See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/360695653

Tell Me a Story! Blending Digital Storytelling Into Marketing Higher Education

for Student Engagement

Article in Journal of Marketing Education · May 2022

DOI: 10.1177/02734753221090419

CITATIONS READS

17 186

3 authors:

Daniela Spanjaard Francine Virginia Garlin

University of Technology Sydney Western Sydney University
18 PUBLICATIONS 221 CITATIONS 11 PUBLICATIONS 376 CITATIONS

SEE PROFILE SEE PROFILE

Mohammed Hossain
Toronto District School Board
18 PUBLICATIONS 207 CITATIONS

SEE PROFILE

All content following this page was uploaded by Francine Virginia Garlin on 17 September 2024.

The user has requested enhancement of the downloaded file.

1090419
research-article2022
JMDXXX10.1177/02734753221090419Journal of Management EducationSpanjaard et al.

Article

Journal of Marketing Education

Tell Me a Story! Blending Digital

2023, Vol. 45(2) 167­–182
© The Author(s) 2022
Article reuse guidelines:
Storytelling Into Marketing Higher sagepub.com/journals-permissions
DOI: 10.1177/02734753221090419
https://doi.org/10.1177/02734753221090419

Education for Student Engagement journals.sagepub.com/home/jmd

Daniela Spanjaard1 , Francine Garlin1,

and Hossain Mohammed1

Abstract
Multimodal learning via the use of smart devices, online social interactions, and intuitive communication platforms are
fundamentally changing teaching and learning settings. Consequently, educators face unique challenges around student
engagement as learners increasingly look to the use of technology-enabled activities for meaningful collaboration. Within this
context, this research explores the role of digital storytelling in promoting perceived individual student and group engagement,
and how these interact with group functioning, in a postgraduate marketing subject. We evaluate the implementation of
this assessment as an avenue to optimize the benefits of the blended learning setting. It contributes to understanding by
incorporating the concept of collective, or collaborative engagement where there is a scarcity of research despite the
widespread application of group assessment in business education, and marketing education in particular. We employed
canonical correlation analysis as an exploratory technique to gain initial insight into its efficacy for student engagement. We
contend that digital storytelling shows promise as an inherently social and interactive learning task to provide an authentic
assessment for a range of marketing problems. In turn, these attributes provide a stimulating vehicle for student engagement
that can promote learning and satisfaction.

Keywords
digital storytelling, canonical correlation analysis, innovative teaching methods, learning approaches and issues, marketing
education issues, technology in classroom, student engagement, assessment, education administration issues

Learning environments are under a state of ongoing change perceptions of individual and group student engagement,
as technology continues to play an important role in provid- which in turn is supported by the functioning of the group.
ing new opportunities to encourage student participation We further investigate the possible influence of student
(Niemi & Multisilta, 2016). Multimodal learning via the use engagement with the DST activity on their beliefs about
of smart devices, online social interactions, and intuitive incremental learning and overall satisfaction with the
communication platforms (e.g., YouTube) are fundamentally quality of the subject.
changing teaching and learning settings as more institutions This paper contributes to theory by extending the context
move away from traditional lecture-style delivery. Educators in which the student engagement principle is applied, taking
face unique challenges around student engagement as learn- into consideration the role and influence of individual and
ers increasingly look to the use of technology-enabled schol- group interactions on student learning. While there are stud-
arly activities for meaningful collaboration, rather than just ies examining such theory at an institutional level, course
another way to access the material (Steen-Utheim & Foldnes, level (Mandernach, 2015), or from the perspective of an
2018). As such, emphasis on the integration of thoughtfully overarching teaching methodology such as blended learning
selected face-to-face and online tools to provide a more (e.g., Vaughan, 2014), this study considers student engage-
holistic learning experience persists (Garrison & Vaughan, ment in the context of a singular assessment task. It further
2007). contributes to understanding by incorporating the concept of
In this study, we evaluate the implementation of a digi-
tal storytelling (DST) assessment as an avenue to optimize 1
Western Sydney University, Australia
the benefits of the blended learning context, and to expose
Corresponding Author:
students to non-traditional methods of assessment design Daniela Spanjaard, School of Business, Western Sydney University, Locked
focused on social marketing theory. In doing so, we Bag 1797, Sydney 2751, Australia.
explore whether such a relatively novel design encourages Email: [email protected]
168 Journal of Marketing Education 45(2)

collective, or collaborative engagement (Järvelä et al., 2016) noting that the use of technology is only a means to an end
an area characterized by a dearth of research despite the (Bechter & Swierczek, 2017) not the main driver of learning
widespread application of group assessment in business edu- design.
cation generally, and marketing education in particular. According to Crittenden (2005), when putting this within
Bandura (1997) observed that group confidence toward its the context of a learning situation, such social activity tran-
ability to complete a task is connected to its ultimate success. spires where the “environment” is that created by the aca-
In addition, when the individuals who form that group share demic, and the “person” is the students who become
the same belief, there is a greater opportunity to manage any immersed in the activity so that their resultant behavior
challenges and produce anticipated results (Donohoo et al., becomes the overall student experience. In our study, the
2018). We consider the multidimensional approach by environment is positioned as the blended learning activity
Bowden et al. (2021) where social, cognitive, emotional, and based around digital storytelling, and where students interact
behavioral influences around student engagement provide an as individuals within the group to form collective engage-
anchor point for a positive learning experience. As such, we ment networks.
propose that attention to student engagement within the con- Research has shown that while this transition to blended
text of an innovative, singular group assessment task, and learning provides an ideal opportunity to enhance student
how students individually interact as part of group function- engagement (Sahni, 2019) often using an applied or experi-
ing, offers the promise of added utility to academics in the ential approach (Zhai et al., 2017), this in itself is not an indi-
pursuit of efficient assessment design. cator of achievement. Increased flexibility in learning can be
seen as a benefit, but it also increases the amount of self-
regulation needed on the part of the student (Boelens et al.,
Blended Learning 2017). To be successful, students may require better time
As noted by Krause (2007), blended learning occurs when management skills, abilities to effectively use technology,
there is an effective integration of different delivery modes, and self-efficacy to exercise control over their learning prac-
together with a diverse range of teaching models and varied tices (McDonald, 2014). Adaptable learning may work well
styles of learning which take a methodical approach toward for students who are high achievers, or for those who already
the application of technology, in combination with face to have such self-regulation abilities, but this may not be the
face collaboration. It is the design of the learning, not the case for those who don’t attain the same level of academic
tools that bring about a positive student experience. In this success (Owston et al., 2013). This means that while contem-
paper, we support the approach put forward by Garrison and porary teaching and learning environments provide new
Vaughan (2007) where blended learning aims to combine the opportunities to broaden their range of skills, the role of stu-
strengths of both online and classroom delivery to such an dent engagement becomes even more important.
extent that it broadens student learning in deep and meaning-
ful ways. If done successfully, blended learning combines Student Engagement, Collective Engagement,
the competencies of face to face and online teaching which
cannot generally be achieved if each is provided separately.
and Group Functioning
It requires the presence of the lecturer and students in the The concept of student engagement and its operationaliza-
classroom but also provides access to an e-learning platform, tion has been contested in the literature since the 1980s.
where students have control over how they manage their Efforts to consolidate divergent perspectives of student com-
learning time, where they learn, and the pace at which they mitment highlight the complexity of this multilevel, multidi-
learn. This, in turn, will influence the student’s learning mensional construct. Collectively, these works acknowledge
experience, engagement, and potentially overall achieve- contributions from the behavioral, psychological, sociologi-
ment (Sahni, 2019). cal, and holistic perspectives (Balwant, 2018; Christenson
The use of technology in blended learning models has et al., 2012; Kahu, 2013).
become the expected norm for those undertaking higher edu- Student engagement has been characterized by the time and
cation studies, and this was rapidly deployed during the effort students dedicate to activities that are empirically asso-
COVID19 pandemic. Yet even before this event, the incorpo- ciated with the required outcomes of a course of study and
ration of learning technologies had been growing in popular- what actions the educational establishment undertakes to
ity (Bliuc et al., 2007; Garrison & Vaughan, 2007; Serrano encourage them to participate in such endeavors (Kuh, 2009).
et al., 2019; Suwardy et al., 2013). Other influences that have Research studies using scales developed from this perspective
driven this change include the growing diversity of the stu- typically measure at the institutional level, which may assist
dent population, higher demand for a technologically skilled with broad policymaking, but is arguably less than useful for
workforce, and an increasing need for a flexible learning the curriculum developer (Eccles & Wang, 2012). More
environment (Serrano et al., 2019). Ideally, such a situation recently, Bowden et al. (2021, p. 1209) define student engage-
encourages increased student motivation and performance, ment as a “student’s positive social, cognitive, emotional, and
Spanjaard et al. 169

