0% found this document useful (0 votes)
29 views

Sullivan Dissertation 2016

Uploaded by

Hamada Elzaiat
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
29 views

Sullivan Dissertation 2016

Uploaded by

Hamada Elzaiat
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 193

COSTS AND BENEFITS OF MIND WANDERING IN A TECHNOLOGICAL SETTING:

FINDINGS AND IMPLICATIONS

Yulia W. Sullivan

Dissertation Prepared for the Degree of

DOCTOR OF PHILOSOPHY

UNIVERSITY OF NORTH TEXAS

August 2016

APPROVED:

Chang E. Koh, Major Professor


Fred D. Davis, Committee Member
Robert Pavur, Committee Member
Mary C. Jones, Chair of the Department of
Information Technology and Decision
Science
Marilyn Wiley, Dean of the College of
Business
Victor Prybutok, Vice Provost of the Toulouse
Graduate School
Sullivan, Yulia W. Costs and Benefits of Mind Wandering in a Technological Setting:

Findings and Implications. Doctor of Philosophy (Business Computer Information Systems),

August 2016, 182 pp., 30 tables, 15 figures, chapter references.

The central purpose of this dissertation is to develop and test a theoretical model of

mind wandering in a technological setting by integrating the emerging work and theory on

mind wandering—a shift of attention from the primary task to the processing of internal goals.

This dissertation is intended to advance our understanding on the costs and benefits of mind

wandering in information systems (IS) research and in turn, contribute to the literature of

cognitive IS research. Understanding the consequences of mind wandering in a technological

setting is imperative because mind wandering plays a vital role in influencing various outcomes

associated with technology use and/or technology learning, such as technology anxiety,

software self-efficacy, and task performance. This dissertation is composed of three essays

which examine the determinants and consequences of mind wandering and focus of attention

on a number of emotional and cognitive outcomes. A multi-method approach (i.e., online

survey and laboratory experiment) across three essays is used to test the research models.
 
 
 
 
 
 
 
 
 
 
 
 
 
Copyright 2016 
by 
Yulia W. Sullivan 
 
   

ii 
 
ACKNOWLEDGEMENTS 

  I would like to gracefully acknowledge various people who have been in a journey with 

me in recent years as I worked on my doctoral degree. I owe an enormous debt of gratitude to 

my family, especially my husband, Barry Sullivan, for his longstanding support for my academic 

degree. He has been a constant source of love, concern, and support all these years.  

  I would like to thank my advisor, Dr. Chang Koh for his continued guidance and support 

during the pursuit of my degree. I have been amazingly fortunate to have an advisor who gave 

me freedom to explore on my own, and at the same time, was willing to walk together with me 

through my dissertation journey. I am also thankful to him for always believing in me.  

  Dr. Fred Davis, one of the best teachers and mentors I have had in my life. I am deeply 

grateful to him for always be there for me to listen and give advises. Despite his busy schedule, 

he still made time to have long discussions with me that had helped me sort out the theoretical 

details of my work. He has taught me how to think outside the box and plan for a long‐term 

research pipeline. I am grateful to him for holding me to a high research standard and teaching 

me how to do research from his own personal experience.  

  I’m grateful to Dr. Robert Pavur for his insightful comments on some of the data 

analyses I did in my dissertation. I am also thankful to him for always monitoring my academic 

status and making sure I was on the right track. I would like to also acknowledge Dr. Daniel Peak 

for helping me purchase the software I used in my dissertation.  

  Last but not least, I would like to thank my friends who have helped me stay sane 

through these years. I will forever cherish their friendship and I deeply appreciate their belief in 

me.  

iii 
 
TABLE OF CONTENTS 

Page 
ACKNOWLEDGEMENTS ................................................................................................................... iii 
LIST OF TABLES .............................................................................................................................. viii 
LIST OF FIGURES ............................................................................................................................... x 
CHAPTER 1 GENERAL INTRODUCTION ............................................................................................ 1 
1.1  References ......................................................................................................................... 4 
CHAPTER 2 MIND WANDERING: TECHNOLOGY AND EMOTIONAL OUTCOMES CORRELATES ...... 5 
2.1   Introduction ....................................................................................................................... 5 
2.2   The Conceptualization of Mind Wandering ...................................................................... 8 
2.3   Content Regulation Hypothesis of Mind Wandering ........................................................ 9 
2.3.1   Technology‐related (IT) versus Non‐technology‐related (Non‐IT) Thoughts ......... 12 
2.4  Emotional Outcomes ....................................................................................................... 16 
2.5   Research Model ............................................................................................................... 18 
2.6  Hypotheses Development ............................................................................................... 19 
2.6.1   Mind Wandering and Technology Anxiety ............................................................. 19 
2.6.2  Focus of Attention and Technology Anxiety ........................................................... 21 
2.6.3   Mind Wandering and User Satisfaction .................................................................. 23 
2.6.4  Focus of Attention and User Satisfaction ............................................................... 24 
2.6.5  Technology Anxiety and User Satisfaction.............................................................. 25 
2.6.6.   Control Variables ..................................................................................................... 26 
2.7   Research Method ............................................................................................................ 27 
2.7.1  Instrumental Development ..................................................................................... 28 
2.7.2  Data Collection ........................................................................................................ 30 
2.8   Data Analysis and Results ................................................................................................ 30 
2.8.1  Assessment of Measurement Validation ................................................................ 30 
2.8.2   Structural Model ..................................................................................................... 37 
2.9   Discussion ........................................................................................................................ 39 
2.9.1  Theoretical Contributions ....................................................................................... 40 
2.9.2  Practical Contributions ............................................................................................ 42 

iv 
 
2.10  Conclusions ...................................................................................................................... 44 
2.11   References ....................................................................................................................... 44 
CHAPTER 3 MIND WANDERING AND COGNITIVE IS LEARNING ................................................... 52 
3.1  Introduction ..................................................................................................................... 52 
3.2   Executive Functions Theory ............................................................................................. 54 
3.2.1   Inhibition ................................................................................................................. 55 
3.2.2   Switching ................................................................................................................. 57 
3.2.3   Updating or Working Memory Capacity ................................................................. 57 
3.3  Mind Wandering .............................................................................................................. 58 
3.3.1   Executive‐Control Failure Theory of Mind Wandering ........................................... 59 
3.3.2  Consequences of Mind Wandering ......................................................................... 62 
3.4   Resource Allocation Theory ............................................................................................ 65 
3.5  Research Model ............................................................................................................... 67 
3.6  Hypotheses Development ............................................................................................... 69 
3.6.1  Executive Functions, Mind Wandering, and Focus of Attention ............................ 69 
3.6.2  Perceived Task Complexity, Mind Wandering, and Focus of Attention ................. 71 
3.6.3  Focus of Attention, Mind Wandering, and Software Self‐Efficacy ......................... 72 
3.6.4  Focus of attention, Mind Wandering, and Learning Performance ......................... 79 
3.6.5  Software Self‐efficacy and Learning Performance ................................................. 81 
3.6.6  Control Variables ..................................................................................................... 82 
3.7  Research Method ............................................................................................................ 83 
3.7.1  Research Context and Sample ................................................................................ 83 
3.7.2  Operationalization of Research Constructs ............................................................ 85 
3.8  Data Analysis and Results ................................................................................................ 87 
3.8.1  Measurement Model .............................................................................................. 87 
3.8.2  Structural Model ..................................................................................................... 93 
3.8.3  Post Hoc Analysis: Moderated Mediation Model ................................................. 102 
3.9  Discussion ...................................................................................................................... 103 
3.9.1  Theoretical Implications........................................................................................ 104 
3.9.2  Limitations and Future Work ................................................................................ 107 
3.9.3  Practical Contributions .......................................................................................... 108 


