is what Zhorai thinks about where “toucan” lives.
Based on the histogram, which ecosystem would Zhorai think a toucan lives
in? (This assesses their understanding of how Zhorai makes a decision.)
4. Which ecosystems do snakes live in? Why might Zhorai have a difficult time
classifying snakes into one ecosystem even if it knew everything there is to
know about them? (This is an open-ended question for assessing mistakes
Zhorai may make.)
5. Have you tried saying “Zhorai” to Zhorai? If not, ask the teacher if you can
try. Does Zhorai recognize its own name? If not, why do you think it doesn’t?
Can you think of another name that Zhorai won’t recognize? (This is an
open-ended question on Zhorai’s internal natural language processing.)
(continued)
References 73

Table 5.4 (continued)

Studies Details of assessments
Robot-based assessment
Williams Which view do you agree with more, or are you somewhere in the middle?
(2018); Robots follow rules/robots do not follow rules. I am smarter than robots/robots
Williams et al. are smarter than me. Robots are like toys/robots are like people. Robots cannot
(2019a) learn new things/robots can learn new things. Robots are like friends/robots are
like adults. Druga et al. (2019) intelligence attribution, truthfulness attribution,
perceived understanding
Druga et al. Intelligence attribution, truthfulness attribution, perceived understanding
(2019)
Video analysis
Kandlhofer To assess students’ joyfulness to explore the different units and understanding
et al. (2016) about the AI fundamental concepts (e.g., carrying out correct actions in the
activities)
Druga et al. Children’s body language and non-verbal interactions
(2022a, b)
Vartiainen Children’s interaction with a Teachable Machine
et al. (2020)

References

Druga, S., & Ko, A. J. (2021, June). How do children’s perceptions of machine intelligence change
when training and coding smart programs?. In Interaction design and children (pp. 49–61).
Druga, S., Vu, S. T., Likhith, E., & Qiu, T. (2019). Inclusive AI literacy for kids around the world.
Proceedings of FabLearn, 2019, 104–111.
Druga, S., Christoph, F. L., & Ko, A. J. (2022a, April). Family as a third space for AI literacies:
How do children and parents learn about AI together?. In CHI Conference on Human Factors
in Computing Systems (pp. 1–17).
Druga, S., Christoph, F. L., & Ko, A. J. (2022b, April). Family as a third space for AI literacies:
How do children and parents learn about AI together?. In CHI Conference on Human Factors
in Computing Systems. ACM publications. (pp. 1–17).
Dwivedi, U. (2021, June). Introducing children to machine learning through machine teaching. In
Interaction design and children. ACM publications. (pp. 641–643).
Eguchi, A., Okada, H., & Muto, Y. (2021). Contextualizing AI education for K-12 students to
enhance their learning of AI literacy through culturally responsive approaches. KI-Künstliche
Intelligenz, 35(2), 153–161.
Kandlhofer, M., Steinbauer, G., Hirschmugl-Gaisch, S., & Huber, P. (2016, October). Artificial
intelligence and computer science in education: From kindergarten to university. In 2016 IEEE
Frontiers in Education Conference (FIE) (pp. 1–9). IEEE.
Kewalramani, S., Kidman, G., & Palaiologou, I. (2021). Using artificial intelligence (AI)-interfaced
robotic toys in early childhood settings: A case for children’s inquiry literacy. European Early
Childhood Education Research Journal, 29(5), 652–668.
Lin, P., Van Brummelen, J., Lukin, G., Williams, R., & Breazeal, C. (2020, April). Zhorai: Designing
a conversational agent for children to explore machine learning concepts. In Proceedings of the
AAAI Conference on Artificial Intelligence, 34(09), 13381–13388.
Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., Altman, D., Antes, G., et al. (2009). Preferred
reporting items for systematic reviews and meta-analyses: The PRISMA statement (Chinese
edition). Journal of Chinese Integrative Medicine, 7(9), 889–896.
Su, J., & Yang, W. (2022). Artificial intelligence in early childhood education: A scoping review.
Computers and Education: Artificial Intelligence, 3, 100049.
74 5 AI Literacy Education in Early Childhood Education

Su, J., & Zhong, Y. (2022). Artificial intelligence (AI) in early childhood education: Curriculum
design and future directions. Computers and Education: Artificial Intelligence, 3, 100072.
Su, J., Zhong, Y., & Ng, D. T. K. (2022). A meta-review of literature on educational approaches for
teaching AI at the K-12 levels in the Asia-Pacific region. Computers and Education: Artificial
Intelligence, 100065.
Tazume, H., Morita, T., & Hotta, H. (2020, June). Young children’s literacy and cognition to
interactive AI robots: A multifaceted study of potential enhancement to early childhood edu-
cation. In EdMedia+ innovate learning (pp. 323–328). Association for the Advancement of
Computing in Education (AACE).
Tseng, T., Murai, Y., Freed, N., Gelosi, D., Ta, T. D., & Kawahara, Y. (2021, June). PlushPal:
Storytelling with interactive plush toys and machine learning. In Interaction design and chil-
dren (pp. 236–245).
Vartiainen, H., Tedre, M., & Valtonen, T. (2020). Learning machine learning with very young
children: Who is teaching whom? International Journal of Child-Computer Interaction, 25,
100182. https://doi.org/10.1016/j.ijcci.2020.100182
Williams, R. (2018). PopBots: Leveraging social robots to aid preschool children’s artificial intel-
ligence education (Doctoral dissertation, Massachusetts Institute of Technology).
Williams, R., Park, H. W., & Breazeal, C. (2019a, May). A is for artificial intelligence: The impact
of artificial intelligence activities on young children’s perceptions of robots. In Proceedings of
the 2019 CHI Conference on Human Factors in Computing Systems (pp. 1–11).
Williams, R., Park, H. W., Oh, L., & Breazeal, C. (2019b, July). Popbots: Designing an artificial
intelligence curriculum for early childhood education. In Proceedings of the AAAI Conference
on Artificial Intelligence, 33(01), 9729–9736.
Yang, W. (2022). Artificial intelligence education for young children: Why, what, and how in
curriculum design and implementation. Computers and Education: Artificial Intelligence,
3, 100061.
Chapter 6
AI Literacy Education in Primary Schools

Based on the literature review in Chap. 5, we learn that even children as young as 4
years old have already grown up with AI. In our rapidly transforming digital world,
equipping young learners with AI knowledge and skills will help ensure their
employability and learning potential in their future. Moreover, AI is already present
in their everyday life such as video games, AI toys, virtual assistants, and smart
devices (e.g., Google Assistant, robotics dogs, Alexa devices). Teaching AI was not
possible in the past; however, with age-appropriate curriculum and tools, primary
students can now know and understand the working principles behind AI, use AI for
learning purposes, and apply their knowledge to create artifacts to solve authentic
problems. As such, there is a need to investigate the pedagogy, learning content,
tools, and assessment methods involved to develop young learners’ AI literacy.
Primary schools have started AI literacy education to foster students’ AI funda-
mentals, applications, and limitations. Nonetheless, the way how educators teach
about AI remains largely unanswered due to the scarcity of research on what pri-
mary school students can learn about AI and what pedagogical strategies or learning
tools are appropriate for teaching primary school students AI (Chai et al., 2021).
Given the challenges in implementing AI education as an immediate result of a
limited amount of literature in this field, there is an urgent need to address AI educa-
tion in primary school settings. This chapter aims to fill the research gaps to evalu-
ate, synthesize, and present literature under the current primary school education
landscape. Four research questions (RQ) were formulated:
RQ1: What are the pedagogical strategies used in AI literacy studies at the pri-
mary level?
RQ2: What learning content is appropriate in AI literacy studies at the primary level?
RQ3: What learning tools have been used in AI literacy studies at the primary level?
RQ4: What assessment methods have been used in AI literacy studies at the pri-
mary level?

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 75

D. T. K. Ng et al., AI Literacy in K-16 Classrooms,
https://doi.org/10.1007/978-3-031-18880-0_6
76 6 AI Literacy Education in Primary Schools

6.1 Method

In this chapter, electronic databases, namely, ACM Digital library, Google Scholar,
IEEE, ProQuest Education Collection, Scopus, and Web of Science, were utilized
for the literature search because of their reliability, as they all include a very high
number of peer-reviewed journals and conferences. This paper focuses on AI educa-
tion in primary school settings and uses the right term as a crucial part of the search-
ing process (Talbott et al., 2018). In this study, the keywords selected were (“AI”
OR “artificial intelligence” OR “artificial intelligence literacy” OR “deep learning”
OR “machine learning” OR “neural network*” OR “natural language processing”
OR “chatbot”) AND (“primary school” OR “elementary school” OR “primary edu-
cation” OR “elementary education” OR “primary school student” OR “elementary
school student” OR “primary school pupil” OR “elementary school pupil”) AND
(“learning” OR “teaching” OR “pedagogy” OR “curriculum” OR “assessment”
OR” challenges”). “Artificial intelligence” as a word or in combination with others
such as “literacy” was also included in the search string. “Deep learning” OR
“machine learning” OR “neural network” OR “natural language processing” OR
“chatbot” were also used as part of the context for the search, because they consti-
tute a subset of artificial intelligence. “Primary” or “elementary” is the context on
which this paper focuses. “Learning” OR “teaching” OR “pedagogy” OR “curricu-
lum” OR “assessment” as the context to which AI education is also applied.
To narrow down the scope of the article search, a protocol was employed in
advance to document the exclusion and inclusion criteria. The following articles
were removed according to the following exclusion criteria: (1) editorials, maga-
zines, books, book chapters, and dissertations were excluded as they are not sub-
jected to scholarly peer reviews and are not related to the research topic or abstract
(n = 36,320); (2) duplicated articles (n = 209); (3) articles that solely focused on
artificial intelligence technologies and were not related to AI literacy (n = 978); (4)
articles whose focus was not related to AI learning, teaching, pedagogy, curriculum,
or assessment (n = 34); and (5) student participants were not within the age range of
6–14 years old (n-5). On the other hand, the inclusion criteria were the following:
(1) the work was written in English; (2) the paper was published in a peer-reviewed
journal or conference; (3) student participants were within 6–14 years old or cur-
rently studying in primary schools; (4) teachers, schools, or any stakeholders
involved in teaching or learning of AI; and (5) the work answered one or more of the
terms related to the topics of the RQs. After applying all these criteria, there were
37 articles to be analyzed. The scoping review in Fig. 6.1 employs the Preferred
Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow dia-
gram (Moher et al., 2009). The article selection procedure included steps such as
identification, screening, eligibility, data extractions, and data analysis.
To verify coding reliability, 12 articles were randomly selected, blinded/coded,
and reviewed by two researchers. Any disagreements and differences in the compre-
hension of the coding system were resolved to ensure inter-rater reliability. The data
were then examined and summarized using a coding scheme.
6.2 Results and Discussion 77

37,583 records identified through database

Identification

searching: ACM (1,203), Google (1,630),

IEEE (2,214), ProQuest(11,415)
Web of Science (403).

36,320 records excluded by title, abstract,

non-peer reviewed, not written in English
and not answer RQs
Screening

1263 records screened

1187 articles excluded:

-209 duplicated copies
-978 solely on AI educational
Technologies (AIED), not related
to AI literacy education
Eligibility

76 full-text articles assessed for eligibility

39 articles excluded:
5 – participants not within 6-14 years old
or in primary school settings
34 – Unrelated to AI learning, teaching,
Included

pedagogy or curriculum or assessment.

37 studies included for review

Fig. 6.1 PRISMA diagram

6.2 Results and Discussion

RQ1: What are the pedagogical strategies used in AI literacy studies?

The most often employed pedagogical strategies among the selected studies are
project-based learning (11), play/game-based learning (11), and collaborative learn-
ing (8) (Table 6.1). First, project-based learning enables students to actively acquire
a deeper knowledge through active exploration of real-world challenges and prob-
lems such as through teachable machine learning projects to understand the three
steps of labeling, training and evaluation classifier (Melsion et al., 2021), and an
assemble and programmable toy car to stimulate students’ curiosity and motivation
to learn about AI (Narahara & Kobayashi, 2018). Various levels of scratch projects
78 6 AI Literacy Education in Primary Schools

Table 6.1 Pedagogical approaches for AI literacy education in a primary school setting
Pedagogies Descriptions Sample studies No
Project-based Enables students to actively acquire a Ali et al. (2019), Han et al. 11
learning deeper knowledge through active (2018), Heinze et al. (2010),
exploration of real-world challenges and Ho et al. (2019), Li and Song
problems (2019), Narahara and
Kobayashi (2018), Melsion
et al. (2021), Ng et al. (2022),
Rodríguez-García (2020,
2021), Vartiainen et al. (2020)
Game-based/ Refers to the borrowing of certain game Ali et al. (2019), Han et al. 11
play-based principles or using play as a context for (2018), Heinze et al. (2010),
learning learning Henry et al. (2021), Ho et al.
(2019), Lee et al. (2020),
Narahara and Kobayashi
(2018), Ng et al. (2022),
Shamir and Levin (2021,
2022), Voulgari et al. (2021)
Collaborative Through peer instruction or human-­ Ali et al. (2019), Eguchi et al. 8
learning/ machine interaction, students teach each (2021), Heinze et al. (2010),
human-computer other, address misunderstandings, Ho et al. (2019), Lee et al.
collaborative clarify misconceptions, and discuss (2020), Li and Song (2019),
learning concepts to solve a problem, complete a Tkáčová et al. (2020),
task or create a product. Toivonen et al. (2020),
Vartiainen et al. (2020)

are also used to teach AI according to the different age characteristics of students
(Li & Song, 2019).
The second commonly mentioned method is game-based learning or playful
learning approaches that refer to the borrowing of certain game principles or using
play as a context for learning. This pedagogy is supported by various recent research.
Henry et al. (2021) designed machine learning concepts in gameplay, and Voulgari
et al. (2021) used ArtBot games to scaffold and introduce supervised learning, rein-
forcement learning, and AI algorithmic bias to students. In addition, Lee et al.
(2020) created a collaborative game-based environment with a learning tool,
PRIMARYAI, to enable upper primary school students to gain experience in image
recognition, machine learning, planning, and automated decision-making.
Collaborative learning is also adopted to maximize students’ learning outcomes.
By allowing students to co-design (Toivonen et al., 2020), co-teach, and peer-teach
the machine learning application (Vartiainen et al., 2020), these studies emphasize
that learning is a collaborative process in which students effectively explore, engage
in play, and apply knowledge (Ackermann, 2001).
When teaching AI to primary school students, several researchers use more than
one pedagogical strategy. For example, building on the constructionist pedagogy of
adopting a project-based approach, Ali et al. (2019) developed a hands-on collab-
orative Droodle Creativity game in his AI curriculum which involves designing new
interfaces for students to explore AI. His study examines the project where students
6.2 Results and Discussion 79

can think creatively and learn about AI by modeling the creative behavior of robots.
The diversity of students between and within schools could be one explanation for
the phenomena of this multiple use of pedagogical strategies (Chiu et al., 2021).
Furthermore, other factors such as the primary school students’ age, gender, back-
ground knowledge, educational surroundings, and available learning tools may have
an impact on their learning styles and motivation to learn.
RQ2: What learning content is appropriate for primary school students in AI lit-
eracy studies?
Adapted from Bloom’s taxonomy’s cognitive levels (Bloom, 1956) and Ng et al.
(2021)’s conceptual definition of AI learning inquiry, the learning content of AI
education activities can be summarized in four categories which are (1) know and
understand, (2) use and apply, (3) create and evaluate, and (4) AI ethics.
Know and Understand The first cognitive level focuses on the AI foundations
which are the basic AI foundations for every student who does not have prior AI or
background and computer science knowledge. Researchers often design AI educa-
tional activities to engage students in acquiring authentic AI concepts, knowledge,
and skills. Heinze et al. (2010) tailored AI activities in an age-appropriate and play-
ful manner in order to effectively engage students to know and understand basic
concepts, vocabulary, the history of science, and building blocks of AI. Furthermore,
robotic exercises (Ho et al., 2019), frequent interactions with AI machines
(Vartiainen et al., 2020), and greater AI exposure such as driving an AI toy car
(Narahara & Kobayashi, 2018) demystify the AI concepts such as facial recognition
and machine learning. These concepts can be explained to students in terms of com-
puter algorithms that stimulate human-like behavior.

Use and Apply The second cognitive level is allowing students to use and apply AI
concepts and the related applications in various contexts. Lee et al. (2020) created
engaging learning experiences that integrate artificial intelligence and life science
for upper primary school students with game-based learning. A study conducted by
Ho et al. (2019) asked six primary school students to turn a number-guessing robot
into a self-learning lawn-bowling robot for a game of accuracy. Furthermore, the
“PepperBot” social robot with AI multi-capabilities is already available in various
Japanese schools. Eguchi et al. (2021) advocate including the PepperBot in the AI
curriculum for primary school students to interact, gain hands-on experience, and
program it while understanding the AI concepts and applying various AI functions.

Create and Evaluate Aside from understanding AI concepts and applying AI

through hands-on activities, a few researchers have extended AI literacy to the four
cognitive levels to provide students opportunities to develop higher-order thinking
and critical skills. There are various efforts underway to design and leverage various
learning tools for AI artifact-making activities such as creating machine learning
models; however, the learning outcomes of students are mixed. Google’s Teachable
Machine, for example, is a web-based tool that allows students to create machine
80 6 AI Literacy Education in Primary Schools

learning models to recognize images and poses (Shamir & Levin, 2021, 2022).
These two findings show that students are highly engaged in understanding machine
learning concepts, whereas students also perceive self-efficacy in constructing and
validating the neural network of an AI algorithm. Alternatively, in co-designing
machine learning application workshops, Toivonen et al. (2020) invited 36 primary
school participants to form groups of 4 or 5 to participate in a co-designing machine
learning application workshop. The model resulted in sound recognition, while the
prediction accuracy of students’ work was low, and only one group fulfilled the
required level of prediction accuracy. The results indicate that this machine learning
content is suitable for students to develop their higher-order thinking skills, but
there are practical pedagogical challenges in teaching AI learning content such as
what tools to use, how to train, how long, and how much is enough for students to
collect large enough and rich enough training data in order to solve the real-world
problems of projects.

AI Ethics Many researchers include AI ethics in empirical studies as overused or

poorly constructed AI could inflict irreparable harm on humans and society (Melsion
et al. (2021). Heinze et al. (2010) discussed with Grade 5 and 6 students the two
types of errors that humans and computers make. The notions that a science can be
proven wrong in the context of hypotheses and a computer can also make mistakes
and what this may signify are taught. For example, Melson et al. (2021) designed a
bias visualization tool of machine learning integrated into an online educational
platform (https://biaix.now.sh) which introduces not only the algorithmic bias
among children concretely but also the notions of ethics and societal impact of
AI. Students can become aware of AI ethics through discussing any concerns that
arise from different AI applications (Han et al., 2018).

RQ3: What learning tools have been used in AI literacy studies?