behavioural investments made when interacting with their concentrations of perceived incremental learning and overall
teaching institution and its focal agents (such as peers, employ- satisfaction with subject quality. Gray and DiLoreto’s (2016)
ees, and the institution itself).” However, these authors mea- study on the effects of student engagement on perceived
sured engagement at the global level, using scales designed to learning and satisfaction found strong, direct effects of stu-
gauge customer engagement (Hollebeek et al., 2014; Vivek dent engagement on both learning and satisfaction. More
et al., 2014) or ones measuring a very different conception of recently, Fisher et al (2021, p. 110) found that the blended
engagement (Schaufeli et al., 2002). learning context contributed to perceived engagement, per-
While definitional consensus remains elusive, the predomi- formance, and satisfaction, and that engagement “contributes
nant view suggests that the concept of student engagement has to feelings of satisfaction even without perceptions of
at least affective, cognitive, and behavioral dimensions (Kahu, improved performance.”
2013). Akin to Balwant’s (2018, p. 235) definition, we conceive
student engagement as “emotional, behavioural and cognitive
involvement in academic activities.” We use the C-OAR-SE
Authentic Assessment
methodology (Rossiter, 2011) to define constructs, which sub- In the context of the classroom, Parsons and Taylor (2011)
sequently guides operationalization. Student engagement is concluded that for effective engagement, learning needs to
accordingly assessed as being an abstract collective object hav- be relevant and real; the environment should include an
ing a collection of concrete constituent objects (academic activi- assorted range of technology; encourage “risk taking” activi-
ties); an abstract dispositional attribute with second-order ties; and promote peer to peer relationships between the aca-
cognitive, affective, and behavioral components, involving an demic and the students. However, to achieve this efficiently,
individual rater. This definition can be applied to any type of the alignment of blended learning tools and meaningful
learning task—group, individual, an in-class learning activity or assessments that measure learning outcomes is not without
an out-of-class assessment task. challenge. Other studies (Chapman, 2003; Fisher et al., 2021)
Most studies of student engagement focus on that of the focus on engagement being a balance of cognitive and emo-
individual student (Järvelä et al., 2016). Of special interest in tional commitment. Students will adapt their degree of
this research is the concept of collaborative engagement as it engagement as a means to satisfy their goals (Yorke, 2006)
may be apparent in a group learning task; and how collabora- choosing what to participate in, contribute to and complete,
tive engagement may in turn influence individual engage- to enhance their learning experience (Kinzie & Gonyea,
ment. Research on group learning “has shown that individual 2009; Steele, 2015). Ultimately, deep learning practices will
students’ interpretations can be positive, leading to increased encourage engagement (Hockings et al., 2008) so that stu-
motivation and engagement in group activities; and, alterna- dent-teacher interaction, therefore, becomes a fundamental
tively, that individual students’ interpretations can be nega- component where the actions of the academic remain a cen-
tive and lead to demotivation and withdrawal” (Järvelä et al., tral focus to high student commitment (Zepke et al., 2014).
2016, p. 40). Such investigations suggest there is evidence to Kuh (2009) proposes that there are two major aspects to
expect a relationship between individual and collaborative effective student engagement through learning activity
engagement; and, that the effects of such a relationship may design. First, those undertakings which are completed in-
be subject to students’ perceptions. class are balanced by those that are more suited to be done as
We further suggest that perceptions of group functioning co-curricular or out-of-class. Both activities are guided by
may play a significant role in collaborative engagement. the academic managing the delivery of the subject. This
Group functioning is defined here as the application of inter- leads to the importance of assessment and its associated
dependent skills to achieve a predetermined common goal. design. Assessment construction and how this integrates
The object of this construct is concrete in the context of the within the classroom experience is pivotal to students’ desire
production of the DST artifact. It further contains a set of to engage in their learning, and yet can be a complex process
specific perceptual attributes to represent interdependent as tasks attempt to meet the needs of individual, departmen-
skills such as cooperation and organization. We were unable tal, and institutional requirements (Bearman et al., 2017).
to find any significant body of work that considers the role of According to Carless (2015), design is one of the most essen-
group functioning as an influencing factor on student engage- tial features of assessment practice, and yet can be the most
ment, yet we speculate that group functioning is an important problematic particularly for those that follow a non-tradi-
inclusion in the context of an assessable group task. Good tional format (Postareff et al., 2012), especially when ensur-
working relations within teams undoubtedly contribute to ing students understand the purpose of the assessment and
engagement at both the individual and group level, and this how it contributes to their learning.
potential effect makes group functioning a worthy addition Authentic assessment has the potential to make a valuable
to the conceptual framework of the study. contribution to the learning experience through designs that
The conceptualization extends to the proposition that facilitate student engagement. Using a blend of everyday
individual and collective engagement may lead to higher life, where knowledge can be applied analytically to solve a
170 Journal of Marketing Education 45(2)

problem (Villarroel et al., 2018), authentic assessment can will also inspire the less involved students by stimulating
add to the quality of student experience (Dochy & McDowell, their emotional engagement with the topic (Suwardy et al.,
1997) and allows them to develop employability skills 2013). This form of learning was selected as the preferred
around problem-solving, critical thinking, communication, assessment method for this study as it was determined to be
and teamwork (Singh et al., 2014). Utilizing such techniques effective in capturing not only group functioning as students
provides the opportunity to open up the possibility for better designed, built, and presented their digital story, but it would
student engagement (Vaughan, 2014). We propose that with also provide insights into how the group and individuals
a carefully designed assessment, taking an authentic engaged with their team members, the learning materials,
approach, DST provides students with the ability to show- and assessment requirements.
case their narrative and insights into a social marketing issue. Generally, DST uses multimedia resources which, in a
learning environment, tends to be produced by students to
provide an original artifact on a topic that is important to
Digital Storytelling
them (Clarke & Adam, 2012). With the gradual integration
Storytelling is an efficient way to convey information, of blended learning techniques into the classroom, adapting
beliefs, and traditions (Suwardy et al., 2013). It is an influen- storytelling into a digital format is a natural progression.
tial and lasting method of communication where the blend of According to Wang and Zhan (2010), the application of
verbal and visual educates, informs, and entertains an audi- DST enables the amalgamation of student engagement,
ence. A “good” story should make the creation by the story- deeper learning, project-based knowledge, and the effective
teller, and the story interpretation by the receiver, believable use of technology to provide a more meaningful outcome.
and enjoyable (Van Laer et al., 2014). It has been recom- Moreover, DST fits well with the notion of a learning task
mended that storytelling and other forms of narrative trans- where the problem covers numerous resolutions that require
portation represent a potent mechanism to frame an event learners to connect material from a range of sources, thus
which in turn can be experienced by others (Van Laer et al., necessitating student engagement with the learning process
2014). Such storytelling has been used extensively within toward achieving a meaningful solution (Järvelä et al.,
consumer research and marketing (McDougal et al., 2021; 2016). DST moves students from being passive to active
Shankar et al., 2001; Woodside, 2010). Stories around con- learners and has been used across of range of disciplines
sumer decision making and consumption experiences are including health (Moreau et al., 2018), communication
also not uncommon in academia (Shankar et al., 2001) and (Wang & Zhan, 2010), cultural studies (Burgess, 2006) and,
research has shown that storytelling can be a valuable educa- of course, education (Schmoelz, 2018). Surprisingly, there
tional device aimed at encouraging interaction between the is less research on the use of DST in marketing education
storyteller and the audience to interpret their own perspec- (Greene et al., 2015). While more recent research has used
tives and experience (Sarris, 1990). digital storytelling within the context of digital marketing
Alterio and McDrury (2003) take a slightly different (Fusté-Forné, 2021), tourism marketing (Bassano et al.,
approach and suggest that when embedding storytelling into 2019), and marketing communications (Fernandes, 2019),
a higher education context, students often end up attaching there are still limited studies on the application within mar-
the story with their own understandings and thus bring about keting higher education, where it would seem to be a natu-
a profound learning experience. Those who partake in creat- ral fit (McDougal et al., 2021). We propose that using DST
ing digital stories acquire increased communication skills as provides an opportunity for educators to design assess-
they express their ideas and construct narratives. By sharing ments that challenge students’ creative abilities outside the
their creations with their peers, they gain valuable under- standard written format.
standing in being able to critique their own and others’ work. In this paper, attention will be given to how the manage-
This in turn develops emotional intelligence and social learn- ment of an authentic assignment utilizing in-class activities
ing (Robin, 2008, 2016). balanced by an out-of-class assessment within a blended
Digital storytelling has been defined as a brief narrative learning subject has resulted in higher student engagement in
presented as a short movie (Davis, 2004) where the use of a postgraduate subject. This is important because research in
technology and digital imagery projects the story of the this area, while growing, generally tends to focus on student
author (Porter, 2005). The product may be instructional, per- engagement, transition, and retention in first-year undergrad-
suasive, historical, or reflective (Wang & Zhan, 2010). DST uate courses (Nelson & Clarke, 2014; Zepke, 2013). Less has
encourages students to organize and express their opinions in been covered about how the implementation of blended
a meaningful way. For business students, it can help them to learning specifically encourages learner engagement at more
develop reasoning and critical thinking in an organizational advanced levels which at the same time, provides greater
setting (Greene et al., 2015; Lukin, 2016). Consequently, stu- opportunity for students to do well in their grades. The exam-
dents are more likely to be absorbed in their learning. In ple presented here seeks to understand how postgraduate
addition, it has been suggested that successful storytelling business students respond in this context.
Spanjaard et al. 171