 
3.10  Conclusions .................................................................................................................... 110 
3.11  References ..................................................................................................................... 110 
CHAPTER 4 MIND WANDERING AND IS DECISION MAKING PERFORMANCE ............................. 121 
4.1  Introduction ................................................................................................................... 121 
4.2  Executive Control Theory of Mind Wandering .............................................................. 123 
4.3  Unconscious Thought Theory ........................................................................................ 125 
4.4  Research Model ............................................................................................................. 126 
4.5  Hypotheses Development ............................................................................................. 128 
4.5.1  Task Complexity and IS Decision Making Performance ........................................ 128 
4.5.2  Focus of Attention and Performance Efficiency ................................................... 128 
4.5.3  Mind Wandering and Performance Accuracy ....................................................... 129 
4.5.4  The Moderating Effect of IS Task Complexity and Focus of Attention ................. 130 
4.5.5  The Moderating Effect of IS Task Complexity and Mind Wandering ................... 131 
4.6  Research Method .......................................................................................................... 133 
4.6.1  Experimental Design ............................................................................................. 133 
4.6.2  Experimental Procedure ....................................................................................... 135 
4.6.3  Manipulation Check .............................................................................................. 136 
4.6.4  Sample ................................................................................................................... 138 
4.7  Data Analysis and Results .............................................................................................. 138 
4.7.1  Performance Accuracy .......................................................................................... 139 
4.7.2  Performance Efficiency ......................................................................................... 144 
4.8  Discussion and Implications .......................................................................................... 146 
4.8.1  Theoretical Implications........................................................................................ 147 
4.8.2  Practical Implications ............................................................................................ 148 
4.8.3  Limitations............................................................................................................. 149 
4.9  Conclusions .................................................................................................................... 151 
4.10  References ..................................................................................................................... 151 
APPENDIX A TESTING FOR COMMON METHOD BIAS (WITH A COMMENT LATENT FACTOR) ... 157 
APPENDIX B MEASUREMENT ITEMS ........................................................................................... 159 
APPENDIX C LEARNING PERFORMANCE MEASURES .................................................................. 163 
APPENDIX D EXECUTIVE FUNCTIONS TASKS ............................................................................... 166 

vi 
 
APPENDIX E SAMPLE EXPERIMENT MATERIALS ......................................................................... 171 
APPENDIX F INFORMED CONSENT NOTICE ................................................................................ 176 
REFERENCES FOR APPENDICES ................................................................................................... 181 
 
   

vii 
 
LIST OF TABLES 

Page
Table 2.1 Summary of Mind Wandering Studies in the Context of Emotions……………………….........11 
Table 2.1 Construct Definition………………………........………………………........………………………................15 
Table 2.3 Sample Characteristics………………………........………………………........………………………............31 
Table 2.4 EFA Result………………………........………………………........………………………........……………………..32 
Table 2.5 Summary of Model Fit for All CFAs Examined………………………........……………………….........33 
Table 2.6 Summary of Confirmatory Factor Analysis (8‐Factor Model) ………………………..................34 
Table 2.7 Correlation Matrix and Descriptive Statistics………………………........………………………..........36 
Table 2.8 Chi‐Square, Goodness‐of‐Fit Values, and Model Comparison Tests………………………........39 
Table 2.9 Hypotheses Results………………………........………………………........………………………........……….39 
Table 3.1 Costs and Benefits of Mind Wandering According to the Previous Studies…………………..62 
Table 3.2 Construct Definitions………………………........………………………........………………………..............68 
Table 3.3 Interactions Between Focus of Attention and Mind Wandering (Dependent Variable: 
Software Self‐Efficacy) ………………………........………………………........………………………........…………………79 
Table 3.4 Executive Functions Measures………………………........………………………........………………………86 
Table 3.5 Confirmatory Factor Analysis (PLS Results) ………………………........……………………….............90 
Table 3.6 Correlation Matrix………………………........………………………........………………………...................92 
Table 3.7 The Effect of Mind Wandering and Focus of Attention on the Dependent 
Variables……………………………………………………………………………………………………………………………………..96 
Table 3.8 Hypotheses Summary………………………........………………………........………………………............101
Table 3.9 Direct and Indirect Effect of Focus of Attention (IT) Mediated by SSSE on Learning 
Performance………………………........………………………........………………………........……………………….........103 
Table 3.10 Direct and Indirect Effect of Focus of Attention (Non‐IT) Mediated by SSSE on Learning 
Performance……………………........………………………........………………………........………………………………….103
Table 4.1 Construct Definition………………………........………………………........………………………........…….127 
Table 4.2 Website and Product Criteria………………………........………………………........………………………135 
Table 4.3 Measurement Items and EFA Results………………………........………………………....................137 
Table 4.4 Group Comparison (ANOVA) Results………………………........……………………….....................139 

viii 
 
Table 4.5 GLM Results for Performance Accuracy………………………........………………………................140 
Table 4.6 Regression Analysis with Dummy Coding for Task Complexity………………………..............141 
Table 4.7 Regression Equation for Each Condition (DV: Performance Accuracy……....………………..142 
Table 4.8 Interaction Contrast Analysis Results………………………........………………………....................143 
Table 4.9 Regression Analysis with Dummy Coding for Task Complexity (DV: Performance 
Efficiency) ………………………........………………………........………………………........………………………............145 
Table 4.10 Regression Equation for Each Condition (DV: Performance Efficiency).……………….……145 
Table 4.11 Hypotheses Summary………………………........………………………........………………………..........146
 