This section identifies several age-appropriate learning tools under the current AI
literacy primary schools landscape. An overview of these AI learning tools named
in these selected studies is presented in Table 6.2.
Among these four categories, the introduction of intelligent agents (11) has
become the most popular in primary school classrooms. The findings reveal that the
main goal of suggesting intelligent agents is to teach machine learning concepts, a
subset of AI concepts, which refer to a kind of data-driven thinking approach to
execute real-time computation and make decisions. Intelligent agents such as
Learning ML (Rodríguez-García et al., 2020, 2021; Voulgari et al., 2021) and
Google Teachable Machine (Ali et al., 2019; Ng et al., 2022; Toivonen et al., 2020)
are the appropriate learning tools to make such machine learning concepts
understandable.
Other than intelligent agents that focus on data-driven AI techniques, many hard-
ware, software, and unplugged activities are also targeted to scaffold AI concepts,
such as the ethical aspects of AI. In terms of hardware-focused learning tools, robot-
ics is commonly mentioned and seems to be promising in terms of supporting AI
6.2 Results and Discussion 81

Table 6.2 Learning tools in AI literacy primary school education

Type No Definitions Learning tools Studies
Hardware-­ 8 Use physical artifacts ArtBot, Jibo, and PopBots; Ali et al. (2019),
focused in AI literacy studies Zumi robot; Micro:Bit; Eguchi et al. (2021),
such as robots, small HuskyLens by DFRobot; Heinze et al. (2010),
single-board Pepper robot by SoftBank Ho et al. (2019), Lin
computers or robotics; Lego Mindstorm; et al. (2021), Narahara
microcontroller Lawn-bowling robot; and Kobayashi
devices Raspberry Pi (2018), Ng et al.
(2021), Voulgari et al.
(2021)
Software or 10 Use digital artifacts or Web application (HTML + Han et al. (2018),
open open-source platforms JavaScript); AI Chatbot, Mariescu-Istodor and
source-­ in AI literacy studies Scratch, Snap!, App Inventor Jormanainen (2019),
focused such as programming Grad-CAM; Python and Marques et al. (2020),
language, neural Tensorflow; Quick Draw, Melsion et al. (2021),
networks, machine image stylizer, gamified Narahara and
learning software Kahoot game; Thing Kobayashi (2018), Ng
toolkits Translator; Watson Speech et al. (2022), Shamir
to Text; Watson Visual and Levin (2021),
Recognition; Watson Natural Tkáčová et al. (2020),
Language Understanding; Toivonen et al.
Google Maps, Google (2020), Wei et al.
Drawings, and Google (2020)
Trends; Connect the Dots;
ITMS software, virtual
simulation of teaching
platform, SmartVoice
Intelligent 11 Intelligent agents to Google Teachable Machine; Ali et al. (2019),
agents teach AI by Droodle Creativity; Gong et al. (2020),
performing real-time AI-in-a-Box; Squirrel AI; Lee et al. (2020), Li
computation based on Primary AI; Scratch; code. and Song (2019), Ng
the inputs to make org; LearningML platform et al. (2021, 2022),
decisions such as (https://learningml.org.) Rodríguez-García
machine learning (2020, 2021), Shamir
trainers, chatbots to and Levin (2021),
build machine learning Toivonen et al.
models without coding (2020), Voulgari et al.
(2021)
Unplugged 8 Unplugged activities Abstract drawing; paper Ali et al. (2019),
to teach AI studies prototyping activity; writing Heinze et al. (2010),
without a computer, story with robots; role-­ Henry et al. (2021),
such as role-playing playing games; YouTube Ho et al. (2019),
and storytelling learning video; hand-­ Mariescu-Istodor and
drawing illustration; Jormanainen (2019),
unplugged activities Lucas (2012), Ng
(csunplugged.org); digital et al. (2021, 2022)
story creation

education. Jibo, PopBots (Ali et al., 2019), and lawn-bowling robots (Ho et al.,
2019) motivate and engage students in AI learning through games and competition.
On the other hand, software-focused learning tools such as Grad-Cam (Melsion
82 6 AI Literacy Education in Primary Schools

et al., 2021), Quick Draw and image stylizer (Ng et al., 2022; Tkáčová et al., 2020),
and AI Chatbots (Shamir & Levin, 2021) provide a convenient way to access, visu-
alize, and produce basic AI learning content.
In addition, Shamir and Levin (2021) designed an AI curriculum for Israeli stu-
dents to engage in conversations with AI Chatbot in order to understand the Turing
test concept. Alternatively, a few researchers have offered unplugged activities to
foster their conceptual understanding of AI such as abstract drawing, hand-drawing
illustration (Ali et al., 2019; Mariescu-Istodor & Jormanainen, 2019), and writing
stories with robots (Heinze et al., 2010; Ng et al. (2022).
RQ4: What assessment methods have been used in AI literacy studies?
As shown in Table 6.3, qualitative (n = 12), mixed methods (n = 7), and quantita-
tive (n = 5) are used in AI literacy empirical studies for data collection. It is noted
that the discussion paper and review papers are excluded in this section.
Qualitative evidence is collected through assessing students’ artifacts, role-play
performance, games and competition, interviews, and video recordings to examine
students’ motivation and engagement in learning AI as well as evaluate the suitabil-
ity of the teaching methods in primary school students’ classrooms. For example,
Lee et al. (2020) revealed that the test run driving a toy car on a miniature track
which involves students assembling, training, and testing the AI model can stimu-
late students’ curiosity to learn about AI. Others use the activity of role-play games,
students’ artifact for a lawn-bowling robot competition, and music gamification
competition to arouse students’ motivation, enhance their engagement, and cultivate
their basic AI literacy and computational thinking (Henry et al., 2021; Han et al.
2018; Ho et al. (2019). Vartiainen and Valtonen’s (2020) study focuses on evaluating
the content and teaching method through interviews and video recordings to
observe, explore, and explain the human-computer relationship when students inter-
act with a teachable machine. The result shows that students have “fun” and a “nice”

Table 6.3 Assessment methods

Assessment No
methods (37) Studies
Qualitative 12 Dai et al. (2020), Han et al. (2018), Heinze et al. (2010), Henry et al.
(2021), Ho et al. (2019), Mariescu-Istodor and Jormanainen (2019), Lee
et al. (2020), Li and Song (2019), Lucas (2009), Ottenbreit-Leftwich et al.
(2021), Toivonen et al. (2020), Vartiainen and Valtonen (2020)
Quantitative 5 Chai et al. (2020a, b, 2021), Lin et al. (2021), Rodríguez-Garciá et al.
(2021)
Mixed 7 Ali et al. (2019), Gong et al. (2020), Melsion et al. (2021), Ng et al.
method (2022), Shamir and Levin (2021, 2022), Voulgari et al. (2021)
Discussion 7 Eguchi et al. (2021), Narahara and Kobayashi (2018), Rodríguez-García
papers (2020), Steinbauer et al. (2021), Tkáčová et al. (2020), Touretzky et al.
(2019), Wei et al. (2020)
Review 6 Marques et al. (2020), Ng et al. (2021), Sanusi and Oyelere (2020), Su
articles et al. (2022), Tedre et al. (2021), Yang (2019)
6.3 Conclusions 83

learning experience which suggests that a teachable machine fosters the intellectual
curiosity of students to learn about AI (Vartiainen et al., 2020).
On the contrary, quantitative methods collect data through surveys and question-
naires to examine the learning outcomes of students and their perception of AI edu-
cation. For example, online pre- and post-questionnaires are designed for students
to evaluate their understanding of the concept of the machine learning metric and
data training (Rodríguez-Garciá et al., 2021). Others use surveys to examine stu-
dents’ perception of their behavioral intention, motivation, readiness, relevance, and
anxiety regarding AI education (Chai et al., 2020a, b, 2021) and the effectiveness of
the motivational model design in the context of AI learning in primary schools (Lin
et al., 2021).
Also, seven studies used mixed methods to collect data from multiple sources
including focus groups, questionnaire surveys, field visits, interviews, and artifact
assessments (Gong et al., 2020; Shamir & Levin, 2021; Voulgari et al., 2021).
Overall, artifact-based assessment, interviews, surveys, and questionnaires are
often used in data collection procedures, whereas games and competition have been
increased in the research design method recently. Apart from the Torrance test (Ali
et al., 2019), pre- and post-assessment of gender bias assessment (Melsion et al., 2021)
and online knowledge assessment (Rodríguez-Garciá et al., 2021) for evaluating stu-
dents’ learning outcomes, many studies use subjective measures such as self-reported
surveys or artifact assessment. These studies reveal the inadequacy of rubric-based,
evaluative mechanisms to assess the quality of suggested methodologies.

6.3 Conclusions

This paper presents a systematic review mapping process and offers an exploratory
perspective on AI education in today’s primary school setting. The learning content
described in the reviewed articles was grouped into four categories of know and
understand, use and apply, create and evaluate, and AI ethics. Surveys and question-
naires, artifact-based assessment, and interviews are often used in data collection
for assessment purposes. This study, in addition to shedding light on the learning
content and assessment methodologies of AI education in primary education, also
reveals the most commonly used pedagogical approaches as being project-based,
play/game-based, and collaborative/human-computer interactive teaching strategies
that will be suitable in primary school contexts. It illuminates that many researchers
use a combination of different pedagogical approaches and that the taught content
may change to meet the diverse requirements and cognitive skills of students and
collection procedure, whereas games and competition have been used in research
methods recently.
Although this review contributes to offering an exploratory perspective on AI
education in today’s primary school setting, one limitation lies in the scarcity of
available literature. AI literacy learning and teaching in primary school settings,
however, appear to be in its infancy. Future directions can be focused on this to
84 6 AI Literacy Education in Primary Schools

enrich the field of AI literacy in primary education. To advocate AI literacy educa-

tion for all, priority should be placed on developing an AI curriculum that includes
primary school students to guide educators to have lesson designs with age-­
appropriate pedagogies, learning tools, and assessment methodologies. Moreover,
the implementation of AI education also requires more professional development
for pre-service and in-service primary school teachers. For the purpose of develop-
ing AI educational initiatives, curriculum materials, and pedagogies that are appli-
cable to a wide range and diversity of educators and primary school students, further
studies on the teachers’ perspectives, instructional practices, and effectiveness of
pedagogical strategies on AI education are also of high relevance to promote AI
education in the primary schools landscape. We hope that this review will inspire
researchers, educators, and the government to begin the discussion on what to teach
and how to teach, implement, and evaluate AI literacy education for primary school
students in the future. In fact, countries have started their AI curricula in primary
school settings that are designed by researchers and collaborators and endorsed by
the government. Educators should start by helping students solidify their cognitive
development about technological knowledge and skill acquisition, learning, and
innovation skills. Students should identify AI’s societal impacts in their everyday
life. Moreover, we develop children’s learning about how technology works through
powerful metaphors, playful experience, and simulations in the form of hands-on
experiential activities. Finally, we work with students to develop their understand-
ing of how design choices impact others so that children can design their own intel-
ligent robot solutions to various problems.

References

Ackermann, E. (2001). Piaget’s constructivism, Papert’s constructionism: What’s the difference?

Future of Learning Group Publication, 5(3), 438.
Ali, S., Payne, B. H., Williams, R., Park, H. W., & Breazeal, C. (2019). Constructionism, eth-
ics, and creativity: Developing primary and middle school artificial intelligence education. In
International workshop on education in artificial intelligence k-12 (eduai’19) (pp. 1–4).
Bloom, B. S. (1956). Taxonomy of educational objectives: The classification of educational goals.
Cognitive domain.
Chai, C. S., Lin, P. Y., Jong, M. S. Y., Dai, Y., Chiu, T. K., & Huang, B. (2020a, August). Factors
influencing students’ behavioral intention to continue artificial intelligence learning. In 2020
International Symposium on Educational Technology (ISET) (pp.147–150). IEEE.
Chai, C., Lin, P., Jong, M., Dai, Y., & Chiu, T. (2020b). Primary students’ readiness for learning of
artificial intelligence: A case study in Beijing. In Proceedings of the 28th international confer-
ence on computers in education. Asia-pacific Society for Computers in education.
Chai, C. S., Lin, P. Y., Jong, M. S. Y., Dai, Y., Chiu, T. K., & Qin, J. (2021). Perceptions of and
behavioral intentions towards learning artificial intelligence in primary school students.
Educational Technology & Society, 24(3), 89–101.
Chiu, T. K., Meng, H., Chai, C. S., King, I., Wong, S., & Yam, Y. (2021). Creation and evalua-
tion of a pretertiary artificial intelligence (AI) curriculum. IEEE Transactions on Education.,
65, 30–39.
References 85

Dai, Y., Chai, C. S., Lin, P. Y., Jong, M. S. Y., Guo, Y., & Qin, J. (2020). Promoting students’
well-being by developing their readiness for the artificial intelligence age. Sustainability,
12(16), 6597.
Eguchi, A., Okada, H., & Muto, Y. (2021). Contextualizing AI education for K-12 students to
enhance their learning of AI literacy through culturally responsive approaches. KI-Künstliche
Intelligenz, 35(2), 153–161.
Gong, X., Tang, Y., Liu, X., Jing, S., Cui, W., Liang, J., & Wang, F. Y. (2020, October). K-9 arti-
ficial intelligence education in Qingdao: Issues, challenges and suggestions. In 2020 IEEE
International Conference on Networking, Sensing and Control (ICNSC) (pp. 1–6). IEEE.
Han, X., Hu, F., Xiong, G., Liu, X., Gong, X., Niu, X., … & Wang, X. (2018). Design of AI+ cur-
riculum for primary and secondary schools in Qingdao. In 2018 Chinese Automation Congress
(CAC) (pp. 4135–4140). IEEE.
Heinze, C. A., Haase, J., & Higgins, H. (2010). An action research report from a multi-year
approach to teaching artificial intelligence at the k-6 level. In First AAAI symposium on educa-
tional advances in artificial intelligence.
Henry, J., Hernalesteen, A., & Collard, A. S. (2021). Teaching artificial intelligence to K-12
through a role-playing game questioning the intelligence concept. KI-Künstliche Intelligenz,
35(2), 171–179.
Ho, J. W., Scadding, M., Kong, S. C., Andone, D., Biswas, G., Hoppe, H. U., & Hsu, T. C. (2019).
Classroom activities for teaching artificial intelligence to primary school students. In
Proceedings of international conference on computational thinking education (pp. 157–159).
Lee, S., Mott, B., Ottenbriet-Leftwich, A., Scribner, A., Taylor, S., Glazewski, K., … & Lester,
J. (2020, June). Designing a collaborative game-based learning environment for ai-infused
inquiry learning in elementary school classrooms. In Proceedings of the 2020 ACM conference
on innovation and technology in computer science education (pp. 566–566).
Li, K., & Song, S. (2019, June). Application of artificial intelligence in primary and secondary
schools: A case study of scratch. In International conference on applications and techniques in
cyber security and intelligence (pp. 2026–2030). Springer
Lin, P. Y., Chai, C. S., Jong, M. S. Y., Dai, Y., Guo, Y., & Qin, J. (2021). Modeling the structural
relationship among primary students’ motivation to learn artificial intelligence. Computers and
Education: Artificial Intelligence, 2, 100006.
Lucas, J. M. (2009). K-6 outreach using “computer science unplugged”. Journal of Computing
Sciences in Colleges, 24(6), 62–63.
Mariescu-Istodor, R., & Jormanainen, I. (2019, November). Machine learning for high school stu-
dents. In Proceedings of the 19th Koli calling international conference on computing education
research (pp. 1–9).
Marques, L. S., Gresse von Wangenheim, C., & Hauck, J. C. (2020). Teaching machine learning in
school: A systematic mapping of the state of the art. Informatics in Education, 19(2), 283–321.
Melsión, G. I., Torre, I., Vidal, E., & Leite, I. (2021, June). Using explainability to help children
understand gender bias in AI. In Interaction design and children (pp. 87–99).
Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & PRISMA Group*. (2009). Preferred reporting
items for systematic reviews and meta-analyses: The PRISMA statement. Annals of Internal
Medicine, 151(4), 264–269.
Narahara, T., & Kobayashi, Y. (2018). Personalizing homemade bots with plug & play AI for
STEAM education. In SIGGRAPH Asia 2018 technical briefs (pp. 1–4).
Ng, D. T. K., Leung, J. K. L., Chu, S. K. W., & Qiao, M. S. (2021). Conceptualizing AI literacy: An
exploratory review. Computers and Education: Artificial Intelligence, 2, 100041.
Ng, D. T. K., Luo, W. Y., Chan, H. M. Y., & Chu, S. K. W. (2022). An examination on primary
students’ development in AI literacy through digital story writing. Computers & Education:
Artificial Intelligence, 100054.
Ottenbreit-Leftwich, A., Glazewski, K., Jeon, M., Hmelo-Silver, C., Mott, B., Lee, S., &
Lester, J. (2021, March). How do elementary students conceptualize artificial intelligence?
86 6 AI Literacy Education in Primary Schools

In Proceedings of the 52nd ACM technical symposium on computer science education

(pp. 1261–1261).
Rodríguez-García, J. D., Moreno-León, J., Román-González, M., & Robles, G. (2020, October).
Introducing artificial intelligence fundamentals with LearningML: Artificial intelligence made
easy. In Eighth international conference on technological ecosystems for enhancing multicul-
turality (pp. 18–20).
Rodríguez-García, J. D., Moreno-León, J., Román-González, M., & Robles, G. (2021, March).
Evaluation of an online intervention to teach artificial intelligence with LearningML to
10-16-year-old students. In Proceedings of the 52nd ACM technical symposium on computer
science education (pp. 177–183).
Sanusi, I. T., & Oyelere, S. S. (2020, October). Pedagogies of machine learning in K-12 context.
In 2020 IEEE Frontiers in Education Conference (FIE) (pp. 1–8). IEEE.
Shamir, G., & Levin, I. (2021). Neural network construction practices in elementary school.
KI-Künstliche Intelligenz, 35(2), 181–189.
Shamir, G., & Levin, I. (2022). Teaching machine learning in elementary school. International
Journal of Child-Computer Interaction, 31, 100415.
Steinbauer, G., Kandlhofer, M., Chklovski, T., Heintz, F., & Koenig, S. (2021). A differentiated
discussion about AI education K-12. KI-Künstliche Intelligenz, 35(2), 131–137.
Su, J., Zhong, Y., & Ng, D. T. K. (2022). A meta-review of literature on educational approaches for
teaching AI at the K-12 levels in the Asia-Pacific region. Computers and Education: Artificial
Intelligence, 100065.
Talbott, E., Maggin, D. M., Van Acker, E. Y., & Kumm, S. (2018). Quality indicators for reviews
of research in special education. Exceptionality, 26(4), 245–265.
Tedre, M., Toivonen, T., Kahila, J., Vartiainen, H., Valtonen, T., Jormanainen, I., & Pears, A. (2021).
Teaching machine learning in K–12 classroom. In Pedagogical and technological trajectories
for artificial intelligence education. IEEE Access, 9, 110558-110572.
Tkáčová, Z., Šnajder, L. U., & Guniš, J. (2020, October). Artificial intelligence–a new topic in
computer science curriculum at primary and secondary schools: Challenges, opportunities,
tools and approaches. In 2020 43rd International Convention on Information, Communication
and Electronic Technology (MIPRO) (pp. 747–749). IEEE.
Toivonen, T., Jormanainen, I., Kahila, J., Tedre, M., Valtonen, T., & Vartiainen, H. (2020, July).
Co-designing machine learning apps in K–12 with primary school children. In 2020 IEEE 20th
International Conference on Advanced Learning Technologies (ICALT) (pp. 308–310). IEEE.
Touretzky, D., Martin, F., Seehorn, D., Breazeal, C., & Posner, T. (2019, February). Special ses-
sion: AI for K-12 guidelines initiative. In Proceedings of the 50th ACM technical symposium
on computer science education (pp. 492–493).
Vartiainen, H., Tedre, M., & Valtonen, T. (2020). Learning machine learning with very young
children: Who is teaching whom? International Journal of Child-Computer Interaction,
25, 100182.
Voulgari, I., Zammit, M., Stouraitis, E., Liapis, A., & Yannakakis, G. (2021, June). Learn to
machine learn: Designing a game based approach for teaching machine learning to primary
and secondary education students. In Interaction design and children (pp. 593–598).
Wei, Q., Li, M., Xiang, K., & Qiu, X. (2020, August). Analysis and strategies of the profes-
sional development of information technology teachers under the vision of artificial intelli-
gence. In 2020 15th International Conference on Computer Science & Education (ICCSE)
(pp. 716–721). IEEE.
Yang, X. (2019). Accelerated move for AI education in China. ECNU Review of Education, 2(3),
347–352.
Chapter 7
AI Literacy Education in Secondary
Schools

As AI literacy has grown its popularity across countries and regions around the
world to design and implement AI curricula in secondary school levels. According
to the report of UNESCO (2022), 11 member states have designed, endorsed, and
implemented AI government-endorsed curricula. In the review of Ng et al. (2021b),
over 14 countries around the world (including the United States, China, Spain,
Hong Kong, Finland, Brazil, and Germany) have begun to promote secondary stu-
dents’ AI competences and equip them with related knowledge, skills, and attitudes.
Based on the literature review in the previous chapters, we can find that the learn-
ing content, pedagogy, and tools of AI literacy education are still continuing to
develop to best foster students’ AI literacy. This chapter provides an overview of
how AI literacy is developed at junior and secondary levels. After illustrating the
research method including literature search and data analysis, this chapter reviews
the pedagogy, content, tools, and assessment methods used in the selected studies.
We analyzed the current state of AI literacy education and suggested future direc-
tions regarding how to best teach and learn AI in secondary levels. Four research
questions (RQ) formed the basis of this review:
RQ1: What are the pedagogical strategies used at the secondary level?
RQ2: What learning contents are appropriate for students at their junior and senior
secondary level?
RQ3: What are the learning tools used at the secondary level?
RQ4: What are the assessment methods used at the secondary level?

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 87

D. T. K. Ng et al., AI Literacy in K-16 Classrooms,
https://doi.org/10.1007/978-3-031-18880-0_7
88 7 AI Literacy Education in Secondary Schools

7.1 Method

To ensure that the search encompassed all of the evidence-based SSCI literature, the
researchers used two trusted citation index databases, Web of Science and Scopus.
First, we searched the two databases for publications published between 2016 and
2022 using the phrase (“AI” OR “artificial intelligence” OR “ artificial intelligence
literacy” OR “deep learning” OR “machine learning” OR “neural network*” OR
“natural language processing” OR “chatbot”) AND (“secondary school” OR “mid-
dle school” OR “secondary education” OR “pre-tertiary education” OR “secondary
student” OR “middle school student” OR “pre-tertiary student”) AND (“learning”
OR “teaching” OR “pedagogy” OR “curriculum”) in either the title, the abstract,
main text, or keywords were downloaded and reviewed by the researchers. As of 3
March 2022, this gave us a total of 307 articles in the 2 databases.
Then, two experienced researchers decided if they were appropriate for the goal
of this study. During this search, a set of inclusion and exclusion criteria were estab-
lished to avoid biases in the articles selection. To begin, all of the studies that were
chosen had to be journal articles, discussion papers, case studies, or conference
papers from the aforementioned databases. Second, the studies have to be relevant
to AI literacy and teaching/learning AI concepts in the field of education (e.g., arti-
ficial intelligence, machine learning, deep learning, natural language, neural net-
works, chatbots). Sun’s article, for example, was excluded because it used 5G and
AI technologies in English instruction instead of teaching AI concepts. However,
because of a lack of peer review, editorials and novels are not included. Following
the exclusion of irrelevant studies, a total of 38 articles were discovered. An over-
view of the search protocol in a PRISMA diagram is presented in Fig. 7.1.
The selected papers were qualitatively categorized using the constant compara-
tive method espoused by Glaser (1965). We examined the major content of the arti-
cles and identified related meaningful concepts for thematic analysis. To verify
coding reliability, all of the papers were reviewed by two researchers who resolved
conflicts by discussing the disputed studies in order to ensure inter-rater reliability.
The data were then examined and summarized using a coding scheme. The scheme
was modified from Ng’s (2021a, b) scheme, which includes pedagogical approaches,
technologies/tools used, learning material, and learning outcomes.

7.2 Results and Discussion

RQ1: What are the pedagogical strategies used in AI literacy studies?