This study explores the role of DST in promoting both The Assessment Task
perceived individual student and group engagement in a
postgraduate marketing subject. Accordingly, the study con- The group assessment task required students to develop a
tributes to a largely neglected segment of higher education short movie (5–6 minutes) around the social impact of mar-
research, where the focus of student engagement has cen- keting on society. This included aspects of sustainability,
tered on the transition and retention of first-year undergradu- social awareness, marketing ethics, and social responsibility.
ate courses. In addition, few studies have explored the As part of the assessment, each group was to provide a con-
relationship between the perceptions of individual and group ceptual tale and translate this into a storyboard before the
engagement within a specific assessment task; nor has there final submission. The ultimate digital story was to be devel-
been consideration given to the role of group functioning in oped in any broad-based software (e.g., iMovie, Canva,
the process. Adobe Spark) which the students were familiar with. Those
who needed assistance were provided with online tutorials
on how to use software to design their story via the e-learn-
Research Questions ing student system. Students were also provided with dedi-
While this study aims to primarily investigate the capacity of cated class time to work through their story in addition to
a novel, group assessment task to elicit individual and col- being provided with online learning materials which they
laborative engagement within this, it also presents further could access during their own time. This was to ensure that
opportunity to understand the relationship between individ- adequate application of marketing theory was applied to the
ual and collaborative engagement, the role of group func- story, and their focus was not just on the use of the software.
tioning, and the subsequent influence of these constructs on The majority were comfortable in applying the technology
incremental learning and overall student satisfaction with given their familiarity with social media and other online
subject quality. Accordingly, three research questions arise: platforms (e.g., YouTube) particularly as they had the free-
dom to choose the tool most relevant to their skill sets. They
Research Question 1 (RQ1). To what extent does the were permitted to use open access resources, or they could
DST assessment task stimulate student engagement? create their own material via animation or utilize team mem-
Research Question 2 (RQ2). Is there a relationship bers to generate content. Teaching staff encouraged students
between individual student engagement, collective stu- to consider how their group could work together to create a
dent engagement, and group functioning? digital artifact that effectively communicated their chosen
Research Question 3 (RQ3). Do perceptions of student topic and ideally to tap into the emotions of the viewer to
engagement and group functioning contribute to percep- increase awareness of the social issue. All digital stories
tions of incremental learning, overall student satisfaction required a hyperlink to be embedded in their final submis-
with the subject’s quality? sion. Students were graded on how persuasive their story
was, their analysis of the social marketing situation, their
recommendations to address this and how well they commu-
Method nicated with the audience.
The task suited the expected knowledge befitting a post-
The Context graduate cohort on the basis that digital storytelling requires
The DST assessment task was trialed in a postgraduate mar- a relatively high level of creative thinking and engagement
keting subject that aimed to teach students the principles of translated into visual texts, audio recording, and role-playing
marketing systems, and the social impact these have on com- (Robin, 2008). While the task fits the premise around authen-
munities. While this subject is primarily designed for stu- tic assessment, it should not be mistaken as another alterna-
dents undertaking postgraduate study, it also forms part of tive for experiential learning. This tends to occur where there
the Business Administration stream for those who are inter- is an opportunity to replicate “real-world situations, build the
ested in marketing but do not want to commit to an entire formation of abstract concepts and if possible, test new situ-
degree. Thus, there is a mix of students within the cohort ations” (A. Y. Kolb & Kolb, 2005; D. A. Kolb, 1984). DST
other than those training to be marketing specialists. It was entails setting ideas and thoughts into digital themes as well
deemed appropriate that students should have experience of as forming a coherent script to enhance relevant theoretical
tertiary study so they would grasp the complexities of pro- concepts, thus making it primarily an abstract conceptualiza-
posed work, but that it was unlikely they had been exposed to tion of authentic scholarship.
the nuances of DST. Students were introduced to the concept The format of the assessment aimed to minimize the com-
of the storytelling task via a teacher-produced story based mon practice where students divide the group reports into
around a social marketing topic, which they were required to “sections” for individual completion that would be combined
replicate as a group assessment. To encourage creativity, just before submission. This tactic of the partitioning of work
they were able to choose their own theme upon which to minimizes face-to-face interaction and reduces critical team-
build their “story.” building and communication skills. By putting students into
172 Journal of Marketing Education 45(2)

a situation where team collaboration is crucial to the success well they believe their team functioned as a group to produce
of high-level creative thinking, the goal of enhanced learner the storytelling artifact on the same 8-point scale. Two
engagement was more likely. Each submission was to be pre- dependent measures—a measure of incremental, “new” per-
sented back to the class in conjunction with an explanation of ceived learning in the subject and overall satisfaction with
their project topic. The digital stories were also uploaded to the quality of the subject—were included using a 4-point
the class online site for sharing and online discussion. Some ordinal scale (see Bass et al., 1974).
examples of student stories include the environmental impact The learning construct was defined as having a concrete
of plastic packaging, raising awareness of domestic violence, object (the subject), and concrete perceptual attributes.
promoting support services for the homeless and the dark Being “doubly-concrete,” a single-item measure is suffi-
side of disposable fashion. cient using this methodology (Bergkvist & Rossiter, 2007;
Rossiter, 2011. See also Morrison et al., 2006). In line with
Bacon’s (2016) call, we explicitly measure student learning
The Research Instrument as perceived and not actual learning but include an indicator
Items were developed to measure perceptions of individual of actual learning in the analysis by assessing differences in
engagement, collaborative engagement, and group function- students’ subject marks pre and post the DST assessment
ing with respect to the DST task as independent measures. task. Measuring perceived learning is appropriate given the
Given the paucity of established measures of student engage- context of authentic assessment, wherein students must
ment at the level of a singular learning activity, the items appreciate how a task contributes to their learning.
were informed by the results of a separate qualitative study For the measure of student satisfaction with the quality of
of the target sample. This involved running a series of three the subject, we used a single global, “overall” measure
focus groups (n = 36) attended by students who had partici- (Fisher et al., 2021). While the learning construct was also
pated in the DST assessment. The focus groups were held at defined with concrete properties, the measure also replicates
the conclusion of three teaching periods and after the final that used in the University’s student feedback surveys, which
grades for the subject had been released. Students were asked provides the benefit of student familiarity with the item, tak-
to self-report their understanding of digital storytelling, and ing advantage of well-formed attitudinal schemata.
its implications for engagement and sharing of ideas with Finally, demographic data were collected on gender, age,
other students, particularly within the context of group func- and international student status. The questionnaire was
tioning, and noting their self-efficacy around collaborative administered at the beginning of class, where consenting stu-
interaction. dents were assured of the voluntary and anonymous nature of
In line with the C-OAR-SE procedure,1 scale develop- their participation.
ment was based on construct definitions, with student
engagement scales parsimoniously constructed to contain
Results
concrete measures of its cognitive, affective, and behavioral
constituents. The items were informed by the qualitative The final sample consisted of 154 responses, of which 55.8%
results while the literature further substantiated the determi- were male. Some 72.8% identified as international students.
nation of cognitive, affective, and behavioral items (Balwant, Of those students nominating their country of origin, the
2018; Kahu, 2013; Wang et al., 2016). For instance, the cog- majority were from South-Asian countries (60.4%), mostly
nitive measures such as “I tried to connect other things . . . to India (68.9%). The largest age group belonged to the 25 to 29
the DST activity,” “I tried to think about different ways to category (43.1%) with 91.5% of the sample indicating they
develop the DST,” and the behavioral measure “I discussed were under 35 years of age. To preserve the sample size, a
the DST with others at every opportunity” accorded with small proportion of missing values (<1%) were replaced
associated items in Wang et al.’s (2016) engagement scales. with pooled multiple imputation values based on linear
While items like “My group enjoyed working on the DST regression prior to analysis.
activity”; “Working on the DST activity made me feel more We employed canonical correlation analysis (CCA) as an
anxious than other assessments in the subject” are well sup- exploratory technique to gain some initial insight into the effi-
ported as affective measures of engagement (see Kahu, cacy of DST for eliciting student engagement. CCA provides
2013), these items also reflect key themes emerging from the a means to examine the relationship between multiple depen-
qualitative data. dent and predictor variables simultaneously, and as such
Students were explicitly asked to compare the DST enables the control of Type I errors arising from separate
assessment task with other group projects they had done on analyses using single dependent variables (Thompson, 1991).
items using an 8-point scale from 0 (“Never”) to 7 (“Always”). Other multivariate techniques are considered to be the lineage
For example, “Compared to other group projects I’ve done, I of canonical correlation (Attaway et al., 1987), and while its
tried to connect things I have observed in my everyday life to complexity is rewarded by its capacity to reveal elaborate
the DST activity.” Subsequently, students also rated how interrelationships between variables of interest (Sherry &
Spanjaard et al. 173