   

ix 
 
LIST OF FIGURES 

Page
Figure 1.1. Dissertation Framework……………………….........................................................................3 
Figure 2.1. Research Model………………………........………………………........………………………....................19 
Figure 2.2. Structural Model…………………........………………………........………………………........................38 
Figure 3.1. Graphic Illustration of Executive‐Control Failure Theory…………………………………………….61 
Figure 3.2. Research Model………………………………………………………………………………………………………….68
Figure 3.3. Structural Model………………………………………………………………………………………………………..94 
Figure 3.4. The Moderating Effect of Mind Wandering (IT) on the Relationship between Focus of 
Attention (IT) and Software Self‐Efficacy……………………………………………………………………………………..97 
Figure 3.5. The Moderating Effect of Mind Wandering (Non‐IT) on the Relationship between Focus 
of Attention (IT) and Software Self‐Efficacy…………………………………………………………………………………98 
Figure 3.6. The Moderating Effect of Mind Wandering (Non‐IT) on the Relationship between Focus 
of Attention (Non‐IT) and Software Self‐Efficacy………………………………………………………………………….98 
Figure 3.7. The Moderating Effect of Mind Wandering (IT) on the Relationship between Focus of 
Attention (Non‐IT) and Software Self‐Efficacy……………………………………………………………………………..99 
Figure 3.8. The Moderating Effect of Mind Wandering (IT) and Mind Wandering (Non‐IT) on the 
Relationship between Focus of Attention (IT) and Software Self‐Efficacy………………………………….100 
Figure 3.9. The Moderating Effect of Mind Wandering (IT) and Mind Wandering (Non‐IT) on the 
Relationship between Focus of Attention (Non‐IT) and Software Self‐Efficacy…………………………..100 
Figure 4.1. Research Model……………………………………………………………………………………………………….127 
Figure 4.2. The Moderating Effect of Task Complexity between Mind Wandering and Performance 
Accuracy……………………………………………………………………………………………………………………………………143 
Figure 4.3. The Moderating Effect of Task Complexity between Focused Attention and 
Performance Accuracy………………………………………………………………………………………………………………144 


 
CHAPTER 1 

GENERAL INTRODUCTION 
 
The faculty of voluntarily bringing back a wandering attention, over and 
over again, is the very root of judgment, character and will…  
William James, 1890 
 

The central purposes of this dissertation are to develop and test a theoretical model of 

mind wandering in a technological setting by integrating the emerging work and theory on 

mind wandering (i.e., “a shift of executive control away from a primary task to the processing of 

personal goals” (Smallwood & Schooler, 2006, p. 946)) and information systems (IS). This 

dissertation is intended to advance our understanding on the costs and benefits of mind 

wandering and in turn, contribute to the literature of cognitive IS research. Research indicates 

that people spend up to 50 percent of their day‐to‐day lives engaged in mind wandering and 

this experience occurs frequently in all forms of activity (Killingsworth & Gilbert, 2010; Randall, 

Oswald, & Beier, 2014). Research also shows that mind wandering is associated with significant 

costs (e.g., disrupted task performance, poor psychological wellbeing, and unhappiness) (e.g., 

Killingsworth & Gilbert, 2010; McVay & Kane, 2010) and benefits (e.g., mental break, problem 

solving, and future planning) (e.g., Mooneyham & Schooler, 2013; Smallwood & Schooler, 

2015). Given these potential costs and benefits can occur when users interact with technology, 

understanding the consequences of mind wandering in a technological setting is imperative 

because mind wandering plays a vital role in influencing various outcomes associated with 

technology use and/or technology learning, such as technology anxiety, users’ satisfaction, self‐

efficacy, and task performance (Smallwood & Schooler, 2015). 


 
Although mind wandering has been recently studied in the reference disciplines, such as 

cognitive psychology and neuroscience, this topic has been overlooked in IS research—

regardless of the fact that distraction could automatically take place internally in our mind as 

we continuously interact with technology. Therefore, the studies reported herein are intended 

to shed light on the costs and benefits of mind wandering in a technological setting. To achieve 

this goal, the following aims have been identified and will be reported: 

Aim #1: To test the influence of previously identified mind wandering and focus of 

attention determinants in a technological setting.  

Aim #2: To develop and test a research framework that explains the effect of mind 

wandering and focus of attention on emotional outcomes and cognitive learning 

outcomes.  

Aim #3: To examine the effect of mind wandering and focus of attention on IS decision 

making performance. 

Together, these objectives are selected to understand the potential implications of mind 

wandering in the technology domain. In order to achieve these objectives, this dissertation is 

composed of three essays which examine the effect of mind wandering on emotional, 

cognitive, and performance outcomes. At the beginning of Essay 1, scales to measure mind 

wandering are developed and validated. After that, the relationship between mind wandering 

and emotional outcomes (i.e., technology anxiety and user satisfaction) is examined. Essay 2 

specifically focuses on the effect of mind wandering and focus of attention in the IS learning 

context. Essay 2 explores the hypotheses concerning the roles of executive functions (i.e., 

working memory, inhibition, and switching ability) and task complexity in influencing the 


 
occurrence of mind wandering and focus of attention and, in turn, cognitive outcomes, 

including software self‐efficacy and IS learning performance. Essay 3 examines the extent to 

which individuals’ focus of attention and mind wandering influence IS decision making 

performance at different levels of task complexity. The conceptual model underlying this 

dissertation is illustrated in Figure 1.1. 

Figure 1.1. Dissertation Framework 

Drawing upon the theory of mind wandering, executive functions theory, cognitive 

resource allocation theory, and unconscious thought theory, this dissertation establishes a new 

form of mind wandering theory to explain the consequences of mind wandering on a variety of 

outcomes, especially when users utilize technology as a part of their task. The results advance 

IS theories by revealing the importance of mind wandering in determining the success of IS 

users when they perform IS tasks. Further, this dissertation sheds light on the relationship 

between executive functions and the content of mind wandering that technology users 

generally generate when they interact with technology. The findings also explains cognition in 


 
the high‐technology environment, revealing possible ways to minimizing the negative effect of 

mind wandering. 

A multi‐method approach (i.e., online survey and laboratory experiment) across three 

essays is used to examine the antecedents and consequences of mind wandering in a 

technological environment. Each essay, including the model and constructs’ definitions, is 

discussed in more details in the next sections. 

1.1  References 

Killingsworth, M. A., & Gilbert, D. T. (2010). A wandering mind is an unhappy mind. Science, 
330, 932. 
McVay, J. C., & Kane, M. J. (2010). Does mind wandering reflect executive function or executive 
failure? Comment on Smallwood and Schooler (2006) and Watkins (2008). Psychological 
Bulletin, 136, 188‐197. 
Mooneyham, B. W., & Schooler, J. W. (2013). The costs and benefits of mind wandering: A 
review. Canadian Journal of Experimental Psychology, 67(1), 11‐18. 
Randall, J. G., Oswald, F. L., & Beier, M. E. (2014). Mind wandering, cognition, and performance: 
A theory‐driven meta‐analysis of attention regulation. Psychological Bulletin, 140(6), 1411‐
1431. 
Smallwood, J. M., & Schooler, J. W. (2006). The restless mind. Psychological Bulletin, 132(6), 
946‐958. 
Smallwood, J. M., & Schooler, J. W. (2015). The science of mind wandering: Empirically 
navigating the stream of consciousness. Annual Review of Psychology, 66, 487‐518. 
 