As indicated in Table 7.1, this section outlines the three major pedagogical meth-
ods and strategies used in the studies: project-/problem-based learning (24), col-
laborative learning (23), and experiential learning (15).
First, project-/problem-based learning is the most often used pedagogy to foster
students’ AI literacy. For example, Vachovsky et al. (2016) involved 24 girls in
7.2 Results and Discussion 89

Fig. 7.1 PRISMA diagram

Table 7.1 Pedagogical approaches for AI literacy education

Percent
Pedagogies Descriptions Sample studies N (%)
Project-/ Learning approaches to engage Rodríguez-García 24 63.2
problem-based students to gain knowledge and skills (2020),
learning by working to investigate authentic Sakulkueakulsuk et al.
questions, problems, or challenges (2018), Wan et al.
(2020)
Collaborative This pedagogy allows students to Deng et al. (2021), Gao 23 60.5
learning learn how to communicate and work and Wang (2019), Gong
with classmates to gain AI knowledge et al. (2018)
and manipulate with smart devices
Experiential Process of learning by doing through Chiu et al. (2021), 15 39.5
learning hands-on experiences and reflection; Morris (2020), Tamborg
students could gain better et al. (2022)
understanding to connect theories and
knowledge learned in the classroom to
real-world situations
90 7 AI Literacy Education in Secondary Schools

authentic projects to learn computer vision, robotics, and natural language process-
ing in a summer camp. According to the survey, 95.8% of students believe that the
projects they built can benefit society. Furthermore, students said the course was
interesting (83.3%) and increased their confidence in using AI (75%). Rodrguez-­
Garca et al. (2020) presented the LearningML projects (a low-floor high-ceiling
platform to learn machine learning by doing) to provide the principles of machine
learning to students in order for them to acquire knowledge and become critical
thinking citizens. We can observe that project-/problem-based learning gives
authentic settings such as constructing models of athletic moves (Zimmermann-­
Niefield et al., 2019) and meaningful scientific and STEM contexts that could moti-
vate them to learn AI through a sense of authenticity and real-world applicability
(Sakulkueakulsuk et al., 2018; Wan et al., 2020).
Collaborative learning allows students to learn how to communicate and collabo-
rate with classmates to gain AI knowledge and manipulate with smart devices (e.g.,
Deng et al., 2021; Gao & Wang, 2019; Gong et al., 2018). Gong et al. (2018), for
example, engaged students to various positions such as project managers, software
designers, hardware designers, and art designers in order to build smart vehicles in
authentic settings. Another study, done by Gao and Wang (2019), invited students to
act as buyers and sellers in order to identify problems and shortcomings about intel-
ligent functions in smart home systems. Kaspersen (2021) assigned three to four
students in a group to design ML models that predict if a person will vote for a
particular political party. A combination of collaborative learning and project-/
problem-­based learning could significantly improve students’ higher-order thinking
skills, such as problem-solving, leadership, project management, and creativity
(e.g., Deng et al., 2021; Gong et al., 2018).
The third common method is to engage students in playful and experiential
learning by using Teachable Machine (Chiu et al., 2021; Tamborg et al., 2022),
Code.org games (Ng & Chu, 2021), intelligent agents, chatbots, Cognimates (Gong
et al., 2018), and syntax-based programs (e.g., Python) (Gong et al., 2018; Gunasilan,
2021) (Estevez et al., 2019; Kahn et al., 2018). These activities give students hands-
­on experience investigating what AI is, visualizing complex ideas (Reyes et al.,
2020), and developing building ML models (Sakulkueakulsuk et al., 2018).
Although these activities provide students with hands-on experience to scaffold AI,
most studies further applied minds-on collaborative projects to encourage students
to further build knowledge by creating digital and tangible artifacts in construction-
ist ways. In this way, educators could help students achieve higher cognition levels
and apply AI skills and knowledge to address real-world problems for future learn-
ing and career challenges (Chai et al., 2020).
RQ2: What learning tools have been used in AI literacy studies?
As technology advances, more age-appropriate learning artifacts enable students
to visualize the operations of complex concepts that were previously impossible.
Aligned with Sanusi et al. (2021) and Ng et al. (2021a, b), four major categories of
resources and technologies have been identified: conversational agents,
7.2 Results and Discussion 91

Table 7.2 Learning artifacts

Definition Learning artifacts examples Sample studies
Hardware-­ Use physical artifacts to CUHKiCar, educational robot, Chai et al.
focused learn AI such as robotics, drones, VR devices Alpha dog (2020), Chiu
artifacts sensors and Arduino robot, Raspberry Pi Raspbian, et al. (2021),
devices four-wheel drive chassis, Tamborg et al.
Micro:Bits (2022)
Software-­ Use digital artifacts to Prolog, image stylizer, Colab Chiu et al.
focused learn AI such as block/ (Python programming), Jupyter (2021), Norouzi
artifacts syntax-based notebooks, Blockly, WebAPPs, et al. (2020),
programming and cognitive services, Google Gunasilan
simulation Collaborator, Scratch, Snap!, MIT (2021)
STEP Lab, game design
Intelligent Use intelligent agents Scratch, Google’s Teachable Rodríguez-­
agents such as expert systems, Machine, LearningML platform, García (2020),
machine learning ecraft2learn, Machine Learning for Sabuncuoglu
trainers, chatbots to build Kids, Cognimates, Code.org AI (2020),
their custom machine Ocean, Face AI Scratch, Machine Wan et al. (2020)
learning models without Learning for kids, AI model trainer,
coding ecraft2learn, SmileyCluster,
chatbots
Unplugged Use learning activities to Lectures, career talk, textbook, case Gunasilan
learn AI without a study, webinar, role-playing, (2021), Deng
computer such as lecture, storytelling, debating et al. (2021),
case study, role-playing, Gong et al.
and storytelling (2018)

programming environments, robotics, and unplugged activity. Table 7.2 is a list of

examples of how to use AI learning artifacts to learn AI.
First, hardware- and software-focused artifacts engaged students in making and
creating AI-driven applications using physical and digital artifacts. Chiu et al.
(2021), for example, constructed a robotic car named CUHKiCar with built-in AI
functions to provide interactive learning experiences for students doing face-­
tracking and line-following tasks. It was discovered that students improved signifi-
cantly in perceived knowledge gain, confidence, motivation, and AI readiness. Chai
et al. (2020) conducted a study in which students could use Alpha dog robots and
design algorithms to recognize physical characteristics such as temperature, voice,
face, and images. It helps students to build AI solutions using AI and mathematics
knowledge (e.g., calculus and statistics).
Second, it has been determined that most of the hardware and software should be
manipulated with built-in and add-on AI-driven features that enable students to
build machine learning intelligent agents and machine learning models without the
need for programming (Chiu et al., 2021; Kaspersen et al., 2021). This provides
students with opportunities to lower the learning barrier and gain access to more
advanced concepts that were previously impossible. In our selected studies, most of
them focused on tool-based learning to provide students with hands-on experiences
to learn the fundamental AI/computer science knowledge, skills, and concepts. The
92 7 AI Literacy Education in Secondary Schools

tools allow students to visualize the complex knowledge and encourage them to col-
lect between AI applications and the underlying knowledge. Chatbots (Rodrguez-
Garca et al., 2020), Scratch, and Teachable Machine (Tamborg et al., 2022), for
example, inspire students to solve authentic problems using these tools. Students
can learn the functional and critical parts of using AI-driven technologies through
tool-based AI learning. In addition to learning technical knowledge and skills (e.g.,
computer vision, virtual reality, art design) (Gong et al., 2018), students’ critical
thinking skills are required to express and apply knowledge, as well as communi-
cate and interact with the tools to solve authentic problems (Kaspersen et al., 2021).
RQ3: What learning contents are appropriate for junior and senior secondary
students in AI literacy studies?
This section provides a summary of the learning content that secondary students
need to master at the junior (14) and senior (22) levels (see Table 7.3). Educators
need to understand the cognitive development of each step of AI learning in order to
design developmentally appropriate instruction. Prior research has shown that
junior secondary students should focus on preliminary and simple AI concepts such
as machine learning, natural language processing, and Turing tests in the junior AI
curriculum (Chiu et al., 2021; Fernández-Martnez et al., 2021; Ng & Chu, 2021).
Educators should design hands-on experiential learning for students to taste and use
related AI applications, as well as explore their benefits, challenges, ethical con-
cerns, and shortcomings (Sabuncuoglu, 2020; Wan et al., 2020). In this manner,

Table 7.3 Learning contents in junior and secondary school levels

Levels Learning contents samples Sample studies
Junior Experiencing AI: Using AI applications, benefits, and Chiu et al. (2021),
level disadvantages of using AI, machine learning, face Fernández-Martínez
recognition, image stylizer, machine generation of creative et al. (2021), Ng and
content, experimentation of using AI technologies Chu (2021)
Simple AI concepts: History/introduction/recent
development of AI and its subareas, differences between
humans and machines, natural language processing, scratch
activities on machine learning and image recognition,
machine learning
Societal impacts and AI ethics: AI ethics, societal impacts
of AI, machine reasoning, problem-solving
Senior Complex AI topics: Natural language processing, computer Kaspersen et al. (2021),
level vision, cognition, biomedical informatics, robotics, Kahn et al. (2018),
information networks, human-robot interactions, Zhang and Du (2008)
computational sustainability
AI technical components: Fisher’s exact test, inductive
reasoning, nearest neighbor algorithm, correlation, graph
search algorithms, computational game theory, optimization,
agent-based modeling, probabilistic reasoning
AI literacy: Understanding how ML works, the process
behind creating ML models, and the ability to reflect on its
personal and societal implications
7.2 Results and Discussion 93

students needed to apply these knowledge and skills to solve problems using well-­
defined hardware, software, and intelligent agents.
Senior secondary students could achieve greater cognition levels to develop tech-
nical algorithms and components (e.g., Fisher’s exact test, inductive reasoning,
nearest neighbor algorithm, correlation, graph search) (Vachovsky et al., 2016).
They could experiment with more complex concepts like computational game the-
ory, agent-based modeling, probabilistic reasoning, and graph theories (e.g., Estevez
et al., 2019; Reyes et al., 2020). Students at both levels could create AI-driven solu-
tions and models after knowing and understanding the AI knowledge, concerts, and
skills process. Furthermore, students at both the junior and senior levels were
required to explore the humanistic, sociological, and ethical implications of tech-
nology (Kaspersen et al., 2021). Additionally, students at all levels need to learn
important competences (e.g., critical thinking, communication, collaboration, and
creativity) in twenty-first century skills that will enable them to succeed in school
and in their future workplace (e.g., Fernández-Martnez et al., 2021).
In general, our findings are consistent with Touretzky et al. (2019)’s five “major
ideas” of AI, which state that students may learn how to utilize AI/computers to
perceive the world using sensors, design with AI agents to maintain representation,
reasoning, and learning from data. They were urged to apply intelligent agents to
interact with humans in a natural way. Finally, students could learn how AI can have
an impact on our digital society in both beneficial and detrimental ways. The cate-
gorization is also consistent with Ng’s three dimensions of AI knowledge: AI con-
cepts, practices, and perspectives (2021a, b). Using a sound cognition framework to
foster AI literacy, educators should select appropriate levels of knowledge, con-
cepts, and skills to meet the students’ learning needs and development.
RQ4: What assessment methods have been used in AI literacy studies?
Researchers employed quantitative (14) and qualitative (27) assessments to
investigate how children enhance their AI literacy skills (see Table 7.4). In this RQ,
we double-classified “mixed-method research” into quantitative and qualitative
evaluations, but the discussion papers were not coded.
Quantitative Methods Surveys and questionnaires were designed to assess stu-
dents’ knowledge acquisition via knowledge tests (e.g., Why do you think large
amounts of data might matter?) and students’ perceived abilities in order to better
understand secondary students’ AI literacy development (e.g., I have general knowl-
edge about how AI is used today.) Zimmermann-Niefield et al., 2019; Chiu et al.,
2021). Chiu et al. (2021) designed a questionnaire to understand students’ intrinsic
motivation, AI readiness, perceived abilities, and confidence. Chai et al. (2020)
designed a 41-item questionnaire to assess students’ AI literacy development, sub-
ject norms, anxiety, perceived usefulness of AI, AI for social good, attitude and
confidence in using AI, and behavioral intention. These questions allowed teachers
to investigate students’ noncognitive perceptions toward AI literacy education.
Sakulkueakulsuk et al. (2018) used surveys to evaluate students’ accuracy rate of
machine learning models to categorize the quality of mangoes throughout various
94 7 AI Literacy Education in Secondary Schools

Table 7.4 Assessment constructs and tools to evaluate Students’ AI Learning

Research Constructs and tools (Ng et al., Sample
methods 2021b) Some examples studies
Quantitative Use knowledge tests to assess “What is the importance of Wan et al.
(14) students’ AI cognitive gain and similarity when clustering a (2020)
abilities dataset?”
Use perceived questionnaire to “The content of this AI class is Chai et al.
assess the noncognitive relevant to my interests.” (2020), Chiu
aspects, including perceived “In this AI class, I prefer the et al. (2021)
ability, confidence in using AI, materials that really challenge me
intelligence, truthfulness, so that I can learn new things.”
perceived understanding,
subjective norms, AI anxiety,
perceived usefulness of AI, AI
for social good, attitude toward
using AI, confidence in
learning AI, learning
behavioral intention, AI
optimism, relevance, AI
awareness, career adaptability
skills
Qualitative Use videos, documents, “From students feedback, Gunasilan
(27) pictures, presentations, students reported that they were (2021), Lee
students’ interactions with AI excited and happy to try an et al. (2021),
agents and projects to examine activity they’ve never tried before Druga (2019)
students’ AI cognitive and in a classroom and most had an
noncognitive abilities underlying sense of
sportsmanship to prove each
other’s views right/wrong”

stages, as well as other aspects such as futuristic thinking, engagement, interactivity,

and interdisciplinary thinking.

Qualitative Methods To investigate students’ satisfaction, motivations, and cogni-

tion levels, qualitative data was obtained through capturing students’ work, field
notes, project presentations, and interviews. Lee et al. (2021), for example, pre-
sented students’ daily reflections and interviews about their workshop experiences
in order to demonstrate their excitement while working on this activity. Students
were also able to internalize their learning experience through interviews and relate
to the ethical implications of technology design (e.g., How accurate is your model?
What is the possibility for bias?), the challenges of learning AI, and their future job
prospects (e.g., What will your future career be?). Sintov et al. (2016) designed a
role-playing board game as a final project to reinforce students’ decision-making.
Druga (2019) used field observations to record students’ interactions with AI
agents and a three-attribute AI perception questionnaire to assess how 102 children
(7–12 years old) interacted with and perceived their AI agents in their lessons. These
three attributes measure whether the agents are smarter, more trustworthy, and
understand them (e.g., “What do you think of Google Voice, an AI-enabled agent?”).
7.3 Conclusions 95

The most fun features, according to the children, were playing beatbox and music,
taking pictures, and playing games. Gunasilan (2021) collected feedback after
debate activities to reflect and refine the instruction design in terms of competitive-
ness, enjoyment, teamwork, self-reflection, and peer assessment in an evaluation
session.

7.3 Conclusions

AI has grown popular and widely used across industries as a result of the fourth
industrial revolution, owing to increased data volumes, advanced algorithms, and
improvements in computing power and storage (Reed & Dongarra, 2015). Countries
have begun to design and implement AI curricula to help students develop technol-
ogy skills that will help them in their future studies and careers.
AI literacy is necessary to update the twenty-first century digital literacy skill
sets for citizens and students so that they could be more competent and ready for
their living, studies, and career in today’s AI-driven world. After several years of
implementation, AI curricula have been implemented to enable students to use AI
knowledge and related technologies to facilitate their learning and build creations.
This is the first review that summarizes the existing evidence of AI literacy educa-
tion in secondary school settings in terms of research backgrounds, methodological
approaches, pedagogical strategies used in the AI courses, learning tools that are
used in the AI courses, learning content, assessment methods, and learning out-
comes. We noticed that our findings were consistent with recent reviews (e.g., Ng
et al., 2021a, b; Marques et al., 2020) that we could adapt the Bloom’s taxonomy to
understand the cognition gains of AI knowledge, concepts, and skills, as well as the
TPACK model to understand the instructional design of selecting appropriate tech-
nologies/tools, pedagogies, and learning contents to teach students AI. Furthermore,
in the twenty-first century competencies that bring up with digital citizens in today’s
global community, we identified important competences such as the 4Cs (commu-
nication, collaboration, critical thinking, and creativity). Students need to be able to
communicate and create their thoughts, ideas, and solutions in order to solve future
challenges and boost their competitiveness, in addition to knowing AI concepts and
using AI applications ethically. This review contributes to providing a summary of
the up-to-date literature to inform researchers, policymakers, and educators about
how to effectively develop students’ AI literacy at the pre-tertiary level.
Several limitations were noted, first, because the majority of the publications
(21) were conference papers and half of the articles (15) used qualitative research
methodologies. It was acknowledged that AI literacy is still an emerging issue and
the majority of study was exploratory in nature. We foresee that future research
design will shift to be more empirical and use rigorous research methods (e.g.,
quasi-experiment, design-based research) using interventions and control groups. A
more comprehensive data analysis (e.g., t-test, ANOVA, factor analysis, regression,
structural equation modeling) should be used. Second, there are few questionnaires
96 7 AI Literacy Education in Secondary Schools

available to evaluate secondary students’ AI literacy, and none of them have been
validated. Future study should focus on developing AI literacy measures, such as
surveys and questionnaires, and assessing the scales’ reliability and validity. Finally,
establishing theoretical and pedagogical frameworks to assist policymakers, educa-
tors, and instructional designers with age-appropriate pedagogies, learning artifacts,
and assessment methods must be prioritized in order to advance this research field.

References

Chai, C. S., Lin, P. Y., Jong, M. S. Y., Dai, Y., Chiu, T. K., & Huang, B. (2020, August). Factors
influencing students; behavioral intention to continue artificial intelligence learning. In 2020
International Symposium on Educational Technology (ISET) (pp. 147–150). IEEE.
Chai, C. S., Lin, P. Y., Jong, M. S. Y., Dai, Y., Chiu, T. K., & Qin, J. (2021). Perceptions of and
behavioral intentions towards learning artificial intelligence in primary school students.
Educational Technology & Society, 24(3), 89–101.
Chiu, T. K., Meng, H., Chai, C. S., King, I., Wong, S., & Yam, Y. (2021). Creation and evalua-
tion of a pretertiary artificial intelligence (AI) curriculum. IEEE Transactions on Education.,
65, 30–39.
Deng, W., Huang, X., Liu, Q., & Wang, Z. (2021, December). Curriculum design of artificial intel-
ligence in middle school-taking posture recognition as an example. In 2021 Tenth International
Conference of Educational Innovation Through Technology (EITT) (pp. 310–315). IEEE.
Druga, S., Vu, S. T., Likhith, E., & Qiu, T. (2019). Inclusive AI literacy for kids around the world.
In Proceedings of FabLearn 2019 (pp. 104–111).
Estevez, J., Garate, G., & Graña, M. (2019). Gentle introduction to artificial intelligence for high-­
school students using scratch. IEEE Access, 7, 179027–179036.
Fernández-Martínez, C., Hernán-Losada, I., & Fernández, A. (2021). Early introduction of AI in
Spanish middle schools. A motivational study. KI-Künstliche Intelligenz, 35(2), 163–170.
Gao, J., & Wang, L. (2019, August). Reverse thinking teaching discussion in high school informa-
tion technology under new curriculum standards. In 2019 14th International Conference on
Computer Science & Education (ICCSE) (pp. 222–226). IEEE.
Glaser, B. G. (1965). The constant comparative method of qualitative analysis. Social Problems,
12(4), 436–445.
Gong, X., Wu, Y., Ye, Z., & Liu, X. (2018, June). Artificial intelligence course design: iSTREAM-­
based visual cognitive smart vehicles. In 2018 IEEE Intelligent Vehicles Symposium (IV)
(pp. 1731–1735). IEEE.
Gunasilan, U. (2021). Debate as a learning activity for teaching programming: A case in the sub-
ject of machine learning. Higher Education, Skills and Work-based Learning, 12, 705–718.
Kahn, K. M., Megasari, R., Piantari, E., & Junaeti, E. (2018). AI programming by children using
snap! Block programming in a developing country. Retrieved from https://ecraft2learn.
github.io/ai/publications/EC-­TEL_2018_source-­files_48%20kk%20edits%20changes%20
accepted.pdf
Kaspersen, M. H., Bilstrup, K. E. K., Van Mechelen, M., Hjorth, A., Bouvin, N. O., & Petersen,
M. G. (2021, June). VotestratesML: A high school learning tool for exploring machine learning
and its societal implications. In FabLearn Europe/MakeEd 2021-An international conference
on computing, design and making in education (pp. 1–10).
Lee, I., Ali, S., Zhang, H., DiPaola, D., & Breazeal, C. (2021). Developing middle school students’
AI literacy. In Proceedings of the 52nd ACM technical symposium on computer science educa-
tion (pp. 191–197).
References 97

Marques, L. S., Gresse von Wangenheim, C., & Hauck, J. C. (2020). Teaching machine learning in
school: A systematic mapping of the state of the art. Informatics in Education, 19(2), 283–321.
Morris, T. H. (2020). Experiential learning–a systematic review and revision of Kolb’s model.
Interactive Learning Environments, 28(8), 1064–1077.
Ng, D. T. K., & Chu, S. K. W. (2021). Motivating students to learn AI through social networking
sites: A case study in Hong Kong. Online Learning, 25(1), 195–208.
Ng, D. T. K., Leung, J. K. L., Chu, K. W. S., & Qiao, M. S. (2021a). AI literacy: Definition, teach-
ing, evaluation and ethical issues. Proceedings of the Association for Information Science and
Technology, 58(1), 504–509.
Ng, D. T. K., Leung, J. K. L., Chu, S. K. W., & Qiao, M. S. (2021b). Conceptualizing AI literacy:
An exploratory review. Computers and Education: Artificial Intelligence, 2, 100041.
Norouzi, N., Chaturvedi, S., & Rutledge, M. (2020, April). Lessons learned from teaching machine
learning and natural language processing to high school students. In Proceedings of the AAAI
conference on artificial intelligence (vol. 34, no. 09, pp. 13397–13403).
Reed, D. A., & Dongarra, J. (2015). Exascale computing and big data. Communications of the
ACM, 58(7), 56–68.
Reyes, A. A., Elkin, C., Niyaz, Q., Yang, X., Paheding, S., & Devabhaktuni, V. K. (2020, August).
A Preliminary work on visualization-based education tool for high school machine learning
education. In 2020 IEEE Integrated STEM Education Conference (ISEC) (pp. 1–5). IEEE.
Rodríguez-García, J. D., Moreno-León, J., Román-González, M., & Robles, G. (2020, October).
Introducing artificial intelligence fundamentals with LearningML: artificial intelligence made
easy. In Eighth international conference on technological ecosystems for enhancing multicul-
turality (pp. 18–20).
Sabuncuoglu, A. (2020). Designing one year curriculum to teach artificial intelligence for middle
school. In Proceedings of the 2020 ACM conference on innovation and technology in computer
science education (pp. 96–102).
Sakulkueakulsuk, B., Witoon, S., Ngarmkajornwiwat, P., Pataranutaporn, P., Surareungchai, W.,
Pataranutaporn, P., & Subsoontorn, P. (2018, December). Kids making AI: Integrating machine
learning, gamification, and social context in STEM education. In 2018 IEEE international con-
ference on Teaching, Assessment, and Learning for Engineering (TALE) (pp. 1005–1010). IEEE.
Sanusi, I. T., Oyelere, S. S., Agbo, F. J., & Suhonen, J. (2021, October). Survey of resources
for introducing machine learning in K-12 context. In 2021 IEEE Frontiers in Education
Conference (FIE) (pp. 1–9). IEEE.
Sintov, N., Kar, D., Nguyen, T., Fang, F., Hoffman, K., Lyet, A., & Tambe, M. (2016, March). From
the lab to the classroom and beyond: extending a game-based research platform for teaching
AI to diverse audiences. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol.
30, No. 1).
Tamborg, A. L., Elicer, R., & Spikol, D. (2022). Programming and computational thinking in
mathematics education. KI-Künstliche Intelligenz, 1–9.
Touretzky, D., Gardner-McCune, C., Martin, F., & Seehorn, D. (2019, July). Envisioning AI for
K-12: What should every child know about AI? In Proceedings of the AAAI conference on
artificial intelligence (vol. 33, no. 01, pp. 9795–9799).
UNESCO. (2022). K-12 AI curricula: A mapping of government-endorsed AI curricula. Retrieved
from https://unesdoc.unesco.org/ark:/48223/pf0000380602
Vachovsky, M. E., Wu, G., Chaturapruek, S., Russakovsky, O., Sommer, R., & Fei-Fei, L. (2016,
February). Toward more gender diversity in CS through an artificial intelligence summer pro-
gram for high school girls. In Proceedings of the 47th ACM technical symposium on computing
science education (pp. 303–308).
Wan, X., Zhou, X., Ye, Z., Mortensen, C. K., & Bai, Z. (2020, June). SmileyCluster: Supporting
accessible machine learning in K-12 scientific discovery. In Proceedings of the interaction
design and children conference (pp. 23–35).
98 7 AI Literacy Education in Secondary Schools