Henson, 2005), this complexity along with inconsistencies in (R_c), which are simply Pearson’s correlations between the
the way that it has been reported in the literature has arguably variates in the function, hence the squared canonical correla-
resulted in a generally patchy application of the technique tions (R_c^2) denote the shared variance between the sets,
(Engellant et al., 2016; Pugh & Hu, 1991). However, as inter- akin to R2 in multiple regression (Sherry & Henson, 2005).
est lies in gaining some preliminary understanding of the In this case, the squared canonical correlations were .543,
interplay between several predictor and criterion variables— .393, .321, .196, .111. .037, .028, and .006, respectively.
in this study the relationships between individual engage- The criteria used to determine the number of functions to
ment, group engagement and group functioning—canonical interpret is based on the significance of the functions
correlation analysis remains the most appropriate multivari- extracted and the size of the canonical correlation (Hair,
ate technique. 2009). Although the first five functions were significant,
A series of three canonical correlation analyses were con- using the recommended threshold (Sherry & Henson 2005),
ducted to examine the relationship between (a) individual only the first three functions have canonical correlations
and group engagement; (b) individual engagement and group above .45. The results for the three functions are reported in
functioning; and (c) group engagement and group function- Table 2.
ing. While CCA has been considered a large sample proce- In reference to Function 1, individual engagement with the
dure (Boedeker & Henson, 2020), there are several examples DST activity was primarily explained by enjoyment and
of studies employing the procedure with smaller samples interaction with others through discussion about the task. The
(Dos Santos & Brandi, 2014; Mahmood & Mann, 1993; Van finding is supported by the magnitude of the function coeffi-
Auken et al., 1993) including those within a higher education cients. Secondary contributions arise from sharing ideas and
context (Engellant et al., 2016; Kim & Shin, 2021; Pugh & the perceived individual effort applied to the exercise. Fan
Hu, 1991; Sahin et al., 2013). However, sample size suffi- and Konold (2010) argue that the small value of the standard-
ciency is important for CCA, and a minimum threshold of 10 ized canonical function coefficient (Coef) suggests that the
observations per variable is recommended (Hair, 2009), contribution from these variables is shared with other vari-
which is approximated at 9.6 for the current study. Although ables in the set due to collinearity. In such situations,
canonical correlation analysis is comparatively robust Thompson (1984) recommends interpretation be based on the
against its statistical assumptions (Hair, 2009), the minimum structure coefficients (rs). The common sign of each of the
sample size together with departures from multivariate nor- variables in the criterion variate (in this case, all negative)
mality and linearity requires a cautious interpretation of the indicates that they are positively related.
results. As for the predictor variate, perceptions of group engage-
Table 1 presents the descriptive statistics and correlations ment with the DST task are most notably explained by group
for the study variables. Overall, the mean results indicate that discussion of the activity, while the other variables with high
students perceive moderately high rates of individual and structure coefficients (rs) are subject to collinearity hence its
group engagement and levels of group functioning. Despite contribution to the variate is overlapping with others in the
the number and degree of significant correlations in the set. The same sign of the structure coefficients between the
matrix, collinearity tests did not indicate multicollinearity of criterion and predictor variates means that each of the group
concern. engagement variables contributes positively to individual
The eight group engagement variables were stipulated as engagement with the DST activity.
predictors of the eight individual engagement variables to The results relating to individual and perceived group
evaluate the multivariate shared relationship between the anxiety about the task were notable. Being reverse-coded,
two sets. CCA depicts the relationship between the two sets lower scores on these measures indicate higher levels of anx-
of variables by producing independent canonical functions, iety associated with the assessment task (refer to Table 1).
maximizing the correlation between the linear composites, or The variables did not feature in the Function 1 results but did
“variates,” in the process. The number of canonical functions feature as providing a substantive contribution to Functions
extracted by the analysis equals the number of variates in the 2 and 3 by the relatively high values of both function and
smallest variate. In this case, the number of variables is the structure coefficients. Anxiety is the primary explanation for
same for each set, hence eight functions were produced. student engagement overall in Function 2, and there is a posi-
Based on Wilks’ lambda, the full model was significant with tive relationship between individual and group anxiety. This
λ = .125, F(64, 802.46) = 5.435, p < .001. Sherry and means that students perceiving group anxiety about the DST
Henson (2005, p. 48) explain that since Wilks’ λ “represents task will also be likely to experience individual anxiety. For
the variance unexplained by the model, 1 – λ yields the full Function 3, perceptions of group anxiety impact individual
model effect size in an r2 metric.” For the set of eight canoni- anxiety, but this is inversely related to individuals’ likelihood
cal functions, the r2 type effect size is .875. That is, the of making connections with everyday observations. This
model explained 87.5% of the variance shared between the finding suggests that, for some students, anxiety has the
two variates. The results produce canonical correlations potential to impact other dimensions of individual
 Table 1. Correlations of Total Sample.

174
Variable M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Individual engagement
   1. I tried to connect other things I have 5.32 1.37
learned in other subjects to the DST
activity (I-CONNECT)
   2. I tried to connect things I have 5.56 1.35 .40**
observed in my everyday life to the
DST activity (I-OBSERVE)
   3. I tried to think about different 5.40 1.48 .25** .42**
ways to develop the DST activity
(I-DIFFWAYS)
   4. I enjoyed working on the DST activity 5.64 1.45 .16* .29** .54**
(I-ENJOY)
   5. Working on the DST activity made 3.58 2.23 −.02 .02 −.06 −.31**
me feel more anxious than other
assessments in the unit (I-ANXIOUS)a
   6. I shared ideas and information about 5.85 1.40 .18* .32** .33** .40** −.21*
DST with other members in my group
(I-SHAREINFO)
   7. I discussed DST with others at every 4.75 1.91 .26** .32** .40** .49** −.06 .39**
opportunity (I-DISCUSS)
   8. I put a lot of effort into the DST 5.53 1.32 .26** .29** .39** .38** −.12 .50** .50**
activity (I-EFFORT)
Group engagement
   9.My group tried to connect other things 5.20 1.55 .24** .39** .22** .44** −.07 .32** .28** .33**
that have been learned in other units
to the DST activity (G-CONNECT)
10. My group tried to connect other 5.51 1.37 .32** .42** .31** .52** −.05 .35** .37** .29** .71**
things that were observed in everyday
life to the DST activity (G-OBSERVE)
11. My group tried to think about 5.63 1.30 .06 .20* .31** .55** −.14 .31** .25** .35** .51** .50**
different ways to develop the DST
activity (G-DIFFWAYS)
12. My group enjoyed working on the 5.51 1.52 .11 .19* .25** .61** −.19* .33** .43** .34** .54** .56** .70**
DST activity (G-ENJOY)
13. Working on the DST activity made my 3.30 2.21 −.10 −.02 −.04 −.18* .57** −.08 −.06 −.07 −.11 −.11 −.10 −.20*
group more anxious (G-ANXIOUS)a
14. My group members shared ideas and 5.67 1.36 .12 .22** .29** .57** −.18* .42** .40** .40** .54** .55** .66** .70** −.20*
information about the DST activity
(G-SHAREINFO)
15. My group discussed the DST activity 5.10 1.65 .14 .18* .32** .52** −.02 .31** .61** .45** .51** .59** .64** .67** −.07 .60**
at every opportunity (G-DISCUSS)
16. My group put a lot of effort into the 5.77 1.30 −.05 .14 .18* .56** −.12 .22** .37** .23** .56** .50** .63** .65** −.08 .62** .61**
DST activity (G-EFFORT)
Group functioning
17. The group functioned cooperatively as 5.76 1.31 .09 .13 .20* .51** −.15 .23** .25** .19* .46** .43** .64** .71** −.13 .58** .55** .60**
a team (COOP)
18. The group was organized (ORGAN) 5.61 1.44 .12 .15 .20* .53** −.17* .29** .42** .27** .51** .45** .52** .72** −.18* .59** .57** .61** .79**
19. The group completed tasks in a timely 5.49 1.49 .07 .22** .19* .50** −.17* .28** .34** .26** .54** .46** .60** .71** −.08 .58** .56** .59** .75** .79**
way (TIMELY)
20. The group adjusted to changes or 5.76 1.31 .06 .07 .18* .40** −.27** .26** .26** .20* .41** .47** .52** .62** −.17* .43** .46** .45** .65** .62** .65**
challenges (ADJUST)

a
Reverse coded.
*Correlation is significant at the .05 level (2-tailed). ** Correlation is significant at the .01 level (2-tailed).
Spanjaard et al. 175

Table 2. Canonical Solution for Group Engagement Predicting Individual Engagement.

Function 1 Function 2 Function 3

2 2
Variable Coef rs rs (%) Coef rs rs (%) Coef rs rs 2 (%) rs 2 h2 (%)
I-Connect 0.04 −0.20 3.92 0.18 0.22 4.62 −0.21 −0.30 9.00 0.09 17.54
I-Observe 0.00 −0.32 10.43 0.37 0.33 10.56 −0.45 −0.52 26.73 0.27 47.72
I-Diffways 0.20 −0.44 19.01 −0.09 0.12 1.39 0.05 −0.15 2.37 0.02 22.77
I-Enjoy −0.71 −0.86 73.27 0.44 0.39 15.52 −0.23 0.02 0.03 0.00 88.83
I-Anxious 0.19 −0.08 0.67 0.61 0.71 50.27 0.77 0.67 44.36 0.44 95.30
I-Shareinfo −0.09 −0.51 26.11 0.12 0.24 5.95 −0.34 −0.22 4.67 0.05 36.73
I-Discuss −0.45 −0.81 65.61 −0.71 −0.31 9.42 0.60 0.22 4.71 0.05 79.74
I-Effort −0.15 −0.58 33.64 −0.05 0.02 0.03 −0.02 −0.06 0.34 0.00 34.00
Rc 2 54.30 39.30 32.10
Adequacy 29.0 12.22 11.53
G-Connect 0.06 −0.61 37.33 0.20 0.36 13.18 −0.42 −0.38 14.44 0.14 64.95
G-Observe −0.16 −0.71 50.69 0.36 0.34 11.56 −0.57 −0.40 16.16 0.16 78.41
G-Diffways 0.18 −0.67 45.43 0.60 0.36 13.10 −0.31 −0.11 1.21 0.01 59.74
G-Enjoy −0.30 −0.85 71.40 0.06 0.27 7.02 0.17 0.08 0.67 0.01 79.10
G-Anxious 0.09 −0.07 0.44 0.58 0.66 42.90 0.71 0.65 42.77 0.43 86.11
G-Shareinfo −0.27 −0.80 64.48 0.04 0.29 8.41 −0.11 −0.03 0.11 0.00 73.00
G-Discuss −0.50 −0.91 82.08 −0.97 −0.19 3.57 0.48 0.10 0.96 0.01 86.62
G-Effort −0.17 −0.76 58.37 0.03 0.15 2.31 0.43 0.11 1.21 0.01 61.89
Adequacy 51.28 12.76 9.81