   


 
CHAPTER 2 

MIND WANDERING: TECHNOLOGY AND EMOTIONAL OUTCOMES CORRELATES 
 
2.1   Introduction 

  Much research in information systems (IS) has focused on identifying factors that 

influence users’ attitude or behaviors toward technology (e.g., Davis, Bagozzi, & Warshaw,, 

1989; Taylor & Todd 1995; Venkatesh, Morris, Davis, & Davis, 2003). Although this stream of 

research has shed light on various cognitive and affective factors affecting users’ decision to use 

technology, it pays little attention to the content of users’ mind (i.e., what’s going on inside 

their head when they use technology). In our study of IS, we generally assume that users are 

capable at times of choosing to act in a rational way, leading to intentional response (Bagozzi, 

2007). However, the way users experience technology may be determined by their inner 

thought contents and these thought contents may influence the outcomes of technology use. In 

a certain sense, you are where your attention takes you. 

  In an effort to understand users experience of using technology, research on how 

attention shifts between external sources (e.g., between one technology to another) has been 

conducted by a number of IS scholars (e.g., Addas & Pinsonneault, 2014; Aral, Brynjolfsson, & 

Van Alstyne, 2012; Speier, Vassey, & Valacich, 2003). However, in a relatively short history of 

information systems (IS), which has inherited much of its models, theories and findings from 

other disciplines, such as management, economics, and psychology, a study on how users’ 

attention shifts between external events and internal thought is virtually absent as a subject of 

research (Sullivan, Davis, & Koh, 2015). This attentional shift is called mind wandering, meaning 


 
that the mind wanders away from its current train of thought (often an external task) to the 

internal processing of personal goals (Smallwood & Schooler, 2015). Given the fact that people 

spend around eleven hours or more per day using technology (Petronzio, 2014), it is likely that 

mind wandering can occur simultaneously when individuals engage in technology use. Although 

previous studies have shown that mind wandering was associated with significant costs, 

including disrupted task performance, poor psychological wellbeing, and unhappiness 

(Killingsworth & Gilbert, 2010; McVay & Kane, 2010; Moberly & Watkins, 2008; Mooneyham & 

Schooler, 2013; Smallwood & Schooler, 2006; Watkins, 2008), the effect of this phenomenon in 

the technology domain is unknown.  

  Research in cognitive psychology and cognitive neuroscience has made a significant 

progress in understanding the phenomenon of mind wandering. Current research in this field 

has revealed that mind wandering has a strong relationship with emotional outcomes. For 

example, Killingsworth and Gilbert (2010) conducted an experience sampling study and found 

that participants were least happy when resting, working or using a home computer and 

reported that mind wandering was the cause of their unhappiness across these activities. 

Another line of research suggests that the content of mind wandering should be taken into 

account before any conclusions on the consequences of such experience can be made 

(Andrews‐Hanna, Smallwood, & Spreng, 2014). The hypothesis suggesting that content of mind 

wandering determines its outcomes is known as the content regulation hypothesis. Specifically, 

it is argued that mind wandering serves an adaptive purpose by allowing individuals to prepare 

for upcoming events, form a sense of self‐identity, and navigate the social world (Andrews‐

Hanna et al., 2014; Smallwood & Andrews‐Hanna, 2013). However, even the existing studies in 


 
the reference disciplines have primarily focused on the frequency of occurrence of mind 

wandering and neglected the importance of the content of mind wandering itself. In order to 

understand the nature, dimensionality, and relative effects of mind wandering on its functional 

outcomes, there is a need to understand the different aspects of mind wandering contents that 

occupy individuals’ minds.  

Studying mind wandering from an IS perspective will advance our understanding on 

cognition and emotion from the actual technology use experience (Ortiz de Guinea & Webster, 

2013) and help us understand the thought patterns that shape users’ experience (Ortiz de 

Guinea & Markus, 2009). Although previous studies have made a significant contribution on the 

cognitive and neural bases of mind wandering (Buckner, Andrews‐Hanna, & Schacter, 2008; 

Smallwood & Schooler, 2006), little is known about its emotional consequences (Killingsworth & 

Gilbert, 2010), especially in an IS setting. Therefore, the purpose of this work is to investigate 

the relationship between mind wandering, either when users mind wander about technology or 

non‐technology‐related subjects, and emotional outcomes. Specifically, this study addresses 

the following research questions: Do the contents of thought during mind wandering episodes 

and focus of attention matter? If so, how do different types of thought content influence users’ 

emotional outcomes (i.e., technology anxiety and users’ satisfaction)?  

This work contributes to the literature in several ways. First, it extends the scope of 

inquiry of mind wandering into a technological context. Investigating how mind wandering is 

related to emotional outcomes would advance our understanding on the consequences of mind 

wandering in everyday life. Second, it also extends and tests the content regulation hypothesis 


 
by differentiating the content of thought into two dimensions: technology‐related (IT) and non‐

technology‐related (non‐IT).  

2.2   The Conceptualization of Mind Wandering1 

  Mind wandering has been framed in the context of a variety of constructs, including 

“task‐unrelated thoughts” (TUT) (Smallwood, Baracaia, Lowe, & Obonsawin, 2003; Smallwood 

et al., 2004), “spontaneous thought” (Christoff, Gordon, & Smith, 2011), and “daydreaming” 

(Giambra, 1979). Mind wandering is generally distinct from focus of attention (Randall, Oswald, 

& Beier, 2014). Whereas focus of attention refers to any thought directed to the primary task, 

mind wandering is attention directed toward concerns that are unrelated to the task at hand 

(Randall et al., 2014). It occurs when the content of a person’s current thought is not associated 

with the primary task (or task‐related thinking) but rather, is filled with other unrelated 

thoughts (Smallwood & Schooler, 2006). When individuals mind wander, they become 

“perceptually decoupled,” showing attention becomes coupled to an internal process and 

decoupled from external stimuli (Smallwood & Schooler, 2015). Because attention is decoupled 

from the primary task, representations of the task may be less detailed than during periods of 

time when attention is focused on the task (Smallwood, Fishman, & Schooler, 2007). 

  An established body of research over the last decade has linked a large‐scale brain 

system referred to as the default mode network (DMN) to mind wandering (Andrews‐Hanna et 

al. 2014; Christoff, Gordon, Smallwood, Smith, & Schooler, 2009). The DMN was initially 

                                                       
1
 Parts of Sections 2.2. and 2.3 are reproduced from Sullivan, Y. W., Davis, F. D., & Koh, C. (2015). Exploring mind 
wandering in a technological setting. Paper presented at the 36th International Conference on Information Systems 
(ICIS), Fort Worth, TX, December 13‐16. 


 
characterized as a network that became active during periods of rest and was absent during 

externally focused goal‐directed tasks (Raichle et al., 2001). However, recent findings reveal 

that the DMN is generally active when individuals are required to engage in relatively complex 

cognitive tasks, including tasks that often require individuals to retrieve episodic, 

autobiographical or semantic information, think about aspects of their personal goals, imagine 

novel ideas, infer the mental states of other people, comprehend narrative, self‐reflect, 

reference information to one’s self, etc. (Andrews‐Hanna et al., 2014). The activation of the 

DMN during mind wandering episodes suggests that this network is important in unconstrained 

internal thought processes (Barron, Riby, Greer, & Smallwood, 2011) and mind wandering 

potentially serves important psychological functions (Andrews‐Hanna et al., 2014).  