Zhang, J., & Du, H. (2008, December). The PBL’s application research on prolog language’s
instruction. In 2008 international workshop on education technology and training & 2008
international workshop on geoscience and remote sensing (vol. 1, pp. 112–114). IEEE.
Zimmermann-Niefield, A., Turner, M., Murphy, B., Kane, S. K., & Shapiro, R. B. (2019, June).
Youth learning machine learning through building models of athletic moves. In Proceedings
of the 18th ACM international conference on interaction design and children (pp. 121–132).
Chapter 8
AI Literacy Education for Nonengineering
Undergraduates

AI literacy is in high demand across industries. Thus, being literate in or learning AI

should no longer be viewed as a specialized field under engineering but an ability
that penetrates all disciplines (Johri, 2020). An analogy to extend this argument is
by viewing traditional literacy. We would expect not only linguistics students to be
competent in literacy, which is the proficiency to read and write, but also an appro-
priate level of literacy across any majors. Similarly, students at all levels and disci-
plines should develop AI literacy to stay competent in today’s world.
In previous chapters, reviews had presented the educational levels across several
years as a single unit (i.e., kindergarten, primary school, and secondary school).
However, when it comes to postsecondary level, this unification becomes inappli-
cable. Students acquire knowledge and develop competencies toward their profes-
sion. The curricula focus on specialized subject domains instead of foundational
and whole-person development as in K–12. Though one may argue about the shift
in the twenty-first century educational paradigm where higher education is being
remodeled to a more dynamic and interdisciplinary state, one of the core objectives
of universities remains to be helping students progress into employment, which
primarily is still driven by their major studies. And hence, it is more logical to con-
tinue the review of AI literacy education in postsecondary level by differentiating
the students’ major of study.
But where should we draw the line? What we would like to investigate is whether,
and to what extent, there is an emergence of AI literacy education even though the
majors “traditionally” do not involve AI. One possibility is to separate AI majors
from all other programs. However, the coverage of AI literacy may spread beyond
AI-specific courses. For example, a command in programming contributes to partial
steps to understand, apply, and create AI, but the introduction to programming, for
instance, is also a typical subject for computer science or even engineering students.
We also found examples where universities have proposed to make AI education
mandatory regardless of the students’ discipline (de Freitas & Weingart, 2021).
Therefore, there is no definite line to draw in this respect. As this chapter will

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 99

D. T. K. Ng et al., AI Literacy in K-16 Classrooms,
https://doi.org/10.1007/978-3-031-18880-0_8
100 8 AI Literacy Education for Nonengineering Undergraduates

explain, a general term, nonengineering undergraduates (NEU in short) is used. The

review uses various approaches in an attempt to capture more evidence on how
educators design learning experiences for NEUs to develop AI literacy.
Many universities are not new to offering programs on AI, but most of them are
exclusive for computer science (CS) majors, CS-related engineering students or
offered to postgraduates (Mishra & Siy, 2020). The literature signifies the trends
and importance of AI education at all levels, while the AI curriculum development
for NEUs is under-explored. This chapter aims to shed light on the instructional
design, pedagogies, learning artifacts, and the topic areas on AI literacy mentioned
in discussion papers and empirical studies related to NEUs. Most research on this
similar topic focuses on K–12, and thus a research gap lies in the current state for
NEUs. This study also represents an extension to Ng et al.’ (2021) work on AI lit-
eracy at different education levels. In their study, only four articles were identified
in the category of higher education. This study aims to broaden the search by includ-
ing more databases and to be more specific to literature that focuses on NEUs. This
study aims to answer the following questions:
RQ1: What pedagogies have been used in teaching AI to NEUs?
RQ2: What aspects of AI literacy have NEUs developed?
RQ3: What learning artifacts have been used in teaching AI to NEUs?

8.1 Methodology

8.1.1 Data Collection

Both electronic databases search and snowball sampling were undertaken in search
for publications within our scope. For database search, relevant peer-reviewed arti-
cles and conference papers were identified from five world’s trusted databases,
namely, Web of Science, Scopus, ERICs, ProQuest, and IEEE. The selected data-
bases ensure the inclusion of evidence-based quality research (Mongeon & Paul-­
Hus, 2016). There were two sets of search phrases. The first set included “AI” OR
“artificial intelligence” OR “artificial intelligence literacy” OR “deep learning” OR
“machine learning” OR “neural network*” OR “intelligent agent”; the other set
included “non-computer science majors” OR “nonmajors” OR “nonengineering.”
Matched results in either the title, abstract, or keywords were downloaded. As of 16
April 2022, 507 records were identified. After the process of initial screening (462),
removing duplicate studies and those unable to access full text (18), and excluding
studies that were not targeted to NEUs or not related to AI education (15), 12 arti-
cles were included for detailed review.
For snowball sampling, it is a non-probability sampling technique that uses the
initial set of samples to introduce, refer, and generate further samples (Parker et al.,
2019). Based on the references from the included articles, eight more articles were
8.1 Methodology 101

included by chain referencing. As a result, 20 articles were reviewed. The search

protocol is summarized using the PRISMA diagram presented in Fig. 8.1.
RQ2 concerns a mapping of what and how AI was taught as described in the
articles into aspects of AI literacy. As illustrated in the UNESCO report, AI literacy
comprises both data literacy and algorithm literacy, and it serves as an orientation of
“knowledge, understanding, skills, and value of AI” (UNESCO, 2022, p. 11). Three
main categories of AI curriculum content were proposed which were further divided
into nine topic areas. We adopted the AI curriculum areas suggested in the UNESCO
report and developed a coding system according to the topic areas as illustrated in
Table 8.1.

507 records identified through database searching:

Identification

Web of Science (29), Scopus (62), ERICs (27),

Proquest (370), IEEE (19)

462 records excluded by title and abstract

Screening

45 records screened

18 articles excluded:
13 – duplicated studies
5 – cannot access full-text
Eligibility

27 full-text articles assessed for eligibility

15 articles excluded:
7 – participants not related to non-
engineering undergraduates
8 – focus not related to AI education
Included

20 studies included for review

(8 articles added from snowball method)

Fig. 8.1 PRISMA diagram of included articles

102 8 AI Literacy Education for Nonengineering Undergraduates

Table 8.1 Coding system of AI literacy topic area

Category Topic area Code
AI foundations Algorithms and programming A1
Data literacy A2
Contextual problem-solving A3
Ethics and social impact The ethics of AI B1
The social or societal implications of AI B2
Applications of AI to domains other than ICT B3
Understanding, using, and developing AI Understanding and using AI techniques C1
Understanding and using AI technologies C2
Developing AI technologies C3
Adapted from UNESCO (2021)

8.1.2 Data Analysis

We conducted document analysis to review and evaluate its content systematically

(Bowen, 2009). The procedure in this study involved an iterative process of (1)
identifying relevant program information, (2) categorizing and coding the content,
(3) interpreting the patterns of program description, and (4) sorting into themes. To
ensure coding reliability, six articles were randomly picked, blind-coded, and ana-
lyzed by two researchers. Agreements were reached for any discrepancies in the
understanding of the coding systems before moving forward to thorough analysis.
Table 8.2 shows the characteristics of the reviewed articles by their published
year, country, type of article, and research methods. As mentioned above, there is an
increasing attention to AI education in recent years, and the growing number of
publications since 2016 is in line with this notion. Moreover, the jump is even more
notable in the year 2021. This is a positive indication that more scholars are putting
effort in discussing AI for nonengineering majors. For the origins of the publication,
the majority of the articles published were from the United States. In terms of
research methods, there is an even spread of qualitative (5), quantitative (6), and
mixed-method (2) papers. Conceptual papers (6) refer to articles that the scholars
tend to share their insights toward the area and their experiences and good practices
in their teaching.

8.2 Results and Discussion

RQ1: What pedagogies were used in teaching AI to NEUs?

The pedagogies described in the reviewed articles are summarized in Table 8.3.
Most studies used project-based or hands-on experiential learning in their courses to
teach AI to NEUs. Most projects involved learning artifacts which aid problems or
advancements of students’ subject domain by introducing possibilities with AI. For
8.2 Results and Discussion 103

Table 8.2 Frequency (N, %) of the characteristics of the reviewed articles

Variables Categories N Percent (%)
Year <2010 3 15
2016 2 10
2017 2 10
2018 1 5
2019 3 15
2020 1 5
2021 6 30
2022 2 10
Countries Austria 1 5
Canada 2 10
China 2 10
Hong Kong 1 5
Portugal 1 5
South Korea 2 10
Taiwan 2 10
United States 9 45
Types of articles Journal articles 9 45
Conference papers 11 55
Research methods Qualitative 5 25
Quantitative 6 30
Mixed research 2 10
Conceptual papers 6 30
Not mentioned 1 5

example, Armstrong (2010) introduced two autonomous robots in his course. The
instructors removed the robots’ outer shells to let students explore the mechanical
components and the controlling sensors. After students learned about the exterocep-
tive and proprioceptive sensors on both, they were tasked with programming tasks
to learn about intelligent systems and related topics. A more recent study by Lin
et al. (2021) introduced hands-on exercises to train an AI model that could identify
directions and use the model to control a motor-controlled car.
A few studies used games and competitions to motivate students’ interest to learn
AI. For example, Rattadilok et al. (2018) developed an intervention named In-Class
Gamified Machine Learning Environment (iGaME) based on a mobile game called
Clash of Clans. Results showed that students engaging in games can enable key
machine learning concepts. The gamified platform also made customized learning
experiences possible. Another study which also combined gamification and
competition-­based learning is by de Freitas and Weingart (2021). NEUs in their
course programmed a rocket landing simulator using Python. Students competed by
applying genetic algorithms to train the rocket to land in the safest and most effi-
cient manner. Their findings indicated that NEUs can understand AI concepts
Table 8.3 Descriptive information of the included studies
104

Author/ Type of Research Learning Key findings and AI

year Article title study design Participant Pedagogy artifacts notions literacy
Armstrong Robotics and Empirical Not mentioned Not mentioned Experiential Robotics, Leveraging the appeal A1, A3,
(2010) intelligent systems learning, ActionScript, of robotics and C2
for social and project-based Python artificial intelligence to
behavioral science improve recruitment to
undergraduates the major and to
expose students in
other disciplines to
computational thinking
8

Au-Yong-­ What can we Discussion Questionnaire 110 mixed N/A N/A AI may lead to B1, B2,
Oliveira expect from the and interviews engineering and negative impact on B3
et al. future? The impact nonengineering work and loss of
(2020) of artificial students control. Humans will
intelligence on be ahead of
society AI. Engineering
students are more
afraid of AI than
nonengineering
students
AI Literacy Education for Nonengineering Undergraduates
Danyluk Using robotics to Discussion N/A N/A Experiential Robotics, The lab provides a A1, C2
(2004) motivate learning learning, LEGO, setting in which
in an AI course for project-based Handyboard, C, students can build
non-majors light sensors confidence; the fun of
working with robots
motivates students to
be more engaged in
lecture; and students
learn some
8.2 Results and Discussion

fundamental
programming concepts
in addition to material
about AI
de Freitas I’m Going to Learn Empirical Assignment 174 freshmen Competition-­ Python, Non-computer A1, A3,
and What?!? Teaching performance based teachable scientists can B1, B2,
Weingart Artificial and machine comprehend AI/ML C1, C2
(2021) Intelligence to questionnaire concepts without being
Freshmen in an overwhelmed by the
Introductory subject material
Computer Science
Course
Eaton et al. Blue sky ideas in Discussion N/A N/A Peer learning N/A Using seminar-style A3
(2017) artificial approach or project-­
intelligence based approach to
education from the connect subject
EAAI 2017 new students with AI
and future AI students
educator program
(continued)
105
Table 8.3 (continued)
106

Author/ Type of Research Learning Key findings and AI

year Article title study design Participant Pedagogy artifacts notions literacy
Fox (2007) Finding the “Right” Empirical Not mentioned Four to six Competition-­ LEGO, There are relatively A1, C2
Robot Competition: teams from three based, gamified Handyboard fewer competitions
Targeting to four schools learning suitable for NEUs to
Non-Engineering introduce them to AI
Undergraduates or robotics techniques
Gil (2016) Teaching Big Data Discussion N/A N/A Experiential WINGS Students will be able A1, A2,
Analytics Skills learning intelligent to get firsthand C2
with Intelligent workflow system experiences with a
8

Workflow Systems breadth of big data

topics, including
multistep data analytic
and statistical
methods, software
reuse and composition,
parallel distributed
programming,
high-end computing
Hu and Study on Teaching Discussion N/A N/A N/A N/A AI experiences a low A1, A2,
Wang Reform of Artificial development process C1
(2021) Intelligence and fails to attract
Education in attention. The
non-Computer popularization of AI
Major education in non-­
computer major
involves the
adjustment of the
whole information
general education
system
AI Literacy Education for Nonengineering Undergraduates
Kandlhofer Artificial Empirical Observations N/A Constructionism, Golog, YAGI, All students A3, C2
et al. intelligence and and interviews demonstrative Unity successfully completed
(2016) computer science hands-on activities the course and gained
in education: From deeper understanding
kindergarten to of AI topics and a
university higher abstraction
level
Kim and Development of an Empirical Questionnaires 88 AR-based AI Unity, Based on TAM and A1, A3,
Shim AR-Based AI nonengineering education for AIEduAR, QR IMMS results, students C1, C2
8.2 Results and Discussion

(2022) Education App for undergraduates visually hands-on code are highly likely to
Non-Majors experiences accept and learn AR
technology in AI
education
Kong et al. Evaluation of an Empirical Questionnaire 120 mixed Flipped classroom Not mentioned Students without prior A1, A3,
(2021) artificial and interviews nonmajors knowledge of B1, B2,
intelligence literacy programming could B3, C1
course for understand AI
university students concepts of machine
with diverse study learning, supervised
backgrounds learning, regression,
classification,
unsupervised learning,
and clustering
(continued)
107
Table 8.3 (continued)
108

Author/ Type of Research Learning Key findings and AI

year Article title study design Participant Pedagogy artifacts notions literacy
Lee and Development of an Empirical Questionnaire 11 non-­ Experiential Python, Perception and A1, A2,
Cho (2021) Artificial and AI computer majors learning teachable understanding of AI B2, B3,
Intelligence comprehension machine were correlated with C2
Education Model test education satisfaction
of Classification
Techniques for
Non-computer
Majors
8

Li (2019) Experience Report: Empirical Qualitative mixed of CS and Explorable Course Catalog The presentation of CS A1, C1,
Explorable Web student nontechnical explanations, Prerequisite content to non-CS C2
Apps to Teach AI feedback students (# not interactive Extraction, students have benefited
to NonMajors mentioned) illustrations, Pattern-­ from use of interactive
self-guided Matching web apps that allow
discoveries Chatbot, students to explore CS
Bayesian concepts without
Network writing code
Calculator
Lin et al. STEM-based Empirical Questionnaires 328 Project-based Custom Vision AI literacy is B1, B2,
(2021) Artificial nonengineering (deep learning), correlated to their C2
Intelligence freshmen Raspberry Pi awareness of AI
Learning in (robot car) ethical issues.
General Education STEM-based AI
for curriculum increased
NonEngineering the awareness of AI
Undergraduate ethical issues among
Students low-AI-literate
learners
AI Literacy Education for Nonengineering Undergraduates
Rattadilok Teaching Students Empirical Assignment Not mentioned Gamified learning, iGaME Students engaging in C1, C2
et al. About Machine performance universal design games can enable key
(2018) Learning Through for learning, machine learning
a Gamified personalized concepts. Games can
Approach learning customize learning
experiences
Shih et al. Learning Ethics in Empirical Questionnaires 328 Situated learning Custom vision A strong link between B1, B2,
(2021) AI—Teaching nonengineering (deep learning), AI understanding and C2
Non-Engineering freshmen Raspberry Pi attitudes and AI ethics
8.2 Results and Discussion

Undergraduates (robot car)

through Situated
Learning
Sulmont What Is Hard Empirical Interview 10 instructors N/A Scikit-learn, Learning goals, such A1, A2,
et al. About Teaching Microsoft Azure as making design A3, C1
(2019a) Machine Learning Machine decisions and
to Non-Majors? Learning comparing/contrasting
Insights from models, are difficult to
Classifying teach. Learning goals
Instructors’ are consistent with the
Learning Goals levels of the Structure
of Observed Learning
Outcomes (SOLO)
taxonomy
(continued)
109
Table 8.3 (continued)
110

Author/ Type of Research Learning Key findings and AI

year Article title study design Participant Pedagogy artifacts notions literacy
Sulmont Can You Teach Me Empirical Structured 10 instructors N/A N/A Student A1, A2,
et al. To Machine Learn? Interview preconceptions include A3, C1
(2019b) ideas that ML is
important, but also not
accessible. […] it is
possible to teach ML
to those with little to
no math/CS
8

background. […]
students having
difficulty appreciating
the human decision-­
making aspects of ML,
and overestimating the
power of ML to solve
real-world problems
Way et al. Machine learning Discussion N/A N/A Problem-based WEKA (Text 12 modules are A1, A2,
(2017) modules for all Classification) designed for use in a A3, C1
disciplines variety of educational
scenarios. Students
experience solid
learning that is
retained and
nontechnical students
generally gain more
than technical students
AI Literacy Education for Nonengineering Undergraduates
Yang et al. Stakeholders’ Review Document 62 publications N/A N/A A significant impact B1, B2,
(2021) perspectives on the analysis on radiology but B3
future of artificial unlikely replacing
intelligence in radiologists. Non-­
radiology: a computer scientists
scoping review have limited AI
knowledge.
Collaboration between
radiologists and AI
8.2 Results and Discussion

specialists is needed
111
112 8 AI Literacy Education for Nonengineering Undergraduates

without a strong background in computer science and that students acknowledged

the importance and value of learning AI through their course.
Within the popular approach of using project-based learning to teach AI, we
speculate a shift in the type of projects introduced in the courses. While the projects
before 2010 tended to link AI with robots, recent studies tended to make use of
practical applications in the specific disciplines. This phenomenon may be explained
by what drove educators to teach AI to NEUs. In early to mid-2000s, there was a
decline in enrollments in computer science departments, and hence the objective
was only to use AI topics to attract students’ interest (e.g., Danyluk, 2004; Fox,
2007). At present, applications of AI are penetrated in most disciplines, and there-
fore educators are embedding AI literacy into part of students’ subject curriculum
(e.g., Kim & Shim, 2022; Kong et al., 2021).
RQ2: What aspects of AI literacy had NEUs developed?
The mapping of AI topic areas is shown in the rightmost column of Table 8.3.
Some articles (e.g., Au-Yong-Oliveira et al., 2020; Eaton et al., 2018) did not center
discussions on the types of AI content but other issues such as instructional strate-
gies and learners’ perceptions. The following discussion is guided based on three
categories: AI foundations, ethics and social impact, and understanding, using and
developing AI (UNESCO, 2021).
For AI foundations, most articles (12) covered topics related to algorithms and
programming (A1), which represent teaching of basic technical engagement with
AI. However, only five articles discussed the instructions for data literacy (A2) and
seven used project-based learning as their key pedagogy to solve problems in
disciplinary-­specific contexts (A3). Most studies (e.g., Kong et al., 2021) designed
instructions such that prior knowledge of computer science or programming was not
necessary. Hence, it is reasonable that instructors choose the minimal level of AI
foundations needed for their students to experience AI.
For ethics and social impact, we see an interesting divide by the year of publica-
tion that is similar to our argument in RQ1. Most recently published articles (e.g.,
Au-Yong-Oliveira et al., 2020; Kong et al., 2021) would address the balance between
the convenience that AI brings to our daily lives as well as the potential threats and
ethical challenges if AI is not implemented properly. Studies earlier than 2010
hardly mentioned any concerns for ethical considerations. The attention to this
aspect would only seem to increase in proportion to the rapid development of AI
applications across industries and in our daily lives.
For understanding, using, and developing AI, most studies (12) used AI tech-
nologies (C2) as a service to address the needs in the field of NEUs. As students
experienced these technologies, some instructors would explain the AI techniques
(C1) behind such that students could try to practice applying the service or replicat-
ing the results on their own. Finally, none of the studies reported to teach their stu-
dents to develop AI technologies (C3). This is not surprising as the target audience
is NEUs and the learning objectives were to understand basic concepts or to apply
AI in their disciplines, rather than reaching a higher level of mastery.
8.2 Results and Discussion 113

RQ3: What learning artifacts were used in teaching AI to NEUs?