Note. Coef = standardized canonical function coefficient; rs = Structure coefficient; rs 2 = squared structure coefficients; h2 = communality coefficient; Rc 2
= squared canonical correlations. Structure coefficients (rs) greater than |.45| underlined. Communality coefficients (h2) greater than 45% underlined.

engagement. It is also noteworthy that the anxiety items were engagement variables positively added to the explanation of
among the highest communality values in the model (95.30% group engagement. For the predictor variate, all group func-
individual and 86.11% group), indicating a high degree of tioning variables contributed to its variate and were posi-
utility of these variables for the solution. tively related to group engagement, most notably from the
The adequacy coefficients included in Table 2 can be used timely completion of group tasks, but also from cooperation
as indicators of the overall representativeness of the canoni- and organization within the group.
cal variates (Boedeker & Henson, 2020). In this instance, the The final canonical correlation explored the relationship
Group Engagement variate in Function 1 has the largest ade- between individual engagement and group functioning. The
quacy coefficient at 0.5128, meaning that, on average, model was significant, λ = .537, F(32, 525.27) = 3.010, p <
51.28% of the variability in the original variable set is cap- .001, equating to a comparatively lower shared variance of
tured by the variate. The remaining adequacy percentages 46.25%. Squared canonical correlations for each function
are relatively weak indicating that some variables are better were .351, .085, .063, and .035. Only the first function was
represented than others as is the case for the anxiety vari- significant, with the results provided in Table 4.
ables in Functions 2 and 3, for example. In this model, individual enjoyment of the assessment
The second canonical correlation addressed the question task is the primary explanation for the individual engage-
of whether the four-item set of group functioning variables, ment variate. Although sharing information with the group,
in turn, influenced group engagement. The full model was discussing the assessment and individual effort also posi-
significant, λ = .259, F(32, 525.27) = 7.259, p < .001, with tively contribute to the variate, they have relatively low
a corresponding effect size of 74.1%. Squared canonical cor- structure coefficients. For the group functioning variate,
relations were .657, .149, .083, and .032, respectively. The group organization is by far the most important variable to its
first two functions were significant, but only the first func- explanation, followed by the timely completion of tasks and
tion was interpreted based on canonical correlation cooperation. Group adjustment to change/challenges pro-
coefficients. vided the least explanation. The adequacy coefficient sug-
As reported in Table 3, the solution to this model showed gests that variate is a good representation of its underlying
that the perception of group enjoyment of the task is the pri- variables at 74.13%. The results indicate that group function-
mary contributor to the group engagement variate. With the ing has a positive overall influence on individual engage-
exception of group anxiety, perceptions of all other group ment with the DST task.
176 Journal of Marketing Education 45(2)

Table 3. Canonical Solution for Group Functioning Predicting Group Engagement.

Variable Coef rs rs 2 (%)

G-Connect −0.16 −0.68 46.24
G-Observe 0.07 −0.62 38.44
G-Diffways −0.12 −0.80 64.00
G-Enjoy −0.62 −0.96 92.16
G-Anxious −0.00 −0.18 3.24
G-Shareinfo −0.05 −0.78 60.84
G-Discuss −0.07 −0.75 56.25
G-Effort −0.19 −0.80 33.64
Rc 2 65.70
Adequacy 49.35
G-Coop −0.34 −0.92 84.64
G-Organ −0.26 −0.91 82.81
G-Timely −0.37 −0.92 84.64
G-Adjust −0.14 −0.76 57.76
Adequacy 77.46

Note: rs = Coef = standardized canonical function coefficient; Structure coefficient; rs 2 = squared structure coefficients; Rc 2 = squared canonical
correlations. Structure coefficients (rs) greater than |.45| underlined.

Table 4. Canonical Solution for Group Functioning Predicting Individual Engagement.

Variable Coef rs rs 2 (%)

I-Connect 0.00 −0.18 3.24
I-Observe −0.03 −0.30 9.00
I-Diffways 0.31 −0.35 12.25
I-Enjoy −0.87 −0.93 86.49
I-Anxious −0.02 −0.31 9.61
I-Shareinfo −0.12 −0.51 26.01
I-Discuss −0.31 −0.68 46.24
I-Effort −0.04 −0.47 22.09
Rc 2 35.08
Adequacy 26.87
G-Coop −0.07 −0.85 72.25
G-Organ −0.67 −0.98 96.04
G-Timely −0.27 −0.89 79.21
G-Adjust −0.06 −0.70 49.00
Adequacy 74.13

Note: Coef = standardized canonical function coefficient; rs = structure coefficient; rs 2 = squared structure coefficients; Rc 2 = squared canonical
correlations. Structure coefficients (rs) greater than |.45| underlined.

To determine whether individual, group engagement and analysis for “Learning” and “Satisfaction” are presented in
group functioning, in turn, influenced perceptions of their Tables 5 to 7, and 8 to 10, respectively.
incremental learning within the subject (“Learning”), and Each model significantly predicted low-high classifica-
satisfaction with the quality of the subject (“Satisfaction”), tion above that of their baseline counterparts. While the
each set of variables were subject to a series of binomial Nagelkerke pseudo-R² may be considered weak to moderate
logistic regressions, with the outcome measures of satisfac- in value (Garson, 2016), the significance of the Hosmer and
tion and learning recoded into two Low-High groups. Again, Lemeshow Test suggests the models were not a poor fit, pre-
the issue of sample size is acknowledged with the use of dicting between 85.7% and 90.9% of cases. In addition, to
logistic regression, although its application with small sam- assess the ability of the models to discriminate between
ple sizes is not unprecedented in marketing education studies cases, receiver operating characteristic curve analysis was
(Basil & Basil, 2008; Dikcius et al., 2020). The results of the conducted. The areas under the curve were significant and
Spanjaard et al. 177

Table 5. Influence of Individual Engagement Variables on Learning.

Variable β SE Wald Sig. Exp (β)

I-Connect −0.56 0.28 3.961 .05 0.57
I-Observe 0.20 0.22 0.792 .37 1.22
I-Diffways 0.02 0.20 0.007 .93 1.02
I-Enjoy 0.44 0.21 4.433 .04* 1.55
I-Anxious 0.18 0.14 1.690 .19 1.19
I-Shareinfo −0.43 0.29 2.173 .14 0.65
I-Discuss 0.30 0.16 3.624 .06 1.35
I-Effort 0.16 0.27 0.357 .55 1.18
Constant 1.40 1.90 0.55 .46 4.06

Note. Dependent variable coded 0 = Low learning and 1 = High learning.

R² = 0.11 (Cox & Snell R²), 0.21 (Nagelkerke). Model χ²(1) = 17.44, p = .03. Hosmer and Lemeshow. χ² = 4.13, p = .85. Overall percentage correct = 87.7%
*Significant at p < .05.

Table 6. Influence of Group Engagement Variables on Learning.

Variable Β SE Wald Sig. Exp (β)

G-Connect 0.49 0.32 2.46 .12 1.64
G-Observe −0.34 0.35 0.93 .34 0.72
G-Diffways 0.20 0.41 0.24 .63 1.22
G-Enjoy −0.33 0.35 0.90 .34 0.72
G-Anxious 0.12 0.18 0.42 .52 1.12
G-Shareinfo −1.08 0.45 5.70 .02* 0.34
G-Discuss 1.04 0.30 11.97 .00* 2.82
G-Effort 1.15 0.37 9.44 .00* 3.16
Constant −2.84 1.97 2.09 .15 0.06

Note. Dependent variable coded 0 = low learning and 1 = high learning.

R² = .26 (Cox & Snell), 0.51 (Nagelkerke). Model χ²(1) = 47.00, p = .00 Hosmer and Lemeshow. χ² = 8.97, p = .35. Overall percentage correct = 90.3%
*Significant at p < .05.

Table 7. Influence of Group Functioning Variables on Learning.

Variable Β SE Wald Sig. Exp (β)

G-Coop 0.01 0.29 0.00 .97 1.01
G-Organ 0.46 0.30 2.33 .13 1.58
G-Timely 0.19 0.28 0.44 .51 1.20
G-Adjust −0.15 .25 0.35 .55 0.86
Constant −0.50 0.98 0.26 .61 0.61

Note. Dependent variable coded 0 = low learning and 1 = high learning.