2.3   Content Regulation Hypothesis of Mind Wandering 

Evidence suggests that mind wandering is associated with unhappy moods and 

emotions (Killingsworth & Gilbert, 2010). Research also reveals that this association depends on 

the content of mind wandering experience (e.g., Smallwood & O’Connor, 2011). The hypothesis 

indicating that the content of mind wandering is crucial in determining its consequences is 

known as content regulation hypothesis. According to the content regulation hypothesis, “the 

relationship between self‐generated thought and psychological well‐being depends on 

assessing how individuals regulate the content of their mental experiences so as to maximize 

thoughts with a productive outcome, and minimize those which are detrimental to their 

happiness or other life outcomes” (Smallwood & Andrews‐Hanna, 2013, p. 4). Although most 

studies have focused on one or two aspects of mind wandering and their interactions, a few 


 
have assessed multiple types of thought content across a large group of individuals (Andrews‐

Hanna et al., 2014).  

Research exploring the content of mind wandering has demonstrated a number of 

general principles that has enhanced our understanding on mind wandering (Smallwood & 

Schooler, 2015). The summary of previous studies investigating the relationship between mind 

wandering and emotional outcomes is presented in Table 2.1. For example, using a thought 

sampling paradigm, Andrews‐Hanna et al. (2013) found that individuals with more negative and 

more personally significant thoughts scored higher on constructs associated with depression 

and trait negative affect, whereas those who characterized their thoughts as less specific scored 

higher on constructs related to rumination. Previous research has also shown that individuals 

who mind wandered about positive events, and about concurrent as opposed to past activities, 

were attributed to boredom and therefore, led to perceived dissatisfaction with an ongoing 

task (Critcher & Gilovich, 2010). Ruby, Smallwood, Engen, and Singer (2013) found that 

emotional content of mind wandering determined individuals’ moods. However, this direct 

relationship was modulated by the socio‐temporal content of thoughts—thoughts that were 

past‐ and other‐related were associated with negative moods, whereas future‐ and self‐related 

thoughts preceded improvements of mood, even when current thought content was negative. 

In sum, these previous studies suggest that (1) the outcomes of mind wandering depend on the 

content of mind wandering (Franklin et al., 2013) and (2) mind wandering can be characterized 

according to multiple interacting dimensions, including its personal significance, temporal 

orientation, valence, social orientation, and representational format (Andrews‐Hanna et al. 

2013). 

10 
 
Table 2.1 

Summary of Mind Wandering Studies in the Emotional Context (In a Chronological Order) 

Study  Cost Benefit


Smallwood,  Mind wandering was associated with poor retrieval  ‐
O’Connor, Sudbery,  of information and it was used as a marker for 
and Obonsawin  depressive thinking. 
(2007) 
Smallwood,  Negative mood reduced the amount of attentional 
Fitzgerald, Miles,  commitment to the primary task by enhancing the 
and Phillips (2009)  focus on task irrelevant personal concerns. 
Baird, Smallwood,  ‐  The reduction in mood associated 
and Schooler (2010)  with engaging in a boring task was 
attenuated when individuals engaged 
in mind wandering. 
Critcher and  Mind wandering about positive events, about ‐
Gilovich (2010)  concurrent activities, and about many events 
(rather than just one event) were attributed to 
boredom and thus, led to perceived dissatisfaction 
with an ongoing event.  
Killingsworth and  Mind wandering was negatively associated with  ‐
Gilbert (2010)  happiness.  
Mar, Mason, and  Mind wandering about people not close to us  Mind wandering about close others 
Litvack (2012)  predicted more loneliness and less perceived social  predicted greater life satisfaction. 
support.  
Andrews‐Hanna et  Participants who mind wandered about negative  Those who mind wandered about 
al. (2013)  and more personally significant events scored  positive, less personally significant, 
higher on constructs associated with Depression  and more specific contents scored 
and Trait Negative Affect.  higher on constructs associated with 
wellbeing and mindfulness. 
Franklin et al. (2013)  Mind wandering generally was associated with a  Mind wandering might have a positive 
more negative mood.  effect on mood when it was rated for 
interest. 
Poerio, Totterdell,  Mind wandering was closely associated with 
and Miles (2013)  negative mood, but it was not inherently 
detrimental to well‐being. 
Ruby et al. (2013)  Thoughts that were past and other‐related were  Future‐ and self‐related thoughts 
associated with subsequent negative mood, even if  were associated with improvements 
current thought content was positive.  of mood, even when current thought 
content was negative. 
Carciofo, Du, Song,  Mind wandering and negative affect yielded a bi‐ ‐
and Zhang (2014)  directional relationship. Specifically, (1) negative 
affect had both significant direct and indirect effect 
on mind wandering through sleep quality and (2) 
mind wandering had both significant direct and 
indirect effect on negative affect through sleep 
quality.  

11 
 
2.3.1   Technology‐related (IT) versus Non‐technology‐related (Non‐IT) Thoughts 

According to the diversion perspective, people remember things by encoding events in a 

specific physical environments, time, and socioemotional context (Delaney, Sahakyan, Kelley, & 

Zimmerman, 2010). These contextual features are encoded along with the event that is in the 

focus of attention, and they serve as retrieval cues during recall processes (Delaney et al., 2010; 

Howard & Kahana, 2002). Context can gradually drift as people mind wander because they 

mentally travel to and immerse themselves in the context of the event (Delaney et al., 2010). 

Such thought shifting will, in turn, influence individuals’ mental processes of retrieval of 

information. However, this effect of mind wandering experiences is influenced by mental 

distance between current on‐task thought and mind wandering contents. For example, Delaney 

et al. (2010) conducted two experiments to investigate the relationship between mind 

wandering and memory. They found that the contents of mind wandering that were more 

different from the current moment (e.g., distance, time, or circumstances) led to more 

forgetting than mind wandering experiences that were less different from the current moment, 

because they required a greater contextual shift (Delaney et al. 2010). Thus, based on the 

diversion paradigm, I postulate that when individuals work on a technology task, mind 

wandering about technology will lead to a less detrimental effect than mind wandering about 

other things.  

The content of mind wandering investigated by the previous studies has been primarily 

non‐technology‐related (e.g., temporal focus, positive and negative events). A recent study 

conducted by Hollis (2013) revealed that mind wandering about technology (i.e., thinking about 

or using another technology while engaging in a primary task) represented the highest level of 

12 
 
off‐task thinking among individuals and thus, this content of mind wandering should be 

included in future mind wandering research. As new technologies emerge and constantly 

outdate one another, it is difficult to imagine anyone living in a modern society without thinking 

about technologies in his/her everyday life. For example, when one is working on a financial 

report using Microsoft Excel, s/he could be thinking about checking his/her email. Thus, it is 

important to include technology as a content of thought in mind wandering research.  