In order to teach AI to those with little background knowledge in mathematics
and programming, using tools to visualize what AI or ML can do is a popular and
practical tactic (Sulmont et al., 2019b). Similar to previous chapters, instructors
choose the types of learning artifacts which match students’ learning goals and
available resources. These artifacts can be categorized into hardware-focused,
software-­focused, and intelligent agents. The definitions and the categorization of
these learning artifacts are shown in Table 8.4.
For hardware-focused artifacts, using robotics is a straightforward representation
of what AI can achieve. Students can build robots to achieve certain mechanical
tasks such as pathfinding by sensors or performing object sorting by image recogni-
tion. For example, Fox (2007) used LEGO-based robots as the tool to attract NEUs
to participate in an AI-related competition. A couple of limitations were mentioned
as NEUs do not have the necessary knowhow to understand and build such sophis-
ticated machines. So, there was a continuous struggle between not setting expecta-
tions too high and providing sufficient autonomy for students to explore the
possibilities. One observation from this review is that few recent articles mentioned
using hardware-focused artifacts as their teaching tool. A possible explanation is
because the focus of AI has shifted from highly associated with humanoid robots in
the early 2000s to the wide range of app-based applications.
For software-focused artifacts, some articles reported using Python as the AI
programming platform. de Freitas and Weingart (2021) taught students to use
Python to write a program for a rocket landing simulator. In the assignment, stu-
dents needed to use AI to generate random locations for the rocket, to track perfor-
mance indicators such as fuel consumption and velocity of the rocket, and to design
controls for the rocket’s thrusters for successful landing. In another example, Way
et al. (2017) proposed several teaching ideas using WEKA Text Classification for
NEUs: art majors were engaged in a classification and clustering module; French

Table 8.4 Learning artifacts mentioned in the reviewed articles

Definition Learning artifacts examples Sample studies
Hardware-­ Use physical artifacts to Robotics, LEGO, Handyboard, Armstrong
focused learn AI such as robotics, light sensors, Raspberry Pi (2010), Danyluk
artifacts sensors and Arduino (2004), Fox
devices (2007)
Software-­ Use digital artifacts to Python, ActionScript, C, Custom de Freitas and
focused learn AI such as block/ Vision (deep learning), WEKA Weingart (2021),
artifacts syntax-based (Text Classification), C++ Lee and Cho
programming and (2021), Lin et al.
simulation (2021)
Intelligent Use intelligent agents such Teachable Machine, Golog, YAGI, Rattadilok et al.
agents as expert systems, Unity, AIEduAR, Course Catalog (2018), Sulmont
machine learning trainers, Prerequisite Extraction, Pattern-­ et al. (2019a),
chatbots to build their Matching Chatbot, Bayesian Kandlhofer et al.
custom machine learning Network Calculator, iGaME, (2016)
models without coding Scikit-learn, Microsoft Azure
Machine Learning
114 8 AI Literacy Education for Nonengineering Undergraduates

literature students could learn about AI Machine Translation embedded in their

writing and stylistics course.
The key difference between software-focused artifacts and intelligent agents is
that the former involves learners in the analytical, mathematical, or algorithmic pro-
cesses of AI. Leaners can use the software artifacts to understand, apply, or create
the “thinking” processes of AI, whereas the latter focuses only on using a packaged
platform or technology for applications without coding. Rattadilok et al. (2018) had
successfully engaged students in a gamified AI learning experience which enhanced
students’ machine learning concepts. They developed an intelligent agent named
iGaME (In-Class Gamified Machine learning Environment) for teaching machine
learning algorithms with the adaptation of an online strategy mobile game called
Clash of Clans. They compared teaching with iGaME to traditional machine learn-
ing teaching and suggested the use of intelligent agents could motivate nontechnical
students to study difficult technologically driven topics such as machine learn-
ing and AI.

8.3 Conclusion

Artificial intelligence (AI) is widely applied in almost any industry, such as manufac-
turing, economy, communications, transportation, information, finance, education,
medical care, etc. (Pan, 2018; Sestino & De Mauro, 2022). Researchers and educators
are in search of ways to define and upskill students’ AI literacy. However, few studies
have visited how NEUs can be introduced to and be literate in AI. In the form of a
review, this chapter consolidated the available publications that address this gap.
The review used database search and snowball sampling and resulted in 20 arti-
cles that targeted their discussions on AI for NEUs. Project-based learning and
hands-on experiences were found to be the most commonly adopted approaches by
instructors. The content described in the articles was mapped into nine topic areas
adapted from the UNESCO report (UNESCO, 2021). The learning artifacts that
aided the teaching of AI were categorized into hardware, software, and intelligent
agents focused.
Notwithstanding, there are a few limitations of this study that should be noted.
First, it was difficult to capture all studies on non-computer majors as there was not
a unified terminology for this group of students. Some studies did not include any
term such as non-computer scientists or nontechnical students in their studies. For
example, Yang et al. (2021) elicited perspectives from radiology students which
were also included in this review. Second, there are a noticeably increasing number
of online, self-paced AI programs offered by universities for diverse audiences. It is
highly possible that these programs will gain popularity among NEUs. Future study
is needed to elicit the impact of these programs on learners’ AI literacy. The study
benefits future educators who intend to design AI learning programs that target
NEUs. This review also contributes to the broader scope of AI literacy development
of students at all educational levels.
References 115

References

Armstrong, T. (2010, June). Robotics and intelligent systems for social and behavioral science
undergraduates. In Proceedings of the fifteenth annual conference on Innovation and technol-
ogy in computer science education (pp. 194–198).
Au-Yong-Oliveira, M., Lopes, C., Soares, F., Pinheiro, G., & Guimarães, P. (2020, June). What can
we expect from the future? The impact of artificial intelligence on society. In 2020 15th Iberian
Conference on Information Systems and Technologies (CISTI) (pp. 1–6). IEEE.
Bowen, G. A. (2009). Document analysis as a qualitative research method. Qualitative Research
Journal, 9, 27.
Danyluk, A. (2004, March). Using robotics to motivate learning in an AI course for non-majors. In
AAAI spring symposium (pp. 22–24).
de Freitas, A. A., & Weingart, T. B. (2021, March). I’m going to learn what? Teaching artificial
intelligence to freshmen in an introductory computer science course. In Proceedings of the
52nd ACM technical symposium on computer science education (pp. 198–204).
Eaton, E., Koenig, S., Schulz, C., Maurelli, F., Lee, J., Eckroth, J., et al. (2018). Blue sky ideas in
artificial intelligence education from the EAAI 2017 new and future AI educator program. AI
Matters, 3(4), 23–31.
Fox, S. E. (2007). Finding the “Right” Robot competition: Targeting non-engineering undergradu-
ates. In AAAI spring symposium: Semantic scientific knowledge integration (pp. 49–52).
Gil, Y. (2016, March). Teaching big data analytics skills with intelligent workflow systems. In
Proceedings of the AAAI conference on artificial intelligence (vol. 30, no. 1).
Hu, Q., & Wang, K. (2021, August). Study on teaching reform of artificial intelligence education
in non-computer major. In The Sixth international conference on information management and
technology (pp. 1–4).
Johri, A. (2020). Artificial intelligence and engineering education. Journal of Engineering
Education, 3, 358–361.
Kandlhofer, M., Steinbauer, G., Hirschmugl-Gaisch, S., & Huber, P. (2016, October). Artificial
intelligence and computer science in education: From kindergarten to university. In 2016 IEEE
Frontiers in Education Conference (FIE) (pp. 1–9). IEEE.
Kim, J., & Shim, J. (2022). Development of an AR-based AI education app for non-majors. IEEE
Access, 10, 14149–14156.
Kong, S. C., Cheung, W. M. Y., & Zhang, G. (2021). Evaluation of an artificial intelligence literacy
course for university students with diverse study backgrounds. Computers and Education:
Artificial Intelligence, 2, 100026.
Lee, Y., & Cho, J. (2021). Development of an artificial intelligence education model of classi-
fication techniques for non-computer majors. JOIV: International Journal on Informatics
Visualization, 5(2), 113–119.
Li, J. (2019). Experience report: Explorable web apps to teach AI to non-majors. Journal of
Computing Sciences in Colleges, 34(4), 128–133.
Lin, C. H., Yu, C. C., Shih, P. K., & Wu, L. Y. (2021). STEM based Artificial Intelligence Learning
in General Education for Non-Engineering Undergraduate Students. Educational Technology
& Society, 24(3), 224–237.
Mishra, A., & Siy, H. (2020, October). An interdisciplinary approach for teaching artificial intel-
ligence to computer science students. In Proceedings of the 21st annual conference on informa-
tion technology education (pp. 344–344).
Mongeon, P., & Paul-Hus, A. (2016). The journal coverage of Web of Science and Scopus: a com-
parative analysis. Scientometrics, 106(1), 213–228.
Ng, D. T. K., Leung, J. K. L., Chu, S. K. W., & Qiao, M. S. (2021). Conceptualizing AI literacy: An
exploratory review. Computers and Education: Artificial Intelligence, 2, 100041.
Pan, Y. H. (2018). 2018 special issue on artificial intelligence 2.0: Theories and applications.
Frontiers of Information Technology & Electronic Engineering, 19(1), 1–2.
116 8 AI Literacy Education for Nonengineering Undergraduates

Parker, C., Scott, S., & Geddes, A. (2019). Snowball sampling. In SAGE research methods
foundations.
Rattadilok, P., Roadknight, C., & Li, L. (2018, December). Teaching students about machine
learning through a gamified approach. In 2018 IEEE international conference on Teaching,
Assessment, and Learning for Engineering (TALE) (pp. 1011–1015). IEEE.
Sestino, A., & De Mauro, A. (2022). Leveraging artificial intelligence in business: Implications,
applications and methods. Technology Analysis & Strategic Management, 34(1), 16–29.
Shih, P. K., Lin, C. H., Wu, L. Y. Y., & Yu, C. C. (2021). Learning ethics in AI – Teaching non-­
engineering undergraduates through situated learning. Sustainability, 13(7), 3718.
Sulmont, E., Patitsas, E., & Cooperstock, J. R. (2019a). What is hard about teaching machine learn-
ing to non-majors? Insights from classifying instructors’ learning goals. ACM Transactions on
Computing Education (TOCE), 19(4), 1–16.
Sulmont, E., Patitsas, E., & Cooperstock, J. R. (2019b). Can you teach me to machine learn?
In Proceedings of the 50th ACM technical symposium on computer science education
(pp. 948–954).
UNESCO. (2021). Survey for mapping of AI curricula. Unpublished (Submitted to UNESCO).
UNESCO. (2022). K-12 AI curricula: A mapping of government-endorsed AI cur-
ricula. Retrieved May 4, 2022, from https://www.unesco.org/en/articles/
unesco-­releases-­report-­mapping-­k-­12-­artificial-­intelligence-­curricula
Way, T., Papalaskari, M. A., Cassel, L., Matuszek, P., Weiss, C., & Tella, Y. P. (2017, June).
Machine learning modules for all disciplines. In Proceedings of the 2017 ACM conference on
innovation and technology in computer science education (pp. 84–85).
Yang, L., Ene, I. C., Arabi Belaghi, R., Koff, D., Stein, N., & Santaguida, P. L. (2021). Stakeholders’
perspectives on the future of artificial intelligence in radiology: A scoping review. European
Radiology, 1–19.
 Part III
AI Literacy for Instructional Designers
Chapter 9
AI Literacy on Human-Centered
Considerations

Part II of this book gave us an overview of AI literacy across educational levels.

Several models of AI literacy education, in particular Bloom’s taxonomy, TPACK
model, as well as the P21’s Framework for the 21st Century Learning (2009) that
comprises key digital competencies that inform instructional designers what knowl-
edge, skills, and attitudes students should equip with.
Although these frames suggest useful instructional resources on a research level,
there is a lack of discussion to inform developers within the industry and teachers at
large about what particular digital competencies they need to facilitate their product
or service development and teaching in classrooms, respectively. Artificial intelli-
gence offers unprecedented opportunities for EdTech companies to develop
AI-based products for K–12 and higher education institutions. However, EdTech
startups and companies may have a broad sense of what the market desires yet not
fully aware of teachers’ and students’ holistic needs when learning AI. As a result,
many EdTech on AI education arises, but few can sustain and oftentimes only
become a gimmick. On the other hand, educational institutions have started to
design professional training programs for the teacher workforce to integrate AI
applications throughout the student life cycle. For example, the MIT Schwarzman
College of Computing considered humanistic perspectives as the keys for under-
standing the implications of computation for knowledge and representation, and
connected AI to the critical process of how knowledge works in culture (MIT, 2019).
Part III (Chaps. 9 and 10) aims to explore how instructional designers (i.e., AI
developers and teachers) prepare themselves to become ready to design develop-
mentally appropriate tools, platforms, services, and curricula to empower students
with AI literacy skills.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 119
D. T. K. Ng et al., AI Literacy in K-16 Classrooms,
https://doi.org/10.1007/978-3-031-18880-0_9
120 9 AI Literacy on Human-Centered Considerations

9.1 Overview

The first half of this part, Chap. 9, focuses more on the human-centered design and
product developmental aspects of AI literacy education. Of which much of the
emphasis is placed on the learning of human-centeredness of AI, i.e., human-cen-
tered AI (HCAI). To achieve this goal, both “what” content should be covered and
“how” it should be taught are equally important. By foregrounding human-cen-
teredness in AI, i.e., Sections 9.2 and 9.3, the content created with this basis can
establish a future-proof foundation for students to appropriately embrace the
impacts of AI. Then in Sect. 9.4, we focus on the pedagogical approach in order to
scaffold students’ literacies. AI, which was once an abstract, highly skilled subject,
is now brought down to K–16 education. Careful implementation of instructional
design is necessary for such learning experiences to be effective.
Based on the notions of HCAI, some design principles that place users and learn-
ers at the center are proposed to aid researchers and developers when designing such
learning tools. Furthermore, the calls for HCAI that meet students’ cognitive and
social needs and the educational kits with considerations of ethical concerns behind
have been rapidly emerging. The chapter will provide insights on the key issues to
be taken into account when designing learning artifacts to help reinforce students’
acquisition of AI literacy skills and facilitate their learning.

9.2 Needs of HCAI in Educational Fields

Nowadays, service and product providers have entered the educational sector to
offer intelligent learning solutions for educators and students. Although these
EdTech applications are innovation rich for the business models of providers and
users, few attention has been paid to the role of human values in developing AI
technology. Some scholars proposed the need to establish models and guidelines to
place human values at the center of AI design and development (Dignum, 2019).
The idea of human-centered design was first suggested to remind developers to put
humans at the center of AI development, rather than considering AI as a replace-
ment for human agency (Renz & Vladova, 2021; Xu, 2019). In the educational sec-
tor, some scholars started to propose the idea of HCAI to enhance teaching/learning
experience. For example, Yang (2021) pointed out that learning technology must be
human-centered because it involves teaching and interacting with people. Rather
than focusing on students’ performance, human feelings and outcomes should be a
major concern in designing smart learning environments. Several key elements of
human-centered considerations were identified in educational fields when develop-
ing appropriate tools and systems for students: human factors designs and values,
human intelligence, ethical and responsible design, as well as AI under human con-
trol and under human conditions.
9.3 Key Elements of Human-Centered Considerations 121

9.3 Key Elements of Human-Centered Considerations

9.3.1 Human Factor Designs and Values

The first framework of HCAI in educational fields is proposed by Xu (2019) who

suggested three ideas regarding how human factor designs and human values are
considered when designing and implementing AI learning technologies. First, these
learning technologies need to be ethically designed that developers should avoid
discrimination and ensure justice and fairness in their AI solutions (Xu, 2019). One
possibility to achieve this is by adopting a design thinking approach during the
development of AI applications and systems. Design thinking is an iterative process
that developers need to empathize with users (e.g., teachers, students), define prob-
lems from users’ perspectives, and create solutions by prototyping and testing (Dam
& Siang, 2018). This is consistent with the notion for learner-centered design that is
guided by a constructivist theory and advancements in computer-assisted technol-
ogy. Learner-centered design focuses on the quality of student learning to enhance
students’ needs, interests, and goals (Webber, 2012). This led to student-centered
learning that requires students to set their own learning goals and determine appro-
priate resources and activities that will facilitate them to meet those targets (Jonassen,
2000; Hannafin & Land, 2000). When learners pursue their own goals or enjoy their
learning, their learning activities would become meaningful to them. For example,
Dhungel et al. (2021) applied a human-centered approach to develop a MOOC
about AI for civil servants. To understand whether this course is suitable for the
target audiences, a learning analysis was conducted to examine their mindsets,
knowledge gained, and attitudes to develop a set of learning units that ground the
abstract AI topics in concrete case scenarios taken from the public sector. Civil ser-
vants need to have a practical understanding of AI including knowing the potentials,
challenges, and limitations of AI, as well as when and how to use it ethically. Yang
et al. (2021) proposed the idea of “precision education” that educators need to adapt
to changes due to learning diversity in four steps: diagnosis, prediction, treatment,
and prevention. Most tools and platforms have the problems with a one-size-fits-all
approach because learners have different educational backgrounds. Educators
should change their pedagogical approaches and apply teaching strategies and
methods wisely to enhance students’ learning outcomes.

9.3.2 Reflect Human Intelligence

Learning technologies should reflect human intelligence and consider human char-
acteristics and needs in the algorithm and system design (Xu, 2019). For example,
Pérez-Ortiz et al. (2021) applied a personalized learning companion using AI algo-
rithms and software tools with the goal of making it intuitive and user-friendly, as
well as making the AI transparent and explainable to the teachers and students. The
122 9 AI Literacy on Human-Centered Considerations

tools combine AI and HCI to create an integrated suite of tools for teachers and
learners to cognitively support them to work with the AI agents. In this example, the
AI agents could make learning feedback and recommendations to complete learning
tasks and provide a content flow bar tool for students to view the learning progress
of what topics they have covered in the platform.
In addition to designing student-centered educational technologies, educators
need to know the working principles behind an AI learning system such as how the
recommendation systems suggest the learners to study or practice a specific content
or how the automated scoring derives individualized comments and feedback.
Otherwise, teachers and students would become passive and nonautonomous with-
out gaining a transparent and comprehensive understanding of their inner workings.
To open this black box of learning, Luckin (2017, p. 1) proposed that AI learning
systems need information about: “(1) the learning curriculum, subject area and
activities that each student is completing; (2) the details of the steps each student
takes as they complete these activities; and (3) what counts a success within each of
these activities and within each of the steps towards the completion of each activ-
ity.” Another study Long and Magerko (2020) proposed the need to promote trans-
parency of the AI design such as reducing black-boxed functionality, sharing creator
intentions, and funding/data sources. In other words, the design of AI should reflect
teacher and student intelligence with explainable reasoning and comprehensive
understanding of what and why they receive the responses according to the AI
agents’ suggestions.
It is also important to note that there is unique human intelligence, such as cre-
ativity, socialization, design insights, and aspirations, which is hard, if not impos-
sible, for AI to take over (e.g., Chen et al., 2022; Miller, 2019). Jain et al. (2021)
builds on Shneiderman’s (2020a, b) advocacy for HCAI that advances “fundamental
human aspirations,” such as expressing people’s creative potentials, forming social
connections, and promoting equity. Although human aspirations could be quantified
in some sense through advancing the automatic recognition (e.g., visual complexity,
colorfulness), teachers and students can use AI as tools to facilitate the representa-
tion and communication of abstract ideas and support their visual expression and
reflective ideation.

9.3.3 Ethical and Responsible Design

Learning technologies with ethical and responsible design such as explainability,

comprehensiveness, usefulness, and usability should be considered. Many educa-
tors show their worries that AI agents would replace their jobs and challenge the
effectiveness of using AI to teach in their classrooms due to a number of concerns
such as black-box machine learning, data bias (Rudin, 2019), trustworthiness, pri-
vacy (Reinhardt et al., 2022), and other ethical concerns (Toreini et al., 2020).
Explainable AI enables teachers to understand the algorithm and the working prin-
ciples behind. For example, learning analytics should provide reasons for students’
9.3 Key Elements of Human-Centered Considerations 123

grade prediction through AI-based assessments. The AI agents should offer trans-
parent and trustworthy feedback to teachers, students, and parents about how the
students learn, what type of support they need, and what is the progress they are
making toward their learning goals (Luckin, 2017). The machine learning process
and its working principles of AI learning technologies should be transparent and be
visualized and explainable for all stakeholders of learning. Furthermore, compre-
hensive AI could adapt to students’ needs and capabilities (e.g., knowledge level
and learners’ interest) to provide individualized feedback and make suitable teach-
ing decisions. Useful AI means that AI should satisfy users’ needs in a particular
scenario of their work, study, and life. An example in the educational context, learn-
ers’ behavioral patterns usage scenarios should be considered when modeling the
users’ learning needs and usage scenarios to make appropriate learners’ reports.
Usable AI indicates an effective HCI and user interface design that offer user experi-
ences that specify users’ requirements (Xu, 2019). Ipsita et al. (2022) designed a
user-friendly interface that considers learners’ behavioral modeling to offer graphi-
cal animations, display data for attributes of virtual reality objects, and manipulate
with data and trigger controls in a mechanical engineering classroom. Big data and
AI have a synergistic relationship that AI requires a massive scale of data to learn
and improve decision-making processes. Therefore, instructional designers need to
encourage students to investigate how data is generated and collected in their learn-
ing projects and the limitations and bias of the datasets (Long & Magerko, 2020).