R² = .08 (Cox & Snell), 0.15 (Nagelkerke). Model χ²(1) = 12.20, p = .02 Hosmer and Lemeshow. χ² = 5.37, p = .72. Overall percentage correct = 88.3%.

ranged from acceptable to outstanding (.72 to .92) based on classification. We can only speculate that this finding may be
the rules of thumb suggested by Hosmer et al. (2013). due to the quality or quantity of ideas/information from oth-
The findings for Learning showed significant relation- ers was not perceived to contribute to incremental individual
ships for individual enjoyment of the task (Table 5), whereby learning per se. None of the group functioning variables was
increasing levels of enjoyment were predictive of the high predictive of the low-high learning classifications (Table 7).
learning group. Similarly, perceived high levels of group dis- As to satisfaction with the overall quality of the subject,
cussion and effort applied to the DST task were associated increases in individual enjoyment (Table 8), perceived group
with the high learning group (Table 6), but group members’ effort (Table 9), and the group’s timely completion of tasks
sharing information and ideas about the task was associated for the DST assessment (Table 10) were associated with a
with a reduction in the likelihood of high learning group high satisfaction classification.
178 Journal of Marketing Education 45(2)

Table 8. Influence of Individual Engagement Variables on Satisfaction.

Variable β SE Wald Sig. Exp (β)

I-Connect 0.17 0.20 0.75 .39 1.18
I-Observe 0.31 0.20 2.29 .13 1.36
I-Diffways −0.39 0.22 3.10 .09 0.68
I-Enjoy 0.16 0.13 9.79 .00* 2.06
I-Anxious 0.17 0.13 1.40 0.24 1.17
I-Shareinfo −0.15 0.24 0.40 .53 0.86
I-Discuss 0.00 0.16 0.00 .98 1.00
I-Effort 0.00 0.25 0.00 .99 1.00
Constant −2.15 1.60 1.80 .18 .12

Note. Dependent variable coded 0 = low satisfaction and 1 = high satisfaction.

R² = .12 (Cox & Snell), 0.21 (Nagelkerke). Model χ²(1) = 18.98, p = .02. Hosmer and Lemeshow. χ² = 13.10, p = .11. Overall percentage correct = 85.7%.
*Significant at p < .05.

Table 9. Influence of Group Engagement Variables on Satisfaction.

Variable β SE Wald Sig. Exp (β)

G-Connect 0.32 0.24 1.70 .19 1.37
G-Observe 0.25 0.29 0.76 .38 1.29
G-Diffways 0.21 0.31 0.48 .49 1.24
G-Enjoy −0.56 0.32 3.15 .08 0.57
G-Anxious 0.20 0.15 1.88 .17 1.22
G-Shareinfo −0.21 0.33 0.40 .53 0.81
G-Discuss 0.26 0.24 1.20 .27 1.30
G-Effort 0.97 0.33 8.81 .00* 2.63
Constant −4.94 1.67 8.81 .00* 0.01

Note. Dependent variable coded 0 = low satisfaction and 1 = high satisfaction.

R² = .22 (Cox & Snell), 0.40 (Nagelkerke). Model χ²(1) = 38.01, p = .00 Hosmer and Lemeshow. χ² = 14.16, p = .08. Overall percentage correct = 90.9%.
*Significant at p < .05.

Table 10. Influence of Group Functioning Variables on Satisfaction.

Variable β SE Wald Sig. Exp (β)

G-Coop −0.31 0.29 1.13 .29 0.73
G-Organ −0.09 0.30 0.09 .77 0.91
G-Timely 1.06 0.31 11.99 .00* 2.88
G-Adjust −0.02 0.24 0.00 .94 0.98
Constant −1.11 1.01 1.21 .27 0.33

Note. Dependent variable coded 0 = low satisfaction and 1 = high satisfaction.

R² = .16 (Cox & Snell), 0.29 (Nagelkerke). Model χ²(1) = 26.70, p = .00 Hosmer and Lemeshow. χ² = 5.08, p = .75. Overall percentage correct = 87.7%.

Finally, as an indicator of learner success, we compared p = <.001, for the annual student cohort in the year immedi-
the overall results in the subject pre- and post-introduction of ately prior to the introduction of the DST (M = 18.47, SD =
the DST assessment. Using the overall results was consid- 3.83) compared with marks for the cohort in the year imme-
ered appropriate as the DST task was the only change in the diately following the final implementation of the task (M =
subject’s delivery, with all other design elements and teach- 21.21, SD = 2.01). These results provide some incidental if
ing staff held constant over the period. However, because the not direct positive evidence of the efficacy of the DST
sample consisted of students from different student cohorts, assessment.
we did not include this proxy for actual learning in our In answer to the research questions, the results of this
research questions. An independent t test showed a signifi- study provide empirical evidence for a DST assessment task
cant difference in overall subject marks, t(75.679) = −5.068, to stimulate both individual and collaborative student
Spanjaard et al. 179

engagement in cognitive, affective, and behavioral ways. attention. Assessment designers may need to improve their
The findings indicate positive relationships between indi- understanding of what aspect of a learning task has the
vidual and collaborative student engagement, as well as for potential to create anxiety for students and groups and
the positive influence of group functioning in the context of develop pre-emptive or supporting instructional material.
each engagement construct. In addition, the results suggest It should be noted that perceptions of anxiety cannot be
student engagement and group functioning may contribute to taken as an indicator of disengagement, particularly in light
incremental learning and subject satisfaction, with some of the nascent view that disengagement is conceptually dis-
objective evidence of learner success by way of cohort result tinct from engagement (Balwant, 2018; Kahu, 2013). It may
comparison. also be the case that measures of anxiety are better conceived
as a separate affective influence on engagement rather than
subsumed within the concept (Balwant, 2018). Nonetheless,
Discussion
the influence of students’ emotional responses that accom-
When effectively integrated, DST allows students to collabo- pany or are proximate to learning activities has received little
rate, and contribute to the development of new information. attention in the literature. More generally, our understanding
Ideally, it offers both subjective and objective perspectives of the role of emotion in student engagement would benefit
that can be brought about by the complexity of the experi- from future studies exploring such relationships in higher
ence, thus enhancing student learning in different ways to the education research (Kahu, 2013).
traditional assessment (Alterio, 2002). DST moves the
emphasis away from the activity being a teacher-driven Concluding Comments and Future
responsibility and places it squarely with the student, thus
encouraging them to build their knowledge creation exper-
Directions
tise. We contend that DST shows promise as an inherently Evidence provided from the results of this study suggests
social and interactive learning task to provide an authentic that DST as a collaborative task not only provides a means to
assessment for a range of marketing problems. In turn, these assess key learning outcomes but is also capable of doing so
attributes provide a stimulating vehicle for student engage- in a way that elevates student engagement, learning, and sat-
ment that can promote learning and satisfaction. isfaction. We believe DST has a promising contribution to
Items designed to measure the affective, behavioral, and make to marketing education if thoughtfully integrated
cognitive dimensions of student engagement featured across within a blended learning repertoire. In addressing the
the results, yet perhaps those relating to the affective mea- demands of graduate employers, DST requires students to
sures of engagement are most notable. The enjoyment harness technology in a creative yet purposeful way. As a
derived from working on the DST task was a dominant fea- result, students have an opportunity to develop additional
ture explaining both individual and collective engagement skills sets within their portfolio. As described, deploying a
and in the context of group functioning, a finding that sup- DST assessment in a range of subjects within marketing edu-
ports the work of Suwardy et al. (2013) on student-led DST cation beyond a broad-based marketing subject is readily
production. Furthermore, individual enjoyment was the only achievable.
significant engagement variable predictive of perceived Applying DST to tell stories within a branding subject, or
incremental learning and overall subject satisfaction. to imagine a customer journey as they study the theories of
Conversely, perceptions of anxiety were prominent in the consumer behavior is conceivable. It would also work well
results pertaining to the relationship between individual and for marketing communications subjects where students
collective engagement, with some evidence of the potential would not only consider the delivery of digital storytelling
for anxiety to impact cognitive engagement. Anxiety and its but also the embedding of communications theory. DST may
impact on cognition are well established in the literature. For be an innovative way to present findings in a marketing
example, in a study on the effects of computer anxiety on research subject or the results of attribution modeling in
task performance, Mahar et al. (1997) concluded that com- omnichannel marketing. The nature of the task does not need
puter anxiety was associated with deficit computer task per- to be limited to just one perspective of marketing theory or
formance after controlling for past experience and individual practice. The application of DST in marketing education is
state anxiety. In our study, student anxiety may be rooted in limited only by the imagination required to construct the tell-
the novelty of the DST task and the high proportion of inter- ing of a story itself.
national students in the sample predominantly from regions The longevity of interest in the student engagement con-
where their prior experience of assessment in education may struct is a testament to its importance to the academic-practi-
have been more conventional. While the difference was not tioner, faculty, institution, and indeed the higher education
significant, international students rated the DST assessment sector. Despite the controversy directed at its conceptualiza-
as generating more anxiety at an individual and group level tion and operationalization, there remain ample opportuni-
than domestic students, and this may warrant further ties to advance student engagement research. A rich body of
180 Journal of Marketing Education 45(2)