As an effort to understand the complexity of thought in the context of interaction with 

technology, Ortiz de Guinea and Webster (2013) categorized the content of users’ thoughts into 

three groups: the computer system (e.g., thinking about how to insert a table), the task for 

which s/he is using the technology (e.g., writing a report), and something else (e.g., mind 

wandering). In their study, Ortiz de Guinea and Webster (2013) referred to the first category as 

computer‐related and the latter two as non‐computer‐related thoughts (see de Ortiz de Guinea 

& Webster, 2013 for details). However, they did not differentiate task‐unrelated thought or 

mind wandering from on‐task thought or task‐related thought. Task‐related thought or on‐task 

thought is theoretically and conceptually similar to cognitive focus or focus of attention (Lee, 

Sheldon, & Turban, 2003), which indicates the capacity to stay focused on the activity one is 

currently engaged in (Dimitrova, 2015). Because focus of attention reflects conscious effort and 

attention directed at task completion, it is generally associated with high task performance and 

problem solving (Randall et al., 2014).  

From a cognitive perspective, Orlikowski and Gash (1994) argued that understanding 

people interpretation of technology is crucial to understand their interaction with it. To interact 

with technology, people have to make sense of it. In this making‐sense process, they develop 

13 
 
subjective assumptions, expectations and knowledge of the technology, which then lead to 

shape subsequent actions toward it. In other words, technology impacts our cognitive thought 

processes. For instance, Ortiz de Guinea and Webster (2013) proposed that when using 

technology has become automatic, individuals tend to engage in non‐computer‐related 

thoughts during regular or expected IT events, whereas computer‐related thoughts will 

dominate individuals’ cognition during unexpected IT events. In order to facilitate the 

unexpected events, users will redirect their thoughts away from other activities (e.g., task) and 

direct their thoughts to the technology and thus, triggering computer‐related thoughts. As time 

passes and problems are solved, individuals are likely to resume their original tasks, triggering 

non‐computer‐related thoughts. Interestingly, these different contents of thought influence 

individuals’ task performance over a period of time. The question arises whether technology‐

related and nontechnology‐related thoughts lead to different cognitive outcomes. If so, we 

would speculate that the content of mind may work as a vehicle for performance outcomes. 

Therefore, given the content of mind wandering can have a distinct effect on cognitive and 

behavioral outcomes (McVay et al., 2009), I extend the content regulation hypothesis by 

categorizing mind wandering into two dimensions: technology related (IT)—task‐unrelated 

thought which occurs spontaneously and the content is related to the aspects of computer 

systems (e.g., email, social media, electronic devices)—and non‐technology related (non‐IT)—

task‐unrelated thought which occurs spontaneously and the content is unrelated to the aspects 

of computer systems.  

Oettingen and Schwörer (2013) discussed different forms of mind wandering and 

suggested that mind wandering in the form of mental contrasting (e.g., mind wandering about 

14 
 
future versus past events, personal versus other people concerns, etc.) led to skilled problem 

solving and substantial behavioral change. Given the perspective of mental contrasting 

potentially allows us to reveal some consequences of mind wandering, I classify the thoughts 

about affective content, temporal content, and self‐relevance into one category—mind 

wandering (non‐IT) and measure them as one dimension. This categorization is also consistent 

with the concept of cognitive interference proposed by Sarason, Pierce, and Sarason (2002). In 

the context of technology interaction, focus of attention can also be differentiated into two 

categories—IT and non‐IT thoughts (Ortiz de Guinea and Webster 2013). Focus of attention (IT) 

occurs when individuals think about the aspects of the computer systems related to the task at 

hand, whereas focus of attention (non‐IT) occurs when individuals spend thinking about the 

task, but these thoughts are not related to the aspects of computer systems (see Table 2.2). 

Table 2.2 

Construct Definition 

Dimension  Definition  Representative Thought 


Mind wandering (IT)  Task‐unrelated thought which occurs  Mind wandering about email, social 
spontaneously and the content is related to  media, and other online activities. 
the aspects of computer systems.  
Mind wandering  Task‐unrelated thought which occurs  Mind wandering about self 
(non‐IT)  spontaneously and the content is unrelated  (Andrews‐Hanna et al., 2013), other 
to the aspects of computer systems.  people (Ruby et al., 2013), past and 
future events (Baird, Smallwood, & 
Schooler, 2011), and emotional 
content (Poerio, et al., 2013). 
Focus of attention  Task‐related thought which occurs when  Task‐related thought about how to 
(IT)  individuals think about the aspects of the  use a technology and its various 
computer systems used to complete the  features (Ortiz de Guinea & 
primary task at hand.  Webster, 2013). 
Focus of attention  Task‐related thought which occurs when  Task‐related thought about how to 
(non‐IT)  individuals think about the primary task, but  solve a work problem and how to 
this thought is not related the aspects of  correctly perform as many steps as 
computer systems.   possible (Kanfer & Ackerman, 1989).  

15 
 
2.4  Emotional Outcomes  

  Emotions are “a mental state of readiness for action that promote behavioral activation 

and help prioritize and organize behaviors in a way that optimize individual adjustments to the 

demands of the environment” (Beaudry & Pinsonneault, 2010, p. 689). From the goal‐directed 

behavior perspective, the objective of emotions is to assist individuals with limited resources in 

the management of multiple goals under an uncertain environment (Bagozzi & Pieters, 1998). 

Positive emotions are associated with goal attainment, which usually leads to a decision to 

continue with the current plan, whereas negative emotions are the results of problems 

associated with ongoing plans and failures to achieve desired goals (Bagozzi & Pieters, 1998). In 

the context of technology use, technology artifacts can trigger emotional reactions from users 

when they interrupt users’ intended goals (Beaudry & Pinsonneault, 2010). Emotions are 

generally generated based on users’ perceptions of the features of new IT and on the use of 

new IT (Beaudry & Pinsonneault, 2010). Some of the emotions that have been extensively 

studied in IS research are computer anxiety (e.g., Compeau & Higgins, 1995; Thatcher & 

Perrewe, 2002; Venkatesh, 2000), satisfaction (e.g., Bhattacherjee, 2001; Doll, Xia, & 

Torkzadeh, 1994; Wixom & Todd, 2005), and enjoyment (e.g., Davis, Bagozzi, & Warshaw, 1992; 

Koufaris, 2002; Agarwal & Karahanna, 2000).  