9.3.4 AI Under Human Control and Under Human Conditions

Renz and Vladova (2021) further proposed the teaming ideas of education from two
theoretical perspectives (i.e., AI under human control and AI under human condi-
tions). The ideas are similar to Xu (2019) regarding the two perspectives. The for-
mer is subject to judgment based on the degree of human control over AI. In school
settings, AI should be controllable and understood by users (e.g., school administra-
tors, educators, students, parents) to support daily automation to improve educa-
tional administration and take control of their learning and teaching. The second
idea “AI in the human condition” refers to the design of AI algorithms and working
principles that should be explained, interpreted, and refined based on student learn-
ing needs, contexts, and phenomena.
Renz and Vladova (2021) also proposed the concept of “teaming” and suggested
the importance of collaborating between users and AI agents. AI is an intelligent
agent not only to perform administrative and learning tasks for users (teachers, par-
ents, and students) but also to work with users as teaching assistants, learning facili-
tators, and moderators to build positive learning communities. We could find the
coherence in other publications in our review that also raise the importance of vital
roles, responsibilities, and characteristics of a collaborative AI agent. For example,
Morrison (2021) followed the HCAI approach to develop AI learning companions
to facilitate children who were born blind to experience social agency and develop
124 9 AI Literacy on Human-Centered Considerations

the range of social attention skills to initiate and maintain their interactions and
explore concepts learned incidentally through vision by using alternative perceptual
modalities. Another study Pérez-Ortiz et al. (2021) designed a human-centered AI
platform to provide learners with a number of educational tools to interact with
open educational videos and tools to suit their pedagogical preferences of learners.
Yang et al. (2021) proposed the importance of smart learning analytics and assess-
ment that enables students to take control of their learning, know how they are
performing compared to their peers, and help teachers to identify gaps in students’
prerequisite knowledge and key skills.
With AI and HCAI knowledge, AI developers and educators could understand
the individualized characteristics and needs of targeted learners so as to design posi-
tive human-centered AI-empowered learning environments, services, and designs
that provide values to diversified learners, organizations, and society as a whole.
Building on the consolidated domains and definition of HCAI, here we put forward
a summary of five key domains of HCAI in education (see Table 9.1). Instructional
designers can consider these factors to help design their curriculum, teaching tools,
and platforms when implementing AI literacy education in their classrooms.

Table 9.1 A summary of HCAI in education

HCAI domains HCAI in education
Human factors  Instructional designers and teachers should consider the roles of AI when
design designing learning environments, tools, and technologies. A set of human
factors design such as explainability, comprehensiveness, and transparency
should be considered
Reflect human  AI reflects human intelligence (including teachers and students) which
intelligence cannot be bypassed by AI and considers student-centered characteristics and
needs (e.g., students’ technical background, age, cultural background) in the
AI-driven instructional design
Ethical and  AI should facilitate computational techniques that are both innovative and
responsible responsible. It needs to prioritize issues of fairness, accountability,
design transparency, and ethics as they relate to AI, ML, and NLP by drawing on
fields with a sociotechnical orientation
AI under human  AI should be controllable by users (e.g., administrators, teachers) to support
control daily administration and decision-making to facilitate their work
 A comprehensive reasoning of why the recipients (e.g., students, parents)
receive the treatments should be given
 Other ethical concerns such as privacy, security, and safety were taken into
account in the AI-driven learning systems
AI under human  AI algorithms and working principles of the learning technologies that
conditions should be explained, interpreted, and refined according to students/users’
needs and learning scenarios
 A feedback mechanism is necessary for developers to improve the systems
and student learning conditions based on learner individualized needs
9.4 Scaffolding Support 125

9.4 Scaffolding Support

While we have investigated the properties of a humanistic, ethical, and responsible

design for AI literacy education, K–16 education has consistently viewed scaffold-
ing as essential to support students to master AI knowledge and skills like other
subject disciplines. To effectively design instructions and products that help stu-
dents learn, knowing your end users is important. This section suggests four major
considerations that help developers and educators to design their activities, pedago-
gies, assessments, and teaching materials and customize them to meet student
needs: students’ backgrounds, interests, and abilities, as well as parental involvement.

9.4.1 Knowing Learners’ Backgrounds

Students come to the classroom with a diversity of background knowledge and

experiences which influence how they see the world around them and how they
believe it works. Barriers and inequities exist when learners lack the critical back-
ground knowledge to assimilate or apply new knowledge (Ives & Castillo-Montoya,
2020). To create a better learning experience, instructional designers should con-
sider students’ developmental backgrounds such as age, prior knowledge, gender,
and cultural background as well as their preconceptions when designing tools and
platforms for end users.
First, instructional designers need to consider students’ developmental back-
grounds. Therefore, most of the K–12 AI educational kits for children use Blockly
codes and simple tools that may hide and encapsulate advanced computer science
concepts and technical mechanisms of how exactly AI works. These inevitable
“black-boxes” are necessary to prevent cognitive overload and develop students’ AI
knowledge and skills, especially when students do not have the enough technical
background knowledge to understand how exactly AI works. For example,
ML2Scratch extensions use “train label” block to represent the complex ML mod-
els, and chatbots (e.g., Parami, Kore) use conversation flowcharts to represent natu-
ral language processing. These simplified algorithms and components are black-box
for learners. Some teachers may worry about the transparency of these products and
design and challenge that students are merely using technologies without under-
standing the advanced concepts behind. Although this practice may lead to issues of
transparency and misunderstanding, a lower barrier to entry in learning AI is also
important. A trade-off balance can be achieved by giving learners the option to
inspect and learn about the challenging parts later in higher educational levels or
explaining only a few concepts that connect to their learning experience.
Further, designers and educators need to consider design bias when it comes to
race, ethnicity and socioeconomic class, and gender orientation. According to
Harvard Business Review, there have been many incidents of AI adopting gender
and cultural bias from humans since most machines are not women and black
126 9 AI Literacy on Human-Centered Considerations

researchers contributing. For example, Amazon’s Alexa and Apple’s Siri associate
“men” with “doctors” and “women” with “nurses.” Google’s Photos app wrongly
labeled pictures of black people as “gorillas.” Developers need to make sure training
samples are diverse and from different backgrounds. Moreover, gender disparity
and culture diversity in computing have long been a global concern. To reduce the
learning gaps for girls and black children, additional efforts are necessary to address
the issues and early outreach to them. For example, after-school computer club and
activities, career sharing and mentoring from computer scientists, and culturally
responsive design (e.g., using arts, culture, and music as themes to learn AI) can
support students to create more interest at a younger age.
Finally, knowing students’ preconceptions is important. These days, technology
companies are exploiting the current AI craze by exaggerating the scope and capa-
bilities of AI in their products to stimulate their sales, according to a report from
Gartner (Patrizio, 2017). Attention should be paid to what students may have sensa-
tionalized preconceptions of AI from popular media and advertisements. Educators
need to remind students to think critically about what counts as AI and question the
intelligence, trustworthiness, and ethical aspects of so-called AI-powered solutions,
so that they could be less affected by the media and marketing influences.

9.4.2 Knowing Learners’ Interests and Motivation

Enhancing students’ interests and motivation to sustain their learning is important.

Instructional designers should anchor instruction by linking to students’ interests
and their prior knowledge when designing AI literacy tools and interventions (e.g.,
playing or making games, listening to music, drawing). For example, in the past,
learners needed to train and program their intelligent agents using syntax-based
programming. With technological advances, educational tools provide multiple
interesting ways to visualize the complex concepts and develop students’ AI under-
standings. Nowadays, many AI educational kits use playful experience and simula-
tions to explain how AI works. For example, AI for Ocean in Code.org, image
stylizer and Quick, Draw, and CodeCombat use interactive and engaging games for
students to have fun and explore AI in computer lessons (Ng et al., 2022). Moreover,
programming skills can be kept to a minimum so that students can use Blockly
codes and web-­based tools that help students visualize AI concepts and make their
machine learning models.
Researchers have tried to consider motivational factors when designing AI cur-
ricula and piloting them in school settings. For example, in primary schools, Lin
et al. (2021) proposed six motivational factors and strategies: intrinsic motivation,
career motivation, attention, relevance, confidence, and satisfaction among 420 pri-
mary students from the piloting schools as references for future development of AI
curricula and instruction. In secondary level, Chiu et al. (2022) investigated stu-
dents’ perceived AI knowledge gain, readiness, confidence, relevance, and motiva-
tion toward AI among 335 students. Ng et al. (2021a, b) used an AI-adapted
9.4 Scaffolding Support 127

Motivated Strategies for Learning Questionnaire to measure students’ intrinsic

motivation, learning strategy, self-regulation, satisfaction, and knowledge acquisi-
tion. In higher education, Kong et al. (2021) recruited 120 university students with
diverse backgrounds to evaluate students’ AI concepts, literacy, and empowerment
in an AI literacy program.

9.4.3 Knowing Students’ Learning Progress

Scaffolding is an instructional method that progressively moves students toward

deeper understanding during their learning process. Similar to other disciplines,
educators require scaffolding throughout students’ knowledge construction which
helps students accomplish tasks and assessments. This idea is influenced by
Vygotsky’s (1978) concept of the zone of proximal development. It is important to
acknowledge the areas that learners cannot do, what they can accomplish unaided,
and what they can do with assistance in AI literacy education. Knowing about stu-
dents’ backgrounds, interests, and abilities, designers can enable developers and
educators to refine their learning materials, activities, tools and platforms, and
pedagogies.
Designers should offer ways to customize the display of knowledge through
hands-on and minds-on learning activities. For example, educators can explain how
intelligent agents make decisions via graphical visualizations, simulations, and
multimedia demonstrations. These media-rich materials can aid students in decod-
ing challenging concepts and maximizing their understanding and knowledge trans-
fer. On the other hand, minds-on experiences require the use of higher order skills,
reflection, and thinking (e.g., collaboration, communication, creativity) on the tasks
through embodied and physical interactions. Guided by the constructionist principle
of “learning by doing/making,” students can construct tangible and shareable work
in the authentic world via engaging in artifact creations with craft materials, web-­
based AI tools, LEGO bricks, and sensors. Students can put themselves in the intel-
ligent agents’ “shoes” so that they can make sense of agents’ reasoning, it enables
students to mix their technological and creativity while exploring new ways to con-
struct and actively make their artifacts through student-centered, discovery, and
project-based learning, while in minds-on learning, the learners can further “think
about what they are learning” through students’ collaboration and social interaction
with their peers.

9.4.4 Parental Involvement

At last, parental involvement in students’ learning progress remains the key factor
contributing to their success. However, parents have different academic back-
grounds, and parental acceptance of technology, robots, and AI is diverse. Research
128 9 AI Literacy on Human-Centered Considerations

evidence has shown that screen-based technologies and robots can outperform chil-
dren’s story time (Lin et al., 2021; Zhang et al., 2022). Zhang et al. (2022) designed
an AI-driven interactive storytelling platform called StoryBuddy to act as a compan-
ion or a peer to children as it communicates through natural language dialogues.
The system assists parents in asking their children questions, recommending ques-
tions to use, and proposing follow-up questions. These supports can help parents
come up with better questions to fulfill their skill development and assessment goals
and reduce their cognitive load. In this way, parents can allocate more attention to
interact with their children to build a good parent-child relationship and keep their
child engaged and entertained. Therefore, designers need to work with parents to
design meaningful features to support their children’s AI learning, benefit family
well-being, and enhance equity in children’s access to technology.
Overall, we need to engage teachers, learners, parents, and other education stake-
holders to work with developers, scientists, and policymakers to develop a frame-
work to facilitate students to develop AI literacy skills.

References

Chen, H., Gomez, C., Huang, C. M., & Unberath, M. (2022). Explainable medical imaging AI
needs human-centered design: Guidelines and evidence from a systematic review. npj Digital
Medicine, 5(1), 1–15.
Chiu, M. C., Hwang, G. J., Hsia, L. H., & Shyu, F. M. (2022). Artificial intelligence-supported art
education: A deep learning-based system for promoting university students’ artwork apprecia-
tion and painting outcomes. Interactive Learning Environments, 1–19.
Dam, R., & Siang, T. (2018). Five stages in the design thinking process. Retrieved from https://
www.interaction-­design.org/literature/article/5-­stages-­in-­the-­designthinking-­process.
Dhungel, A. K., Wessel, D., Zoubir, M., & Heine, M. (2021). Too bureaucratic to flexibly learn
about AI? The human-centered development of a MOOC on artificial intelligence in and for
public administration. In Mensch und computer 2021 (pp. 563–567).
Dignum, V. (2019). Responsible artificial intelligence: How to develop and use AI in a responsible
way. Springer Nature.
Hannafin, M. J., & Land, S. M. (2000). Technology and student-centered learning in higher educa-
tion: Issues and practices. Journal of Computing in Higher Education, 12(1), 3–30.
Ipsita, A., Erickson, L., Dong, Y., Huang, J., Bushinski, A. K., Saradhi, S., Villanueva, A., Peppler,
K., Redick, T., & Ramani, K. (2022, April). Towards modeling of virtual reality welding simu-
lators to promote accessible and scalable training. In CHI conference on Human Factors in
Computing Systems (pp. 1–21).
Ives, J., & Castillo-Montoya, M. (2020). First-generation college students as academic learners: A
systematic review. Review of Educational Research, 90(2), 139–178.
Jain, A., Way, D., Gupta, V., Gao, Y., de Oliveira Marinho, G., Hartford, J., et al. (2021).
Development and assessment of an artificial intelligence–based tool for skin condition diag-
nosis by primary care physicians and nurse practitioners in teledermatology practices. JAMA
Network Open, 4(4), e217249–e217249.
Jonassen, D. H. (2000). Computers as Mindtools for schools: Engaging critical thinking. Merrill/
Prentice-Hall.
Kong, H., Yuan, Y., Baruch, Y., Bu, N., Jiang, X., & Wang, K. (2021). Influences of artificial
intelligence (AI) awareness on career competency and job burnout. International Journal of
Contemporary Hospitality Management.
References 129

Lin, P. Y., Chai, C. S., Jong, M. S. Y., Dai, Y., Guo, Y., & Qin, J. (2021). Modeling the structural
relationship among primary students’ motivation to learn artificial intelligence. Computers and
Education: Artificial Intelligence, 2, 100006.
Long, D., & Magerko, B. (2020, April). What is AI literacy? Competencies and design consid-
erations. In Proceedings of the 2020 CHI conference on human factors in computing systems
(pp. 1–16).
Luckin, R. (2017). Towards artificial intelligence-based assessment systems. Nature Human
Behaviour, 1(3), 1–3.
Miller, T. (2019). Explanation in artificial intelligence: Insights from the social sciences. Artificial
Intelligence, 267, 1–38.
MIT. (2019). Computing and artificial intelligence: Humanistic perspectives from MIT. Retreived
from https://news.mit.edu/2019/computing-and-ai-humanistic-perspectives-0924
Morrison, K. (2021). Artificial intelligence and the NHS: A qualitative exploration of the factors
influencing adoption. Future Healthcare Journal, 8(3), e648.
Ng, D. T. K., Leung, J. K. L., Chu, K. W. S., & Qiao, M. S. (2021a). AI literacy: Definition, teach-
ing, evaluation and ethical issues. Proceedings of the Association for Information Science and
Technology, 58(1), 504–509.
Ng, D. T. K., Leung, J. K. L., Chu, S. K. W., & Qiao, M. S. (2021b). Conceptualizing AI literacy:
An exploratory review. Computers and Education: Artificial Intelligence, 2, 100041.
Ng, D. T. K., Luo, W., Chan, H. M. Y., & Chu, S. K. W. (2022). Using digital story writing as a
pedagogy to develop AI literacy among primary students. Computers and Education: Artificial
Intelligence, 3, 100054.
Patrizio, A. (2017). Beware of companies claiming products have AI capabilities. Retrieved from
https://www.networkworld.com/article/3209713/beware-­of-­companies-­claiming-­products-­
have-­ai-­capabilities.html
Pérez-Ortiz, M., Rivasplata, O., Shawe-Taylor, J., & Szepesvári, C. (2021). Tighter risk certificates
for neural networks. Journal of Machine Learning Research, 22.
Renz, A., & Vladova, G. (2021). Reinvigorating the discourse on human-centered artificial intel-
ligence in educational technologies. Technology Innovation Management Review, 11(5).
Reinhardt, K. (2022). Trust and trustworthiness in AI ethics. AI and Ethics, 1–10. https://doi.
org/10.1007/s43681-022-00200-5
Rudin, C. (2019). Stop explaining black box machine learning models for high stakes decisions
and use interpretable models instead. Nature Machine Intelligence, 1(5), 206–215.
Shneiderman, B. (2020a). Human-centered artificial intelligence: Reliable, safe & trustworthy.
International Journal of Human–Computer Interaction, 36(6), 495–504.
Shneiderman, B. (2020b). Human-centered artificial intelligence: Three fresh ideas. AIS
Transactions on Human-Computer Interaction, 12(3), 109–124.
Toreini, E., Aitken, M., Coopamootoo, K., Elliott, K., Zelaya, C. G., & Van Moorsel, A. (2020,
January). The relationship between trust in AI and trustworthy machine learning technolo-
gies. In Proceedings of the 2020 conference on fairness, accountability, and transparency
(pp. 272–283).
Vygotsky, L. S. (1978). Socio-cultural theory. Mind in Society, 6(3), 23–43.
Webber, K. L. (2012). The use of learner-centered assessment in US colleges and universities.
Research in Higher Education, 53(2), 201–228.
Xu, W. (2019). Toward human-centered AI: A perspective from human-computer interaction.
Interactions, 26(4), 42–46.
Yang, S. J., Ogata, H., Matsui, T., & Chen, N. S. (2021). Human-centered artificial intelligence
in education: Seeing the invisible through the visible. Computers and Education: Artificial
Intelligence, 2, 100008.
Zhang, H. T., Park, T. J., Islam, A. N., Tran, D. S., Manna, S., Wang, Q., et al. (2022). Reconfigurable
perovskite nickelate electronics for artificial intelligence. Science, 375(6580), 533–539.
Chapter 10
AI Literacy from Educators’ Perspectives

After understanding AI literacy from the perspective of human-design factor, this

chapter presents a conceptual framework introducing an expanded view of AI lit-
eracy from educators’ perspectives. It moves beyond technological competencies
and tries to identify a more holistic and broader understanding about AI literacy
education. When using these novel AI tools to teach, educators need to be equipped
with adequate technological literacy skills and knowledge. In this way, they can
teach AI literacy and promote other digital competencies such as collaboration and
communication among their students in AI-driven environments. Since teachers
may not have rich technical knowledge to apply AI educational applications to facil-
itate their teaching. As one of the most important twenty-first-century competen-
cies, AI literacy can be conceptualized as the knowledge, skills, and attitudes
necessary to be competitive in the twenty-first-century workforce. Teacher educa-
tion and professional development should be reworked to incorporate training in
teaching key digital competencies.

10.1 Understanding Teachers’ AI Digital Competencies

Artificial intelligence technology has gained its popularity in education in recent

years. It reshapes education which helps educators to automate nonteaching tasks to
reduce their workload and enhance learning analysis and teaching optimization
(Kexin et al., 2020). Further, it helps promote students’ personalized learning
(Ahmad et al., 2022) and advance students’ knowledge acquisition using
AI-empowered systems (e.g., Chen et al., 2020; Hwang et al., 2020). However,
these technologies are novel to educators who may not have enough knowledge and
skills to manipulate these AI-driven educational applications for learning and teach-
ing purposes (Ally, 2019; Seo et al., 2021). They may not have rich technological
experience to design AI-driven learning environments, construct learning analysis,

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 131
D. T. K. Ng et al., AI Literacy in K-16 Classrooms,
https://doi.org/10.1007/978-3-031-18880-0_10
132 10 AI Literacy from Educators’ Perspectives

and set the rules in various AI-enhanced systems to automate responses, assign-
ments, and feedback (Seo et al., 2021), not to mention developing students’ AI lit-
eracy (Yau et al., 2022). Moreover, challenges were identified such as AI-based
misunderstandings and misleadingness, limitations, and hidden ethical issues
behind different platforms (Akgun & Greenhow, 2021; Sijing & Lan, 2018).
To develop student’ AI literacy, educators need to upskill and reskill their knowl-
edge to design relevant instruction (e.g., pedagogy, content, activities) for their stu-
dents (Williamson & Eynon, 2020). Research examples were used to identify what
types of AI digital competencies should be emphasized for teachers. A set of teach-
ers’ AI digital competencies is proposed based on the P21’s Framework for the 21st
Century Learning. This chapter aims to extend the proposal of viewing AI literacy
as an important twenty-first-century skill set (Ng et al., 2021a, b). As mentioned in
Chap. 4, AI literacy moves toward a broader understanding that identifies other
nontechnical, critical, and complex literacy that K–16 learners need to learn to
manipulate AI technologies ethically and effectively. Toward a holistic picture,
teachers should not view AI literacy as an independent domain but an avenue to
develop other important skill sets such as life and career skills, multidisciplinary
skills, learning, and innovation skills. With the P21 framework, teachers can use it
as a guideline to equip themselves with necessary digital competencies to facilitate
better instruction (Lindfors et al., 2021).
Several digital competency frameworks have been used in teacher education,
such as DigCompEdu (Caena & Redecker, 2019) and the International Society for
Technology in Education (ISTE) standards (Crompton, 2017) to update the neces-
sary digital competencies for today’s digital world. These frameworks inform edu-
cators how AI should be incorporated in their subject areas and levels of study.
According to the DigCompEdu standards, teachers’ digital competencies can be
categorized in six areas that facilitate their professional activities (Misthou &
Paliouras, 2022): (1) professional engagement, using AI for communication, col-
laboration, and professional development; (2) resource management, sourcing, cre-
ating, and sharing AI-empowered resources; (3) teaching and learning, managing
the use of AI for teaching/learning; (4) assessment, using AI to enhance automatic
assessment and analysis; (5) empowering learners, Using AI to enhance inclusion,
personalization, and learning engagement; and (6) facilitating learners’ AI compe-
tence, enabling learners to create their own intelligent agents responsibly and use AI
for information sharing, communication, content creation, and problem-solving.
The ISTE standards consider AI as a new digital competence requirement to develop
citizens’ necessary knowledge, skills, and attitudes (Riina et al., 2022). These
frameworks offer a common understanding across the European Union and beyond
of what digital competence needs to frame digital skills policy, curricula develop-
ment, and assessment of digital skills (Riina et al., 2022).
Among these models, this Chapter proposes the P21’s Framework for the 21st
Century Learning illustrates the knowledge and skills that educators need to suc-
ceed in working, learning, and living (National Research Council, 2012). The
framework is famous for educators and business leaders to identify learning stan-
dards for implementing curriculum, instruction, assessments, and learning
10.2 Essential AI Digital Competencies for Educators (P21) 133

environments. This chapter does not focus on the technical perspectives for foster-
ing students’ AI literacy. Instead, it aims to focus on broader digital competencies
(e.g., communication, collaboration, critical thinking, life, and career skills) that
support teachers to conduct teaching using AI technologies and implement AI lit-
eracy education. In this way, teachers should not only know and use AI applications
to empower their students and prepare related teaching resources. They should also
master other AI digital competencies such as applying AI knowledge in their disci-
plines and life and career skills to prepare students to enter the future workforce and
communicating and collaborating with their peers.