work has been established in pre-College populations (refer References

to Christenson et al., 2012), much of which may be applied Alterio, M. (2002). Using storytelling to enhance student learning
to higher education student samples. Using a comparative (p. 5). Higher Education Academy.
framing of survey questions, this study has demonstrated the Alterio, M., & McDrury, J. (2003). Learning through storytelling in
potential for the task itself to elicit more or less engagement higher education: Using reflection and experience to improve
from students. Future research instruments could be designed learning. Routledge.
to capture the comparison between assessment tasks more Attaway, J. S., Russ, R., Hair, J. F., & Burns, A. C. (1987). Instructor’s
explicitly with a 7-point scale from −3 (much less than) to Manual for Multivariate Data Analysis with Readings (2nd ed.).
+3 (much more than). Nevertheless, whether differential New York: Macmillan Publishing Company.
Bacon, D. R. (2016). Reporting actual and perceived student learn-
levels of engagement occur by virtue of the unique compo-
ing in education research. Journal of Marketing Education,
nents of an assessment design, the characteristics of the stu- 38(1), 3–6.
dent, or some other aspect of the learning environment are Balwant, P. T. (2018). The meaning of student engagement and
worthy of further inquiry. disengagement in the classroom context: Lessons from organ-
In the context of the current study, we contend that more isational behaviour. Journal of Further and Higher Education,
research is warranted on collaborative, or collective engage- 42(3), 389–401.
ment, not the least because of the ubiquity of group assess- Bandura, A. (1997). Self-efficacy: The exercise of control. W.H.
ment tasks in business education. For a similar reason, the Freeman.
lack of research on the role of group functioning in the con- Basil, M. D., & Basil, D. Z. (2008). The marketing market:
text of student engagement could be rectified. Work on team Matching academic hiring institutions and job candidates.
effectiveness in the organizational behavior literature may be Journal of Marketing Education, 30(2), 138–149.
Bass, B. M., Cascio, W. F., & O’Connor, E. J. (1974). Magnitude
a useful starting point.
estimations of expressions of frequency and amount. Journal of
In a comprehensive attempt at developing a conceptual Applied Psychology, 59(3), 313–320. https://doi.org/10.1037/
framework of student engagement, its antecedents, and con- h0036653
sequences, Kahu (2013, p. 769) notes that no one study can Bassano, C., Barile, S., Piciocchi, P., Spohrer, J. C., Iandolo, F., &
hope to incorporate the full scope of the construct, yet we can Fisk, R. (2019). Storytelling about places: Tourism marketing
advance our knowledge of some aspects “without denying in the digital age. Cities, 87, 10–20.
the existence of the others.” Bearman, M., Dawson, P., Bennett, S., Hall, M., Molloy, E., Boud,
With this in mind, this study has sought to gain insight D., & Joughin, G. (2017). How university teachers design
into an area that has received little attention in the literature assessments: A cross-disciplinary study. Higher Education,
in the examination of student perceptions of individual and 74(1), 49–64.
collaborative engagement with an innovative assessment Bechter, C., & Swierczek, F. W. (2017). Digital storytelling in a
flipped classroom for effective learning. Education Sciences,
task.
7(2), Article 61.
Bliuc, A.-M., Goodyear, P., & Ellis, R. A. (2007). Research focus
Declaration of Conflicting Interests and methodological choices in studies into students’ experi-
The author(s) declared no potential conflicts of interest with respect ences of blended learning in higher education. The Internet and
to the research, authorship, and/or publication of this article. Higher Education, 10(4), 231–244.
Boedeker, P., & Henson, R. K. (2020). Canonical correlation anal-
Funding ysis. SAGE.
Boelens, R., De Wever, B., & Voet, M. (2017). Four key challenges
The author(s) received no financial support for the research, author-
to the design of blended learning: A systematic literature
ship, and/or publication of this article.
review. Educational Research Review, 22, 1–18.
Bergkvist, L., & Rossiter, J. R. (2007). The predictive validity of
ORCID iD multiple-item versus single-item measures of the same con-
Daniela Spanjaard https://orcid.org/0000-0002-4021-368X structs. Journal of Marketing Research, 44(2), 175–184.
Bowden, L.-H., Tickle, L., & Naumann, K. (2021). The four pillars
of tertiary student engagement and success: A holistic measure-
Note
ment approach. Studies in Higher Education, 46(6), 1207–1224.
1. C-OAR-SE is the acronym for the steps in Rossiter’s (2011, Burgess, J. (2006). Hearing ordinary voices: Cultural studies, ver-
vii) “rational-realist measurement theory,” being Construct nacular creativity and digital storytelling. Continuum, 20(2),
definition, Object representation, Attribute classification, Rater 201–214.
identification, Scale formation, and Enumeration. The theory Carless, D. (2015). Exploring learning-oriented assessment pro-
is based on content validity via conceptual definition (being cesses. Higher Education, 69(6), 963–976.
the OAR component) and rater identification (e.g., individual Chapman, E. (2003). Alternative approaches to assessing stu-
student). The content valid definition guides scale construction dent engagement rates. Practical Assessment, Research and
and variable treatment for analysis where appropriate. Evaluation, 8, Article 13.
Spanjaard et al. 181

Christenson, S. L., Reschly, A. L., & Wylie, C. (2012). Handbook Hockings, C., Cooke, S., Yamashita, H., McGinty, S., & Bowl, M.
of research on student engagement. Springer. (2008). Switched off? A study of disengagement among comput-
Clarke, R., & Adam, A. (2012). Digital storytelling in Australia: ing students at two universities. Research Papers in Education,
Academic perspectives and reflections. Arts and Humanities in 23(2), 191–201. https://doi.org/10.1080/02671520802048729
Higher Education, 11(1–2), 157–176. Hollebeek, L. D., Glynn, M. S., & Brodie, R. J. (2014). Consumer brand
Crittenden, W. F. (2005). A social learning theory of cross-functional engagement in social media: Conceptualization, scale development
case education. Journal of Business Research, 58(7), 960–966. and validation. Journal of Interactive Marketing, 28(2), 149–165.
Davis, A. (2004). Co-authoring identity: Digital storytelling in Hosmer, D. W. Jr., Lemeshow, S., & Sturdivant, R. X. (2013).
an urban middle school. THEN: Technology, Humanities, Applied logistic regression (Vol. 398). Wiley.
Education, & Narrative, 1(1), 1–12. Järvelä, S., Järvenoja, H., Malmberg, J., Isohätälä, J., & Sobocinski,
Dikcius, V., Urbonavicius, S., Adomaviciute, K., Degutis, M., M. (2016). How do types of interaction and phases of self-
& Zimaitis, I. (2020). Learning marketing online: The role regulated learning set a stage for collaborative engagement?
of social interactions and gamification rewards. Journal of Learning and Instruction, 43, 39–51.
Marketing Education, 43(2), 159–173. Kahu, E. R. (2013). Framing student engagement in higher educa-
Dochy, F. J., & McDowell, L. (1997). Introduction: Assessment as tion. Studies in Higher Education, 38(5), 758–773.
a tool for learning. Studies in Educational Evaluation, 23(4), Kim, S.-H., & Shin, S. (2021). Social–emotional competence and
279–298. academic achievement of nursing students: A canonical cor-
Donohoo, J., Hattie, J., & Eells, R. (2018). The power of collective relation analysis. International Journal of Environmental
efficacy. Educational Leadership, 75(6), 40–44. Research and Public Health, 18(4), Article 1752.
Dos Santos, S. F., & Brandi, H. S. (2014). A canonical correlation Kinzie, J., & Gonyea, R. (2009). An introduction to the National Survey
analysis of the relationship between sustainability and com- of Student Engagement [Paper presentation]. Annual Forum of the
petitiveness. Clean Technologies and Environmental Policy, Association for Institutional Research, June, Atlanta, GA [online].
16(8), 1735–1746. https://scholarworks.iu.edu/dspace/bitstream/handle/2022/23688/
Eccles, J., & Wang, M.-T. (2012). Part I commentary: So what is An%20introduction%20to%20the%20National%20Survey%20
student engagement anyway? In S. Christenson, A. Reschly, & of%20Student%20Engagement.pdf?sequence=1
C. Wylie (Eds.), Handbook of research on student engagement Kolb, A. Y., & Kolb, D. A. (2005). Learning styles and learning
(pp. 133–145). Springer. spaces: Enhancing experiential learning in higher education.
Engellant, K. A., Holland, D. D., & Piper, R. T. (2016). Assessing Academy of Management Learning & Education, 4(2), 193–
convergent and discriminant validity of the motivation construct 212. https://doi.org/10.5465/AMLE.2005.17268566
for the technology integration education (TIE) model. Journal of Kolb, D. A. (1984). Experiential learning. Englewood Cliffs.
Higher Education Theory and Practice, 16(1), 37–50. Krause, K. (2007). Griffith University blended learning strategy
Fan, X., & Konold, T. R. (2018). Canonical correlation analysis. (Document number 2008, 16252).
In The Reviewers guide to quantitative methods in social sci- Kuh, G. D. (2009). What student affairs professionals need to
ences, (pp. 29–41). Routledge. know about student engagement. Journal of College Student
Fernandes, M. S. D. M. R. (2019). The power of stories to con- Development, 50(6), 683–706.
nect with customers: Digital storytelling as a communication Lukin, A. (2016, October 25). “He” vs “She” in Australian Media
marketing tool to engage with millennials and its impact on the coverage: What the language of news tells us about gender
creation of value for firms [Doctoral dissertation, Universidade imbalance. The Conversation, 26.
Católica Portuguesa]. Mahar, D., Henderson, R., & Deane, F. (1997). The effects of
Fisher, R., Perényi, Á., & Birdthistle, N. (2021). The positive rela- computer anxiety, state anxiety, and computer experience on
tionship between flipped and blended learning and student users’ performance of computer based tasks. Personality and
engagement, performance and satisfaction. Active Learning in Individual Differences, 22(5), 683–692.
Higher Education, 22(2), 97–113. Mahmood, M. A., & Mann, G. J. (1993). Measuring the organiza-
Fusté-Forné, F. (2021). The portrayal of Greenland: A visual analy- tional impact of information technology investment: An explor-
sis of its digital storytelling. Current Issues in Tourism, 1–6. atory study. Journal of Management Information Systems,
Advance online publications 10(1), 97–122.
Garrison, D. R., & Vaughan, N. D. (2007). Blended learning in higher Mandernach, B. J. (2015). Assessment of student engagement in
education: Framework, principles and guidelines. Jossey-Bass. higher education: A synthesis of literature and assessment tools.
Garson, G. (2016). Logistic regression: Binary & multinomial: International Journal of Learning, Teaching and Educational
2016 edition (statistical associates “Blue Book Series”). Research, 12(2), 1–14.
Statistical Associates Publishers. McDonald, P. L. (2014). Variation in adult learners’ experiences of
Gray, J. A., & DiLoreto, M. (2016). The effects of student engage- blended learning in higher education. Blended Learning, 2(6),
ment, student satisfaction, and perceived learning in online 215–234.
learning environments. International Journal of Educational McDougal, E. R., Syrdal, H. A., Gravois, R., & Kemp, A. (2021).
Leadership Preparation, 11(1), 1–20. Telling the tale: Applying a strategic brand storytelling process
Greene, H., Koh, K., Bonnici, J., & Chase, J. (2015). The value for STP planning. Journal of Strategic Marketing, 1–21.
of storytelling in the marketing curriculum. Journal of the Moreau, K. A., Eady, K., Sikora, L., & Horsley, T. (2018). Digital
Academy of Business Education, 16, 111–128. storytelling in health professions education: A systematic
Hair, J. F. (2009). Multivariate data analysis (7th ed.). Prentice Hall. review. BMC Medical Education, 18(1), 1–9.
182 Journal of Marketing Education 45(2)