  Emotions can be measured as either the outcomes of users’ perception or the 

antecedents of users’ attitudes and or behaviors (Beaudry & Pinsonneault, 2010). For example, 

studies have shown that general emotions, such as computer anxiety and computer 

playfulness, influence users’ attitudes and use of a specific technology (e.g., Compeau & 

Higgins, 1995; Compeau, Higgins, & Huff, 1999; Venkatesh, 2000). Other emotions, such as 

16 
 
satisfaction, have been associated with users’ perceived usefulness and perceived ease of use 

(Bhattacherjee, 2001). Emotions have also been associated with users’ response toward 

external events. Specifically, emotions are the unique psychological appraisal created by an 

individual as an effort to evaluate and interpret the events or circumstances (Bagozzi, Gopinath, 

& Nyer, 1999). For instance, emotional reactions are triggered when a new technology 

interrupts the sequence of events in one’s routine (Ortiz de Guinea & Webster, 2013).  

  Although the importance of emotions has been recognized in IS research, they have 

been studied under the assumption that emotions are triggered on the basis of external events 

or technology. Recently, mind wandering, as a stimulus‐independent thought, has been linked 

to various emotions, including satisfaction and happiness (Killingsworth & Gilbert, 2010). 

However, the association between mind wandering and emotions has not been investigated in 

IS research. Given users’ emotional reactions can be triggered by users’ internal thought 

processes, which are independent from external events or technology (i.e., generally the 

primary tasks), there is a need for research to investigate the occurrence of mind wandering 

and explain its consequences on users’ emotional outcomes. Thus, in this study, I sought to 

shed light on the impact of mind wandering as well as focus of attention on users’ emotional 

outcomes.  

Although negative emotions and moods have been shown to be the cause of mind 

wandering (Smallwood et al., 2009; Smallwood & O'Connor, 2011), time‐lag analyses conducted 

by Killingsworth and Gilbert (2010) strongly suggested that mind wandering was generally the 

cause, and not the consequence, of emotional outcomes. Thus, in this current study, mind 

wandering is regarded as a predictor of emotional outcomes associated with technology use—

17 
 
technology anxiety and user satisfaction. These two outcomes are chosen for their prevalence 

in IS research and because they represent two major categories of emotions (Beaudry & 

Pinsenneault, 2010). Technology anxiety refers to “anxiety about the implications of computer 

use such as the loss of important data or fear of other possible mistakes” (Thatcher & Perrewe, 

2002, p. 383) and user satisfaction is defined as users’ affect with (feelings about) the use of 

technology (Bhattacherjee, 2001). In IS research, technology anxiety is an important construct 

and negatively associated with IT use (e.g., Compeau & Higgins, 1995; Compeau et al., 1999). If 

a new technology is perceived to be a threat and users feel that they have some control over its 

consequences, anxiety will be experienced in this situation (Beaudry & Pinsonneault, 2010). In 

contrast, achievement emotions, such as satisfaction are the results of the appraisal of a new 

technology that will generate positive outcomes (Beaudry & Pinsonneault, 2010). Satisfaction is 

used as an evaluation of emotions experienced by users after they interacted with technology 

(i.e., whether IT use experience was as pleasurable as expected) (Bhattacherjee, 2001). 

Whereas technology anxiety is used to capture users’ negative emotions, satisfaction is used to 

measure users’ positive emotions. Given the critical role of users in information processing, 

understanding how users’ cognitive processes affect both users’ anxiety level and satisfaction 

with technology is an important step in appropriating benefits from current technology 

environments (Tarafdar, Tu, & Ragu‐Nathan, 2010).   

2.5   Research Model  

As shown in Figure 2.1, the current model aims to explain the effect of mind wandering 

and focus of attention on users’ emotional outcomes. As discussed previously, I classify the 

18 
 
content of mind into four categories: (1) mind wandering (IT); (2) mind wandering (non‐IT); (3) 

focus of attention (IT); and (4) focus of attention (non‐IT). Consistent with the content‐

regulation hypothesis, I hypothesize that each category of mind will yield a different effect on 

users’ emotional outcomes. Consistent with prior cognitive theories (e.g., Kanfer and 

Ackerman’s (1989) cognitive resource theory), I conceptualize that mind wandering and focus 

of attention are distinct processes and only modestly correlated constructs (Randall et al., 

2014). Hence, they will also yield different results on the associated outcomes.  

Figure 2.1. Research Model  

2.6  Hypotheses Development 

2.6.1   Mind Wandering and Technology Anxiety 

  The negative implications of mind wandering have been widely reported across 

different task contexts (see Table 2.1). A hypothesis that mind wandering is associated with 

emotional costs is built upon a longitudinal study conducted by Killingsworth and Gilbert 

(2010). Using an experience sampling technique, they contacted participants through their 

19 
 
mobile devices at random moments during the day and asked them to report their mind 

wandering experiences and their state of emotions. They found that people are less happy 

when their minds were wandering, regardless their activities. It is likely that “what people were 

thinking” was a better predictor of their emotional outcomes than “what they were doing” 

(Killingsworth & Gilbert, 2010).  

Klinger (1994) investigated the relationships between test anxiety, preparation, thought 

content, and academic performance during a college exam. During a course of exam, students 

were signaled by auditory probe to stop working and record the content of their thoughts. They 

found that mind wandering had no significant negative effect on performance, but had a 

significant positive effect on test anxiety (see Klinger, 1994). People who experience mind 

wandering often describe themselves as troubled and unhappy about their inability to become 

absorbed in significant aspects of their lives (Sarason, Sarason, Keefe, Hayes, & Shearin, 1986). 

Building on these previous works, I hypothesize that mind wandering (both IT and non‐IT) 

positively influences technology anxiety.   

According to the content regulation hypothesis, mind wandering can be associated with 

either negative or positive emotions, depending on the content of thought when the mind 

wanders (Smallwood & Schooler, 2015). As discussed previously, this hypothesis suggests that 

the cost of such mind wandering experiences can only be properly understood by taking into 

account the content of the experiences (Smallwood & Schooler, 2015). In line with this 

hypothesis, I predict that individuals who are engaged in either technology or non‐technology 

thoughts would exhibit different consequences on emotional outcomes of technology use. 

Given mind wandering (IT) discussed in this study primarily focuses on the thoughts about 

20 
 
technology unrelated to the task at hand, users who wander about these contents tend to have 

a positive attitude toward technology. It could be an indication that they use technology in a 

regular basis and are less intimidated by a new technology (Tu, Tarafdar, Ragu‐Nathan, & Ragu‐

Nathan, 2007). To a certain degree where users are becoming familiar with the technology, the 

thought process can occur unconsciously—it will demand minimum cognitive resources (Ritter 

& Dijksterhuis, 2014). However, thinking about other things unrelated to the technology may 

require more cognitive resources than thinking about technology. Thus, I hypothesize that 

although both mind wandering (IT) and (non‐IT) are positively associated with technology 

anxiety, the effect of mind wandering (non‐IT) on technology anxiety is stronger than the effect 

of mind wandering (IT) on technology anxiety.   