10.2 Essential AI Digital Competencies for Educators (P21)

Recent scholars proposed the need to reconsider teachers’ professional develop-

ment programs to meet the educational standards and expectations in the new digital
era. For example, Trujillo-Cabezas (2020) suggested some teacher’s AI digital com-
petencies such as using basic applications, managing information, creating learning
content, and connecting their students via technology. Ng et al. (2021b) incorpo-
rated AI literacy development into the TPACK model which guides teachers to con-
sider related teaching content, pedagogy, and technology. Using the P21’s
Framework, this section investigates what teacher competencies are desired for
effective online teaching with AI technology. Four key competencies are identified.
The key competencies of teachers should focus not only on the acquisition of basic
AI knowledge and skills but also on the cultivation of the capabilities necessary for
the teacher’s adaptation to, survival in, and control over the future society and their
lifelong professional development.
Core Subjects, 3Rs, and Twenty-First-Century Themes To be a successful stu-
dent and adult in the future, educators have a role to help students to master the core
subjects and twenty-first-century themes. AI has commonly become a topic in com-
puter science education, but it has already become part of these core subjects of
education. In schools, teachers are from diverse educational backgrounds and sub-
jects, and they need to collaborate and work together to adopt, adapt, and implement
interdisciplinary materials and approaches. Artificial intelligence technologies
could be new to most teachers, and it takes effort for educators to incorporate AI
into their knowledge domains and other twenty-first-century themes such as global
awareness and financial, environmental, civic, and health literacy. For example, Ng
et al. (2022) designed a digital story writing program to develop students’ AI con-
cepts, and students can use these concepts to write their stories. When writing their
books, translation tools and recommendation systems can help students reduce
grammatical errors and generate ideas and artworks in their digital stories which
involve the efforts of language and computer teachers.
Moreover, teachers who are more capable of using AI can empower students’ learn-
ing using AI-driven technologies (e.g., recommendation systems, intelligent agents)
134 10 AI Literacy from Educators’ Perspectives

in their classrooms. They can enable students to apply their multidisciplinary

knowledge and help them reach higher cognition levels such as critical thinking,
creativity, collaboration, and communication. In higher education, AI literacy edu-
cation is addressed across different subject areas such as healthcare (Rampton et al.,
2020), financial and business (Guerrero-Roldan et al., 2021), and legal education
(Demchenko et al., 2021). Higher education educators need to empower their stu-
dents to use AI applications throughout their life cycle, harness opportunities of
using AI technologies in their fields, and be aware of ethical implications and risks
(Zawacki-Richter et al., 2019). Concurrently, teachers, regardless of their subject
domains, need to develop certain AI digital competencies in order to instruct and
assess their students using AI technologies. Ultimately, teachers should be able to
guide students on how to manipulate different AI-driven systems and machines that
would play a part in their future workplaces.
Nowadays, AI can empower students’ reading, writing, and arithmetic abilities
(3Rs). For example, e-book systems consist of AI-empowered recommendations
that suggest reading materials, practices, and assignments based on students’ read-
ing habits (Wu & Peng, 2017). Moreover, AI writing software helps people generate
ideas and content for writing articles, novels, blogs, and emails to optimize their
work. It can produce compelling articles with the right tone and style, ensure them
to be grammatically correct, paraphrase texts, reduce grammatical errors, generate
citations automatically, and enhance formatting that improve readability (Nazari
et al., 2021; Lin & Chang, 2020). In arithmetic calculation, other applications (e.g.,
Wolfram Alpha, Symbolab) enable students to take photos of mathematical formu-
las and questions to generate steps and solve them automatically. However, these
applications are too convenient for students to plagiarize in their assignments and
examinations (Wu et al., 2021). Teachers should be more sensitive to these
AI-enhanced submissions and differentiate whether it is the students’ original ideas.
It is important for teachers to remind students with correct mindsets and ethical
concerns when using these applications to facilitate their learning and living.
Learning and Innovation Skills (4Cs) Scholars have strived the importance of
critical, creative thinking, collaboration, and communication in particular to prepare
students for today’s ever-changing workforce (Van Laar et al., 2017). The skill sets
can enable students to create and adapt to changes and encourage students to apply
their existing knowledge with peers to solve problems creatively and critically.
Studies have shown the effectiveness of using AI technologies to enhance students’
learning and innovation skills. For example, when AI is used in music education, it
empowers student learning experience through interactive composition, perfor-
mances, mixing, and appreciation (Zulić, 2019).
Demchenko et al. (2021) identified a digital transformation in legal education that
teachers need to equip their students with AI digital competencies. There is a need
for law schools to innovate and form stronger interdisciplinary collaboration with
AI expertise to enhance their effective professional and everyday activities such as
using AI-based tools for criminal justice and identifying algorithmic bias (Deeks,
2019). In business education, AI has a wide range of uses in business to complete
10.2 Essential AI Digital Competencies for Educators (P21) 135

authentic tasks such as aggregating business data, managing customer relationships,

and predicting future trends. Business educators need to update their knowledge and
enable students to integrate AI into their workplace and create a new user experi-
ence for their clients (Williamson & Eynon, 2020; Uzialko, 2022). In medical edu-
cation, there is a greater demand for digital health applications to promote social
distancing during the pandemic. Medical educators need to timely update their
knowledge and skills to help their e-patients via AI-driven medical applications
such as infection rate prediction, computed tomography image recognition, robotics
for patient assessment and drug delivery, and X-ray interpretation (Ahuja & Nair,
2021). These knowledge and skills are useful for teachers and students to become
professionals/leaders in their knowledge fields to implement complex cognitive and
decision-making tasks and adapt to present scenarios.
Information, Media, and Technology Skills Educators need to prepare them-
selves to become digital ready so that they are able to teach students related skills
such as information, media, and ICT literacy (Gleason & Von Gillern, 2018).
Especially in recent years, teachers and students need to adapt to digital transforma-
tion and develop related technological skills. In AI-driven classrooms, teachers need
to manipulate different AI-enhanced systems to design assessments and examine
students’ performance using their historical and current data using the adaptive
learning system (Guerrero-Roldán et al., 2021). In another study, teachers adopt an
automatic mode in an AI-driven service called IBM RXN to enable students to draw
target molecules and generate chemical reactions and structure representations
(Healy & Blade, 2020). Kostopoulos et al. (2021) mentioned the use of an AI-driven
system called DevOps to equip smart city professionals and educators with ade-
quate technological skills to visualize urban innovation in an 11-week online course.
Further, teachers need to acquire better technology skills and enable them to inter-
vene early to enhance students’ performance and retention (Jokhan et al., 2022).
They need to learn how to use different AIED technologies to design their instruc-
tion, examine students’ learning outcomes, and premise intelligent support for col-
laborative working. These examples showed that university educators from different
disciplines need to equip themselves with technological skills to enable their stu-
dents to express knowledge, solve problems, and manipulate AI-driven applications
so that they are ready to work in AI-driven environments.

Life and Career Skills Life and career skills are important to prepare students to
engage as citizens in a dynamic global community and meet different challenges
and opportunities in the workforce. Students need to develop positive mindsets,
attitudes, and other competencies (e.g., flexibility, adaptability, self-direction, social
skills, productivity, responsibility) to navigate complex life and work environments
(Van Laar et al., 2017). First, studies suggested that AI has the potential to transform
youth employment and students need to develop relevant skills to adapt to this
change. For example, Singh et al. (2020) suggested that AI profiling will move away
from merely information collection about formal qualifications to a more holistic
approach of capturing skills and life experiences. Educators need to upgrade their
136 10 AI Literacy from Educators’ Perspectives

students to fit the future job market. Second, improving students’ self-efficacy and
self-regulation is important when using AI-driven systems to support students’
online learning since these systems usually do not consist of a physical teacher to
monitor their learning (Guerrero-Roldán et al., 2021). Third, Cetindamar et al.
(2022) highlighted four sets of workplace capabilities associated with AI: techno-
logical skills (e.g., data collection, analytics, ethics, security), work-related skills
(e.g., decision-making, critical thinking, teamwork), human-machine interaction
(e.g., situation assessments, affordance analysis, adaptive expertise), and learning-­
related capabilities (e.g., lifelong learning, self-learning ability). Other studies also
mentioned the importance of life and career skills in the fourth industrial revolution
such as problem-solving (Mohammed et al., 2021), emotional intelligence, judg-
ment, service orientation, negotiating and cognitive flexibilities (Webber-Youngman,
2017), as well as communication and teamwork skills (Seo et al., 2021). Teachers
can enable their students to become adaptive thinkers who equip themselves with
technological literacies to solve problems, think critically, lead their teammates, and
implement reflective practice (Li & Du, 2017). With these life and workplace skills,
students become more digitally ready to contribute to their fields and companies
after graduation.
These four essential digital competencies suggest how higher education policymak-
ers consider the necessary educational standards and goals in their universities and
schools and provide relevant professional training to develop teachers’ readiness in
today’s AI world. At the classroom level, the framework serves as a guideline to
help instructional designers to design suitable curriculum and materials for their
students and create positive learning environments. Moreover, such educational
standards are important to serve as a basis of educational reform and digital transi-
tion to help practitioners to identify the necessary learning outcomes to meet the
goals set by the governments, regions, and markets. For example, it helps fulfill job
demands in the market, enhance learners’ competitiveness, equipping students with
futuristic skills, and educate the next generation to become responsible citizens. To
achieve these goals, educators need to cultivate their TPACK knowledge to incorpo-
rate meaningful learning elements (e.g., assessments, curriculum, instruction) for
their students. Since teachers may not be familiarized with these novel technologies,
practices for teacher education (e.g., teacher program, professional community,
guideline) are important to support educators to equip with necessary knowledge
and skills for students’ learning and achieve teaching and learning outcomes.
Moreover, collaboration from different professionals from higher education institu-
tions, schools, government, industries, and companies could co-design these profes-
sional development programs and guidelines to develop meaningful materials,
tools, and platforms to support teachers AI literacy education.
References 137

10.3 Conclusion

The existing frameworks focused more on students’ learning; few of them investi-
gate how to enhance teachers’ AI digital competence in professional development
programs. However, it is important to promote frontline educators and decision-­
makers’ professional growth so as to improve their pedagogical, content, and tech-
nological knowledge to incorporate AI into their classrooms. Research shows that
professional development could lead to better instruction, thus improving student
learning outcomes (e.g., Kutaka et al., 2017; Gore et al., 2021). Teachers become
more capable of connecting their teaching materials to AI and understand the edu-
cational standards that guide their teaching, assess, and design students’ learning
experiences with novel technologies. Further, they can use it wisely to solve teach-
ing problems (e.g., lack of social isolation and motivation in online learning). With
AI, teachers can analyze students’ behaviors and performance to adapt their teach-
ing and give students immediate assistance to meet learning needs for individual
learners.
This Chapter proposed the P21’s Framework for 21st Century Learning. Research
examples are suggested for each competence to show what teachers should learn
before teaching their students. Teachers need to learn technological skills to access
AI devices and software (e.g., learning analytic) to improve their teaching and
working efficiency (Kexin et al., 2020). On top of technical skills, there is a need to
include broader digital competencies such as communication, life and career skills,
ethical concerns, teacher identity, attitudes, and mindsets as components of teacher
education in AI literacy education (e.g., Akgun & Greenhow, 2021; Seo et al., 2021).

References

Ahmad, S. F., Alam, M. M., Rahmat, M. K., Mubarik, M. S., & Hyder, S. I. (2022). Academic and
administrative role of artificial intelligence in education. Sustainability, 14(3), 1101.
Ahuja, V., & Nair, L. V. (2021). Artificial intelligence and technology in COVID Era: A narrative
review. Journal of Anaesthesiology, Clinical Pharmacology, 37(1), 28.
Akgun, S., & Greenhow, C. (2021). Artificial intelligence in education: Addressing ethical chal-
lenges in K-12 settings. AI and Ethics, 1–10.
Ally, M. (2019). Competency profile of the digital and online teacher in future education. The
International Review of Research in Open and Distance Learning, 20(2).
Caena, F., & Redecker, C. (2019). Aligning teacher competence frameworks to 21st century chal-
lenges: The case for the European digital competence framework for educators (Digcompedu).
European Journal of Education, 54(3), 356–369.
Cetindamar, D., Kitto, K., Wu, M., Zhang, Y., Abedin, B., & Knight, S. (2022). Explicating AI
literacy of employees at digital workplaces. IEEE Transactions on Engineering Management.
Chen, X., Xie, H., Zou, D., & Hwang, G. J. (2020). Application and theory gaps during the rise of
artificial intelligence in education. Computers and Education: Artificial Intelligence, 1, 100002.
Crompton, H. (2017). ISTE standards for educators: A guide for teachers and other professionals.
International Society for Technology in Education.
138 10 AI Literacy from Educators’ Perspectives

Demchenko, M. V., Gulieva, M. E., Larina, T. V., & Simaeva, E. P. (2021). Digital transforma-
tion of legal education: Problems, risks and prospects. European Journal of Contemporary
Education, 10(2), 297–307.
Deeks, A. (2019). The judicial demand for explainable artificial intelligence. Columbia Law
Review, 119(7), 1829–1850.
Gleason, B., & Von Gillern, S. (2018). Digital citizenship with social media: Participatory prac-
tices of teaching and learning in secondary education. Journal of Educational Technology &
Society, 21(1), 200–212.
Gore, J. M., Miller, A., Fray, L., Harris, J., & Prieto, E. (2021). Improving student achievement
through professional development: Results from a randomised controlled trial of quality teach-
ing rounds. Teaching and Teacher Education, 101, 103297.
Guerrero-Roldán, A. E., Rodríguez-González, M. E., Bañeres, D., Elasri-Ejjaberi, A., & Cortadas,
P. (2021). Experiences in the use of an adaptive intelligent system to enhance online learn-
ers’ performance: A case study in economics and business courses. International Journal of
Educational Technology in Higher Education, 18(1), 1–27.
Healy, E. F., & Blade, G. (2020). Tips and tools for teaching organic synthesis online. Journal of
Chemical Education, 97(9), 3163–3167.
Hwang, G. J., Xie, H., Wah, B. W., & Gašević, D. (2020). Vision, challenges, roles and research
issues of artificial intelligence in education. Computers and Education: Artificial Intelligence,
1, 100001.
Jokhan, A., Chand, A. A., Singh, V., & Mamun, K. A. (2022). Increased digital resource consump-
tion in higher educational institutions and the artificial intelligence role in informing decisions
related to student performance. Sustainability, 14(4), 2377.
Kexin, L., Yi, Q., Xiaoou, S., & Yan, L. (2020). Future education trend learned from the Covid-19
pandemic: Take artificial intelligence online course as an example. In 2020 International
Conference on Artificial Intelligence and Education (ICAIE) (pp. 108–111). IEEE.
Kostopoulos, G., Panagiotakopoulos, T., Kotsiantis, S., Pierrakeas, C., & Kameas, A. (2021).
Interpretable models for early prediction of certification in MOOCs: A case study on a MOOC
for Smart City professionals. IEEE Access, 9, 165881–165891.
Kutaka, T. S., Smith, W. M., Albano, A. D., Edwards, C. P., Ren, L., Beattie, H. L., Lewis, W. J.,
Heaton, R. M., & Stroup, W. W. (2017). Connecting teacher professional development and
student mathematics achievement: A 4-year study of an elementary mathematics specialist pro-
gram. Journal of Teacher Education, 68(2), 140–154.
Li, D., & Du, Y. (2017). Artificial intelligence with uncertainty. CRC Press.
Lin, M. P. C., & Chang, D. (2020). Enhancing post-secondary writers’ writing skills with a chat-
bot. Journal of Educational Technology & Society, 23(1), 78–92.
Lindfors, M., Pettersson, F., & Olofsson, A. D. (2021). Conditions for professional digital compe-
tence: The teacher educators’ view. Education Inquiry, 12(4), 390–409.
Misthou, S., & Paliouras, A. (2022). Teaching Artificial Intelligence in K-12 Education:
Competences and Interventions. In Interactive Mobile Communication, Technologies and
Learning, Springer, Cham, pp. 887–896.
Mohammed, Z., Arafa, A., Atlam, E. S., El-Qerafi, N., El-Shazly, M., Al-Hazazi, O., & Ewis,
A. (2021). Psychological problems among the university students in Saudi Arabia during the
COVID-19 pandemic. International Journal of Clinical Practice, 75(11), e14853.
National Research Council. (2012). Education for life and work: Developing transferable knowl-
edge and skills in the 21st century. National Academies Press.
Nazari, N., Shabbir, M. S., & Setiawan, R. (2021). Application of artificial intelligence powered
digital writing assistant in higher education: Randomized controlled trial. Heliyon, 7(5), e07014.
Ng, D. T. K., Leung, J. K. L., Chu, K. W. S., & Qiao, M. S. (2021a). AI literacy: definition, teach-
ing, evaluation and ethical issues. Proceedings of the Association for Information Science and
Technology, 58(1), 504–509.
Ng, D. T. K., Leung, J. K. L., Chu, S. K. W., & Qiao, M. S. (2021b). Conceptualizing AI literacy:
An exploratory review. Computers and Education: Artificial Intelligence, 2, 100041.
References 139

Ng, D. T. K., Luo, W., Chan, H. M. Y., & Chu, S. K. W. (2022). Using digital story writing as a
pedagogy to develop AI literacy among primary students. Computers and Education: Artificial
Intelligence, 3, 100054.
Rampton, V., Mittelman, M., & Goldhahn, J. (2020). Implications of artificial intelligence for
medical education. The Lancet Digital Health, 2(3), e111–e112.
Riina, V., Stefano, K., & Yves, P. (2022). DigComp 2.2: the digital competence framework for
citizens – With new examples of knowledge, skills and attitudes. Retrieved from https://publica-
tions.jrc.ec.europa.eu/repository/handle/JRC128415
Seo, K., Tang, J., Roll, I., Fels, S., & Yoon, D. (2021). The impact of artificial intelligence on
learner–instructor interaction in online learning. International Journal of Educational
Technology in Higher Education, 18(1), 1–23.
Sijing, L., & Lan, W. (2018). Artificial intelligence education ethical problems and solu-
tions. In 2018 13th International Conference on Computer Science & Education (ICCSE)
(pp. 1–5). IEEE.
Singh, S., Sharma, P. K., Yoon, B., Shojafar, M., Cho, G. H., & Ra, I. H. (2020). Convergence of
blockchain and artificial intelligence in IoT network for the sustainable smart city. Sustainable
Cities and Society, 63, 102364.
Trujillo-Cabezas, R. (2020, June). Towards the development of future trend scenarios through
dynamic analysis: A proposal of integration of Artificial Intelligence, Data Sciences and the
field of Futures Studies to adapt to new environments. In 2020 15th Iberian Conference on
Information Systems and Technologies (CISTI) (pp. 1–6). IEEE.
Uzialko, A. (2022). How artificial intelligence will transform businesses. Retrieved from https://
www.businessnewsdaily.com/9402-­artificial-­intelligence-­business-­trends.html
Van Laar, E., Van Deursen, A. J., Van Dijk, J. A., & De Haan, J. (2017). The relation between 21st-­
century skills and digital skills: A systematic literature review. Computers in Human Behavior,
72, 577–588.
Webber-Youngman, R. C. W. (2017). Life skills needed for the 4th industrial revolution. Journal of
the Southern African Institute of Mining and Metallurgy, 117(4), iv–v.
Williamson, B., & Eynon, R. (2020). Historical threads, missing links, and future directions in AI
in education. Learning, Media and Technology, 45(3), 223–235.
Wu, J. Y., & Peng, Y. C. (2017). The modality effect on reading literacy: Perspectives from stu-
dents’ online reading habits, cognitive and metacognitive strategies, and web navigation skills
across regions. Interactive Learning Environments, 25(7), 859–876.
Wu, J. S., Chien, T. H., Chien, L. R., & Yang, C. Y. (2021). Using artificial intelligence to predict
class loyalty and plagiarism in students in an online blended programming course during the
COVID-19 pandemic. Electronics, 10(18), 2203.
Yau, K. W., Chai, C. S., Chiu, T. K., Meng, H., King, I., & Yam, Y. (2022). A phenomenographic
approach on teacher conceptions of teaching artificial intelligence (AI) in K-12 schools.
Education and Information Technologies, 1–24.
Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of
research on artificial intelligence applications in higher education–where are the educators?
International Journal of Educational Technology in Higher Education, 16(1), 1–27.
Zulić, H. (2019). How AI can change/improve/influence music composition, performance and edu-
cation: three case studies. INSAM Journal of Contemporary Music, Art and Technology, 1(2),
100–114.
Chapter 11
Summary and Conclusions