Morrison, M., Sweeney, A., & Heffernan, T. (2006). Karns’s learn- Singh, P., Thambusamy, R. X., & Ramly, M. A. (2014). Fit or unfit?
ing styles and learning effectiveness: A rejoinder. Journal of Perspectives of employers and university instructors of gradu-
Marketing Education, 28(1), 64–68. ates’ generic skills. Procedia-Social and Behavioral Sciences,
Nelson, K. J., & Clarke, J. A. (2014). The first year experience: 123, 315–324.
Looking back to inform the future. HERDSA Review of Higher Steele, G. A. (2015). New postgraduate student experience and
Education, 1, 23–46. engagement in human communication studies. Journal of
Niemi, H., & Multisilta, J. (2016). Digital storytelling promot- Further and Higher Education, 39(4), 498–533.
ing twenty-first century skills and student engagement. Steen-Utheim, A. T., & Foldnes, N. (2018). A qualitative investiga-
Technology, Pedagogy and Education, 25(4), 451–468. https:// tion of student engagement in a flipped classroom. Teaching in
doi.org/10.1080/1475939X.2015.1074610 Higher Education, 23(3), 307–324.
Owston, R., York, D., & Murtha, S. (2013). Student perceptions Suwardy, T., Pan, G., & Seow, P.-S. (2013). Using digital storytell-
and achievement in a university blended learning strategic ini- ing to engage student learning. Accounting Education, 22(2),
tiative. The Internet and Higher Education, 18, 38–46. 109–124.
Parsons, J., & Taylor, L. (2011). Student engagement: What do we Thompson, B. (1984). Canonical correlation analysis: Uses and
know and what should we do? University of Alberta. interpretation, (No. 47). Sage.
Porter, B. (2005). The art of digital storytelling. Revista Discovery Thompson, B. (1991). A primer on the logic and use of canonical
Education. http://psdtech.pbworks.com/f/ArtOfStorytelling. correlation analysis. Measurement and evaluation in counsel-
pdf ing and development, 24(2), 80–95.
Postareff, L., Virtanen, V., Katajavuori, N., & Lindblom-Ylänne, S. Van Auken, H. E., Doran, B. M., & Yoon, K.-J. (1993). A financial
(2012). Academics’ conceptions of assessment and their assess- comparison between Korean and US firms: A cross balance
ment practices. Studies in Educational Evaluation, 38(3–4), sheet canonical correlation analysis. Journal of Small Business
84–92. Management, 31(3), Article 73.
Pugh, R. C., & Hu, Y. (1991). Use and interpretation of canonical Van Laer, T., De Ruyter, K., Visconti, L. M., & Wetzels, M. (2014).
correlation analyses in Journal of Educational Research arti- The extended transportation-imagery model: A meta-analysis
cles: 1978-1989. The Journal of Educational Research, 84(3), of the antecedents and consequences of consumers’ narrative
147–152. transportation. Journal of Consumer Research, 40(5), 797–817.
Robin, B. R. (2008). Digital storytelling: A powerful technology Vaughan, N. (2014). Student engagement and blended learning:
tool for the 21st century classroom. Theory into Practice, 47(3), Making the assessment connection. Education Sciences, 4,
220–228. 247–264. https://doi.org/10.3390/educsci4040247
Robin, B. R. (2016). The power of digital storytelling to support Villarroel, V., Bloxham, S., Bruna, D., Bruna, C., & Herrera-Seda,
teaching and learning. Digital Education Review, 30, 17–29. C. (2018). Authentic assessment: Creating a blueprint for
Rossiter, J. R. (2011). Marketing measurement revolution: The course design. Assessment & Evaluation in Higher Education,
C-OAR-SE method and why it must replace psychometrics. 43(5), 840–854.
European Journal of Marketing, 45(11/12),1561–1588. Vivek, S. D., Beatty, S. E., Dalela, V., & Morgan, R. M. (2014). A gen-
Sahin, I., Celik, I., Oguz Akturk, A., & Aydin, M. (2013). Analysis eralized multidimensional scale for measuring customer engage-
of relationships between technological pedagogical content ment. Journal of Marketing Theory and Practice, 22(4), 401–420.
knowledge and educational internet use. Journal of Digital Wang, S., & Zhan, H. (2010). Enhancing teaching and learning with
Learning in Teacher Education, 29(4), 110–117. digital storytelling. International Journal of Information and
Sahni, J. (2019). Does blended learning enhance student engage- Communication Technology Education (IJICTE), 6(2), 76–87.
ment? Evidence from higher education. Journal of e-Learning Wang, M. T., Fredricks, J. A., Ye, F., Hofkens, T. L., & Linn, J. S.
and Higher Education, 2019, Article 121518. (2016). The math and science engagement scales: Scale devel-
Sarris, G. (1990). Storytelling in the classroom: Crossing vexed opment, validation, and psychometric properties. Learning and
chasms. College English, 52(2), 169–185. Instruction, 43,16–26.
Schaufeli, W. B., Martinez, I. M., Pinto, A. M., Salanova, M., & Woodside, A. G. (2010). Brand-consumer storytelling theory and
Bakker, A. B. (2002). Burnout and engagement in university research: Introduction to a Psychology & Marketing special
students: A cross-national study. Journal of Cross-Cultural issue. Psychology & Marketing, 27(6), 531–540.
Psychology, 33(5), 464–481. Yorke, M. (2006). Student engagement: Deep, surface or strategic
Schmoelz, A. (2018). Enabling co-creativity through digital story- [Paper presentation]. Keynote address to the 9th Pacific Rim
telling in education. Thinking Skills and Creativity, 28, 1–13. Conference on the First Year in Higher Education, Griffith
Serrano, D. R., Dea-Ayuela, M. A., Gonzalez-Burgos, E., Serrano- University, Australia.
Gil, A., & Lalatsa, A. (2019). Technology-enhanced learn- Zepke, N. (2013). Student engagement: A complex business support-
ing in higher education: How to enhance student engagement ing the first year experience in tertiary education. The International
through blended learning. European Journal of Education, Journal of the First Year in Higher Education, 4(2), 1–14.
54(2), 273–286. Zepke, N., Leach, L., & Butler, P. (2014). Student engagement:
Shankar, A., Elliott, R., & Goulding, C. (2001). Understanding con- Students’ and teachers’ perceptions. Higher Education
sumption: Contributions from a narrative perspective. Journal Research & Development, 33(2), 386–398.
of Marketing Management, 17(3–4), 429–453. Zhai, X., Gu, J., Liu, H., Liang, J.-C., & Tsai, C.-C. (2017). An
Sherry, A., & Henson, R. K. (2005). Conducting and interpreting experiential learning perspective on students’ satisfaction
canonical correlation analysis in personality research: A user- model in a flipped classroom context. Journal of Educational
friendly primer. Journal of Personality Assessment, 84(1), 37–48. Technology & Society, 20(1), 198–210.

View publication stats