In the end, we hope to have taken our readers through a journey of discovering the
origins and development of AI literacy and, more importantly, positioning the
unique role that AI literacy education plays in our rapidly changing, digitized,
twenty-first-century education. Part I has posited that AI literacy is different from
AI education; while they share overlapping domains, they are differentiated by the
scope, aim, and outcomes of the two. Hence, we borrowed concepts from other lit-
eracies to introduce a holistic view of AI learning for a wide range of learners and
also the stakeholders involved in supporting them. That brings us to Part II, where
AI literacy is being discussed in depth according to four distinct educational levels
(kindergarten, primary, secondary, and nonengineering undergraduate). We applied
similar research methodology throughout the four levels to search for existing pub-
lications, which led us to a considerable number of results for thematic analyses.
The information obtained from Part II acted as a bridge. Not only does it confirm the
arguments and proposals on the growing importance of AI literacy stated in Part I
but also provides an ample amount of evidence to generate insights for us to navi-
gate the avenue towards an AI-literate future. At last in Part III is where we synthe-
sized. We gathered evidence, our insights, and international educational standards
and frameworks such as Bloom’s taxonomy, TPACK model, and P21’s Framework,
altogether and ultimately to suggest methods, approaches, guidelines, etc. for stake-
holders of interest to make efficient use of our data.
Our foremost intention to write this book is to propose a set of guidelines of AI
literacy instructional design that can be applied in today’s K–16 educational set-
tings. For researchers, the book provides research-driven recommendations to con-
sider theoretical and pedagogical bases for future studies. The methodology used
was by no means the only effective approach in consolidating data nor the single
way to interpret it. Plus, such a fast-growing industry requires more scholars to col-
lectively contribute to analyzing the landscape of AI literacy from different angles
methodologically, timely, geographically, etc.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 141
D. T. K. Ng et al., AI Literacy in K-16 Classrooms,
https://doi.org/10.1007/978-3-031-18880-0_11
142 11 Summary and Conclusions

Another notion we advocate is that AI literacy or even AI education should not

be viewed as an independent subject. We used the linkage between characteristics
of AI literacy and twenty-first-century skills as a gateway to enter students’ core and
foundation curriculum. Technological knowledge, skill acquisition (in a digital per-
spective), and broader competencies such as life and career, multidisciplinary, the
four Cs learning, and innovation skills (in a critical perspective) are all important
aspects to the development of appropriate teaching tools and platforms, meaningful
learning activities, and assessments. We have also discussed the instructional
designs in terms of teaching practices and human-centered (or even student-­
centered) considerations to meet students’ needs and scaffold on the foundations of
their learning. On this note, this book has put forward a range of pedagogical rec-
ommendations that facilitate twenty-first-century skills education from early child-
hood education to higher education to meet the needs of the industries and society.
To confirm the significance of this AI literacy arena, prior to the publication of
this book, the DigCompEdu and the International Society for Technology in
Education (ISTE) standards updated their guidelines and plans to include AI liter-
acy as the latest educational standards with regard to the digital upskilling of the
population. Furthermore, national education strategies across the United States,
China, and European Union and beyond have updated their plans to frame policy
and curriculum development to devise and affirm the role of artificial intelligence in
promoting AI literacy to citizens and fostering the next generation’s AI digital com-
petence. Inspired by the National Education Technology Plan, there are some gen-
eral recommendations for different stakeholders in terms of learning, teaching,
leadership, assessment, and infrastructure (Smith, 2015) (see Table 11.1).
Further, human-centered design consideration has a large role in the newly
updated plans, which suggests instructional designers (including developers and
teachers) to design their instruction and learning environments that meet learners’

Table 11.1 Recommendations based on national education technology plan

Domains Recommendations
Learning  Developing high-quality government-endorsed curricula to meet the needs of
their students
 Using AI technology to transform learning experiences with the goal of
providing greater equity and accessibility
 Universally offering equitable access to AI technology inside and outside of
school, regardless of students’ backgrounds
 Implementing universal design principles for student-centered learning
Teaching  Designing teacher professional developments to develop minimum standards
for educators to take full advantage of AI-empowered learning environments
Leadership  Leaders need to create a shared vision for AI technology to best meet the
needs of all students and to devise a plan to execute it
Assessment  Designing useful assessments and tools to examine students’ AI learning
outcomes to improve their learning effectiveness and reduce challenges to
students’ learning
Infrastructure  Establishing a robust AI-driven technology infrastructure for today’s schools,
meeting current connectivity goals, and augmenting it for future demands
11.1 For Teachers 143

needs (Yang et al., 2021). It is also our goal of this book to provide educators across
the globe with pedagogies, learning content, technologies, and assessment strate-
gies supported by state-of-the-art research evidence. We believe that this book will
facilitate stakeholders to design and implement AI literacy education more effec-
tively that grows our students up to become educated digital citizens with positive
knowledge, attitude, and skills learning gain in core AI-empowered curriculum
domains as well as twenty-century technological skills. Here, we summarize and
make recommendations for stakeholders who have responsibility in changing the
AI-driven landscape of today’s digital world. These stakeholders include, but not
limited to, educators, policymakers, researchers, and parents.

11.1 For Teachers

Teachers play an important role in making decisions about what learning elements
such as assessments, curriculum, pedagogy, and technology used in their class-
rooms for the needs to foster students’ AI literacy skills as their learning outcomes.
However, teachers may not have computer science educational backgrounds to
develop suitable technology driven curricula and learning materials for their stu-
dents. To enable a smooth digital transition to make our teaching professionals
ready and confident and help them adapt to changes in this era, teacher education in
AI literacy is important. A particular reason is that supporting students requires
teachers to become more digitally advanced so as to serve more aptly in their com-
petence. As such, professional development programs, guidelines, and schools’ sup-
port are required to upskill and reskill teachers’ technological knowledge, skills,
and attitudes. Moreover, there is a need to establish and maintain a community of
scholars, developers, and educators for sharing and disseminating best practices to
develop and refine AI teaching tools, materials, and assessments that meet students’
learning outcomes and needs. Through collaboration, frontline teachers do not need
to struggle on designing technology-based content and tools for their students.
Instead, they can focus on designing engaging learning activities and materials in
their classrooms via meaningful pedagogical approaches (e.g., collaborative learn-
ing, inquiry-based learning, gamification) to motivate their students in the learning
process.
Students may meet challenges throughout their AI literacy learning journeys.
Research suggests that experiential learning can help students have a taste of
AI-driven technologies (e.g., chatbot, social robots, music, and artwork generator).
In this way, students can scaffold their knowledge and skills whereby students can
learn by doing and reflecting on their experiences. As a form of experiential learn-
ing, hands-on experiences can help students to move away from teacher-centered
approaches and absorb new knowledge with their peers. However, it may not suffice
to define a task for students such as simulating machine learning using web-based
tools, training a model using datasets, designing an AI-enhanced game, and generat-
ing a song using AI. Students need authentic and structured instruction on AI
144 11 Summary and Conclusions

literacy in order to learn how to successfully know, understand, use, apply, evaluate,
and create relevant resources to meet their learning outcomes and create their arti-
facts. Therefore, after experiencing AI, as suggested by our reviews, interventions
involving student inquiry and collaboration to solve authentic problems can stimu-
late students’ positive learning gains (e.g., curiosity, creativity, knowledge acquisi-
tion) to apply their knowledge in real-world contexts. In other words, collaborative
project-based learning is the most sound and commonly used pedagogical approach
that helps students to reach a higher cognitive level.
At the same time, teachers are advised to hold back their worries that AI will
replace their job and overemphasize the limitations of AI. Although it may not be
possible to transform over one night, educators need to see a possibility to use AI
applications to facilitate their everyday routine work, understand learners’ needs
through data analysis, design an interactive learning environment, and provide auto-
matic and timely feedback for their students. They must learn to welcome the chal-
lenge of new technologies that may arise in their teaching process. Teachers who
embrace lifelong learning approaches are successful in creating innovative and
inclusive classrooms to grow students’ digital competence through hard work and
be willing to learn and open to changes. With such drive and faith in teachers’ adap-
tation, our next generation will become enthusiastic about AI and technologies to
help the society grow better and meet today’s educational needs.

11.2 For Higher Education Faculty

Higher education aims to develop talents and professionals who can fill the market
demands in their field. As AI is influencing nearly every discipline, it enhances the
efficiency of employees in the workplace and automates repetitive tasks and frees us
to conduct more complex problems and operations. Through turning complex data
into insights, AI can identify important changes in patterns and perform deeper
content analysis to uncover changes in today’s world and people’s behavior that
affect their industries so as to optimize their performance and profitability. In law
schools, faculty need to innovate and form stronger interdisciplinary collaboration
with AI expertise to enhance their effectiveness and use AI-based tools to help with
criminal justice and identify algorithmic bias (Deeks, 2019). In medical schools,
many health applications are now AI-empowered to help medical practitioners to
help their clients with infection rate prediction, computed tomography image recog-
nition, robotics for patient assessment and drug delivery, and X-ray interpretation
(Ahuja & Nair, 2021). In art schools, AI is drastically changing the nature of cre-
ative processes that facilitate illustrators, designers, and photographers to design
digital art packages (Boden & Edmonds, 2019). Many universities have designed AI
literacy programs to develop students’ skills and mindsets for students from educa-
tional diverse backgrounds to develop fundamental AI knowledge and skills. Some
courses aim to support students with basic programming skills for science and engi-
neering students to learn the underlying computer science concepts behind the AI
11.3 For Policymakers 145

technologies (Long & Magerko, 2020). The AI literacy programs further welcome
students without computer science prerequisite knowledge to empower them to
become AI literates to serve in their future workplaces after graduation.
Teacher training programs should help higher education faculty, postdocs, and
teaching support staff to realize the demands, opportunities, and challenges that
their industries face in the twenty-first century. It is necessary to situate education in
the times of the students to reflect on their teaching philosophies and practices to
better cater for learners’ needs. It helps groom faculty’s ability to critically rethink
education to cultivate and strengthen students with related knowledge and skills that
are needed in today’s world. The updated twenty-first-century skills frameworks
have included AI which are relatively novel to educators who do not receive relevant
training before in their teacher training, especially for senior professors who are
more likely to struggle in the acquisition of IT skills that can support more on how
to make sense of AI technologies into their expertise.

11.3 For Policymakers

Across the country, AI is rapidly changing how businesses operate and help people
solve complex problems and work more efficiently. By 2030, 70% of companies
expect to use AI according to a PwC’s global AI report (2022). Countries are work-
ing their best to gain global competitiveness through economic improvement and
creating innovation to accelerate industrial growth. With more novel technologies,
some of the existing practices and policies will not match past experience.
Tomorrow’s leaders started to discuss the global competition for AI leadership and
the need for educational policy frameworks that support learning and innovation.
Governments need to collaborate with universities to update its educational stan-
dards, policy, frameworks and guidelines, curricula, and evaluation methods indi-
cating the AI literacy interventions.
Policymakers play their roles to present the notable shift in digital skills to close
digital divide gaps in the population and nurture learners’ twenty-first-century skills
to make them ready for tomorrow’s workplace and enhance their competitiveness
compared with their counterparts. From the literature, we note that standardized
frameworks and guidelines are important for educational institutions to start their
AI literacy education. Policymakers should set strategies to develop AI and digital
literacy among teachers and students that require more than just making use of AI
in education. Without sufficient government support, different stakeholders includ-
ing developers, schools, investors, communities, and professional associations can-
not put their concerted efforts to input their resources and human capital to reach the
desired level of digital competence and facilitate the inclusion of AI literacy into
K–16 education. Government’s policies can align with the global needs and con-
sider its resources and budgets to support educational institutions financially to
enhance their infrastructure for AI-empowered classrooms and adjust curriculum
and assessment strategies in collaboration with AI professionals.
146 11 Summary and Conclusions

On top of accelerating talent development, realizing AI ethics and human-­

centered AI is also crucial to make our society safe, fair, and transparent. Many
studies have warned that misuse of AI could cause drastic consequences that make
the economy downturn and human lives (Baum, 2017; Dwivedi et al., 2020; Jobin
et al., 2019). Standardized AI ethical guidelines are important to make people
become responsible and trust their lives to AI-empowered machines. Trust and
transparency are crucial elements for citizens and governments to guard against
potential AI biases, and AI should help support human’s life, rather than replacing
human intelligence. From this perspective, there is a need to promote digital citizen-
ship education that helps people to gain understanding of what AI can do and how
to use it effectively and responsibly.

11.4 Parents

Researchers have identified the usage of AI for children’s health development.

However, whether they feel comfortable with AI use in education is still unknown.
With AI becoming more a part of our everyday lives, this novel technology may
affect the way they parent. In recent years, AI-empowered solutions purport to help
parents to free up their time to rest. For example, it assists parents in understanding
whether their children might be fussy, hungry, or in pain through analyzing the
expressions of babies and acoustic features of the baby’s cry. Another AI nanny can
analyze a baby’s sleep patterns and find reasons for its restlessness at night. Further,
AI robotics and toys can engage, educate, and entertain children to learn program-
ming, computational thinking, and AI literacy. However, when students meet learn-
ing challenges, parents may not know how to deal with them. Ng et al. (2022) found
that students may turn to search online after they know their parents do not know
about the topic, and sometimes it is challenging for young learners to critically think
about the correctness of the knowledge or information found online. If parents are
more technologically educated, they can spend more time with their children to
study the content together, handle the misunderstandings, and discuss with them,
thus having a good parent-child relationship. Further, parents support their children
financially, and parents who are AI literates can think critically about what counts as
AI, question the intelligence, and understand the limitations and ethical concerns
behind so that they could more wisely purchase these so-called AI-empowered
products. Providing support for parents can aid them in supporting their children’s
learning (Long & Margerko, 2020).
When students grow up in primary and secondary schools, educational applica-
tions such as intelligent agents, chatbots, and learning analysis evolve to help par-
ents and teachers to understand their children’s learning outcomes and experience.
However, there have been discussions about parenting and AI about privacy and
transparency issues that children’s behaviors may be monitored by developers and
technology companies. Moreover, the ways how the so-called AI generates such
academic results for students are unexplainable. Sometimes the mechanism of
11.5 Researchers and Developers 147

learning analysis could be a “black-box” or too technical that parents may not have
technology knowledge and background to understand.

11.5 Researchers and Developers

Artificial intelligence literacy education is a discipline that needs experts from dif-
ferent fields including educational researchers, technology developers, education
policymakers, and educators to collaborate and coordinate so as to ensure the qual-
ity of education. On one hand, researchers aim at generating new learning theories
and principles to address the global needs of AI literacy education. Through
research, we can understand the genuine needs of learners to refine the inputs of
education (i.e., pedagogies, learning content, technologies used, assessment strate-
gies, student-centered considerations). In the educational technology field, many
studies draw upcoming design-based reach as a methodological paradigm to refine
learning effectiveness to advance the existing practices to ensure high-quality learn-
ing outcomes for students.
On the other hand, developers should not merely focus on profit making with the
trend of AI. Instead, they need to align with research evidence to learn users’ back-
grounds, interests, and needs and produce and refine the products for students that
facilitate their learning, thus making a meaningful and responsible contribution to
AI literacy education. In this way, human-centered considerations are important to
raise attention to educate citizens to become socially responsible and ethical users
such as inclusiveness, fairness, accountability, transparency, and ethics, instead of
merely enhancing students’ AI abilities and interests.
The evidence referenced in this book employs a systematic scoping method that
summarizes the current state-of-the-art literature. As suggested by many scholars,
“digital literacy” varies, and so does “AI literacy,” and our book has synthesized
some of the theoretical frameworks for twenty-first-century skills (Reynolds, 2016).
Education researchers have to work actively and collaboratively to identify research
questions to provide theory and evidence basis for future research. First, there is a
need to update the theoretical, conceptual, and methodological underpinnings of AI
literacy education. Second, we need to evaluate the design and implementation of
AI literacy education for educators to meet the needs of learners; improve quality of
instruction, curricula, technology, and pedagogy utilized in classrooms; and rede-
fine rubrics and assessments (e.g., knowledge tests, questionnaire protocols, project,
and content analysis).
Several research gaps were found based on the literature review in this book.
First, there is a lack of AI literacy curricula and assessment methods. Although
recent studies have started to propose curricula and educational theories to identify
what types of content knowledge that should be included in K–16 levels, more
empirical studies can be conducted to examine the effectiveness of the learning
implementation. Second, AI literacy education is an emerging field and it is under-
standable that studies could be preliminary and exploratory without rigorous
148 11 Summary and Conclusions

research methods to investigate the theoretical and empirical perspectives in emerg-

ing fields. This suggests why design-based research is useful to explore effective
practices, design, reflect, and refine different approaches to teach and learn AI (Ng
& Chu, 2021). With rigorous evidence-based research using quantitative and quali-
tative methods, the effectiveness of AI curricula is carefully and critically examined.
In the near future, diverse learning conditions in intervention design (e.g., interven-
tion duration, frequency, curriculum structure, learning resources, teacher involve-
ment) were examined to achieve learners’ success for intended learning outcomes.
Further, it is foreseen that research designs will shift to be more empirical and inter-
ventional (e.g., quasi-experiments, design-based research) with clearly documented
treatment and control groups and varied data analysis procedures (e.g., regression,
structural equation modeling, t-test, ANOVA). Lastly, to advance the field of AI
literacy education, priority should be placed on developing an overarching frame-
work to guide educators in creating lesson designs that offer the most appropriate
pedagogies and learning artifacts to foster students’ AI literacy. The frameworks
proposed in this book may need further evidence to support its theoretical basis.
Researchers are encouraged to consider both short- and long-term learning out-
comes to examine how students develop their AI literacy under varying twenty-first
skills frameworks.
We are working to establish a community of researchers to disseminate good
practices and research findings in the countries/regions that promote AI literacy
education rigorously (e.g., the United States, China, Hong Kong). Through interna-
tional collaboration, we strive to share experience and locate comparable studies in
different implementation models, curricula, and practices so as to help policymak-
ers and frontline teachers to adopt effective approaches to maximize students’ learn-
ing outcomes. Comparative studies and analyses across countries/regions and/or
different education settings can deepen our understanding of how cultures and edu-
cation systems take part in AI literacy education that harness twenty-first skills
development.

11.6 What Is Next?

Educational professionals have to bear in mind that there is no perfect formula that
will work in all education systems. Every student is different. They grow differently
and possess capabilities and strengths which require varying learning support from
policies to classroom practices in light of their developmental and cultural diversi-
ties. Although it is not possible to transform our society digitally over one night, a
step-by-step approach can help make our world a better place to live in, and continu-
ous and tremendous efforts for leaders in different fields are necessary to implement
educational policies and practices that develop talents who suit their places. Digital
affordances and constraints of the existing education contexts and AI readiness are
diverse across the world.
References 149

This book serves as a launching point for AI literacy research in the field of
twenty-first skills education. It is possible that in the future more research will be
generated that targets new instructional design and theories and principles that
address effective design of learning technologies. Educational researchers, technol-
ogy developers, education policymakers, and educators should work together and
help ensure that student learning experiences are of a high quality. We hope to pro-
pose useful guidelines to conceptualize AI literacy education, identify reasons why
K–16 learners need such digital competence, its current landscape and situation
across academia and countries, and its theoretical basis to set for future research and
education. Further, the guidelines of pedagogy, content, technology, and assess-
ments used in AI literacy education are put forward to update AI literacy as an
important twenty-first-century skill.
As the authors of this book are preparing the book, we are living in a challenging
time during the pandemic. As educators, we are dedicated to addressing the rapid
changes in today’s digital world and preparing the next generation for tomorrow. A
good education should always think ahead of time and make people more intelligent
to cope with upcoming challenges. When we think about whether, why, and how to
incorporate AI into education, we can rethink Alan Turing (the father of AI)’s words
in 1950 that defines his famous Turing test: “A computer would deserve to be called
intelligent if it could deceive a human into believing that it was human.” Artificial
intelligence empowers machines and computers to imitate humans and perform like
us. It has great potential to become conscious and overcome humans. However,
humans make their world a better place to live in to address good values abstract
ideas like beauty and life and create new insights that are all irreplaceable by
machines. Therefore, on one hand, AI embraces a new vision toward education for
the future and a mindset open to positive change to empower our learners with the
knowledge, skills, and attitudes to thrive in an increasingly complex and fast paced
world. On the other hand, as educators, we need to rethink the purpose of education
critically. Should we merely prepare students for the job market and money? Or
should education prepare students to be good people, critical thinkers, lifelong
learners, responsible citizens, and moral and upstanding human beings?

References

Ahuja, V., & Nair, L. V. (2021). Artificial intelligence and technology in COVID Era: A narrative
review. Journal of Anaesthesiology, Clinical Pharmacology, 37(1), 28.
Boden, M. A., & Edmonds, E. A. (2019). From fingers to digits: An artificial aesthetic. MIT Press.
Baum, S. D. (2017). On the promotion of safe and socially beneficial artificial intelligence. AI &
SOCIETY, 32(4), 543–551.
Deeks, A. (2019). The judicial demand for explainable artificial intelligence. Columbia Law
Review, 119(7), 1829–1850.
Dwivedi, Y. K., Hughes, D. L., Coombs, C., Constantiou, I., Duan, Y., Edwards, J. S., et al.
(2020). Impact of COVID-19 pandemic on information management research and practice:
150 11 Summary and Conclusions

Transforming education, work and life. International Journal of Information Management,

55, 102211.
Jobin, A., Ienca, M., & Vayena, E. (2019). The global landscape of AI ethics guidelines. Nature
Machine Intelligence, 1(9), 389–399.
Long, D., & Magerko, B. (2020). What is AI literacy? Competencies and design considerations. In
Proceedings of the 2020 CHI conference on human factors in computing systems (pp. 1–16).
Ng, D. T. K., & Chu, S. K. W. (2021). Motivating students to learn AI through social networking
sites: A case study in Hong Kong. Online Learning, 25(1), 195–208.
Ng, D. T. K., Luo, W., Chan, H. M. Y., & Chu, S. K. W. (2022). Using digital story writing as a
pedagogy to develop AI literacy among primary students. Computers and Education: Artificial
Intelligence, 3, 100054.
PWC’S Global AI Report. (2022). Sizing the prize What’s the real value of AI for your business
and how can you capitalize? Retried from https://www.pwc.com/gx/en/issues/analytics/assets/
pwc-­ai-­analysis-­sizing-­the-­prize-­report.pdf
Reynolds, R. (2016). Defining, designing for, and measuring “social constructivist digital liter-
acy” development in learners: A proposed framework. Educational Technology Research and
Development, 64(4), 735–762.
Smith, D. F. (2015). 6 ways the new National Education Technology plan could help close
achievement gap. Ed Tech Magazine. Retrieved from http://www.edtechmagazine.com/
k12/article/2015/12/6-ways-new-nationaleducationtechnologyplan-could-help-close-
achievement-gap
Yang, S. J., Ogata, H., Matsui, T., & Chen, N. S. (2021). Human-centered artificial intelligence
in education: Seeing the invisible through the visible. Computers and Education: Artificial
Intelligence, 2, 100008.