H1a: Mind wandering (IT) is positively associated with technology anxiety.  
 
H1b: Mind wandering (non‐IT) is positively associated with technology anxiety. 
 
H1c: The association between mind wandering (non‐IT) and technology anxiety is 
stronger than between mind wandering (IT) and technology anxiety. 
 

2.6.2  Focus of Attention and Technology Anxiety 

  Brockner (1979) suggested that the manipulations of attentional focus can produce 

different levels of anxiety. By directing individuals to pay attention to the relevant task or 

technology, technology anxiety can be inhibited as cognitive resources are allocated to the 

relevant stimuli. Technology anxiety occurs when users feel apprehensive and anxious about 

using a new technology (Chua, Chen, & Wong 1999). When individuals’ focus is narrowed to a 

limited stimulus field, irrelevant thoughts and perceptions, including technology anxiety are 

filtered out (Trevino & Webster, 1992). When users interact with technology, the elements of 

21 
 
technology may serve to help focus the individual’s attention on a limited stimulus field 

(Trevino & Webster, 1992). Thus, focus of attention (IT) helps users become more absorbed in 

their activity and in turn, reduce their anxiety level.  

  In addition to focus on technology‐related thoughts, users also engage in non‐

technology‐related thoughts. At the early stage of technology use, users tend to focus on 

navigating different features of technology. Thus, their focus of attention is on technology. 

Once the skills become automatic, then their focus of attention can be allocated to higher levels 

of abstraction, such as problem solving strategies (Kanfer & Ackerman, 1989). Such focus of 

attention is externally oriented and is concerned with maximizing goal‐directed behavioral 

outcomes (Robinson & Tamir, 2011). Because focus of attention is likely to support higher levels 

of approach motivation, it is conducive to positive emotions and hinders negative emotions 

(Robinson & Tamir, 2011).  

  Theory and research suggests that anxiety will trigger psychological distancing—

directing one’s attention away from the situation (Beaudry & Pinsonneault 2010). Thus, anxiety 

about the prospective negative consequences of a new IT is likely to trigger psychological 

distancing (Beaudry & Pinsonneault, 2010). However, when one’s attention is narrowly directed 

to technology‐relevant cues (relative to task‐relevant cues), individuals will have more 

confidence in their ability to learn about the technology. Consequently, focusing one’s 

attention on technology‐related cues will lead to lower technology anxiety than on non‐

technology‐related cues, at least during the early stage of technology use.  

H2a: Focus of attention (IT) is negatively associated with technology anxiety.  
 
H2b: Focus of attention (non‐IT) is negatively associated with technology anxiety. 
 

22 
 
H2c: The negative association between focus of attention (IT) and technology anxiety is 
stronger than between focus of attention (non‐IT) and technology anxiety. 
 

2.6.3   Mind Wandering and User Satisfaction 

  In the context of human‐computer interaction, users’ satisfaction is usually associated 

with one’s feelings or attitudes toward a variety of factors affecting that situation (Bailey & 

Pearson, 1983). According to the user satisfaction literature, user satisfaction is typically viewed 

as ‘an object‐based attitude’ (Wixom & Todd, 2005) and has been measured by various subsets 

of beliefs including flexibility, accuracy, timeliness, reliability, and other related factors (DeLone 

& McLean, 1992, 2003; Wixom & Todd, 2005). Expectation‐confirmation theory (ECT)—a theory 

used to study consumer satisfaction—proposes that users form an initial expectation of a 

specific system before they decide to use the system. After they used the system, they form 

perceptions about its performance and then assess its perceived performance in comparison to 

their original expectation and determine the degree to which their expectation is confirmed. 

They form a satisfaction based on their confirmation level and expectation on which that 

confirmation is created (Bhattacherjee, 2001). This theory assumes that satisfaction is 

determined by users’ deliberate mental processes in comparing their initial perceptions with 

their actual experience. In order to successfully perform such mental processes, users’ focus of 

attention is expected to be on‐task. Thus, any mental distractions, including mind wandering 

activities will interfere with users’ ability to accurately evaluate the system.  

Although the effect of mind wandering on user satisfaction has not been empirically 

tested, mind wandering has been associated with dissatisfaction with one’s life (Mar et al., 

2012). Mind wandering is also considered a manifestation of frustration and stress (Mar et al., 

23 
 
2012). Given mind wandering occurs when individuals engage in other activities, it might be 

used as a cue to understand one’s attitude toward an ongoing activity (Critcher and Gilovich, 

2010). The specific content of users’ mind wandering experiences can be informative in 

signaling their feelings toward their ongoing activities (Critcher & Gilovich, 2010). Given 

satisfaction is the result of the appraisal of the upcoming event and has been associated with 

technology use (Bhattacherjee, 2001; Wixom & Todd, 2005), I postulate that when a wandering 

mind focuses on irrelevant technology‐related activities, it may be interpreted as a signal of 

boredom or dissatisfaction with the current systems because such thoughts represent a coping 

mechanism to the problem at hand (Critcher & Gilovich, 2010). In contrast, mind wandering 

that focuses on other issues unrelated to technology may be seen as less informative—as idle 

thoughts that can have any types of thoughts (Critcher & Gilovich 2010). In that sense, I 

hypothesize that mind wandering (IT) has a stronger negative effect on user satisfaction than 

mind wandering (non‐IT).  

H3a: Mind wandering (IT) is negatively associated with user satisfaction.  
 
H3b: Mind wandering (non‐IT) is negatively associated with user satisfaction. 
 
H3c: The negative association between mind wandering (IT) and user satisfaction is 
stronger than between mind wandering (non‐IT) and user satisfaction. 
 

2.6.4  Focus of Attention and User Satisfaction 

  There is an implicit assumption that users will have a higher level of satisfaction if their 

interaction with the system is incorporated with specific organizational and psychological 

parameters (e.g., users’ cognitive processes) (Mahmood, Burn, Gemoets, & Jacquez, 2000). 

From a disruption literature perspective, interruptions reduce the satisfaction of users with the 

24 
 
systems they employ for their tasks and their ability to benefit from them (Tarafdar, Tu, Ragu‐

Nathan, & Ragu‐Nathan, 2011). It is likely because under a high complexity condition, users 

perceive focus of attention (i.e., cognitive effort) as a major determinant of both performance 

and satisfaction. Thus, users are primarily interested in finding and processing information 

relevant to clarifying task demands (Butler, 1993).  

  According to the cognitive resource literature, when users first encounter a new 

technology, the attentional load on users is relatively high (Kanfer & Ackerman, 1989). As users 

begin to understand the demands of the task, they start learning how to use the systems by 

allocating their attention to the detailed procedures associated with systems use. During this 

phase, focus of attention (IT) is likely to have higher influence on user satisfaction than focus on 

other subjects unrelated to the system (e.g., focus on the strategy of how to solve the task 

more effectively). Given the scope of the study is at the initial phase of technology use, I 

hypothesize that both focus of attention (IT) and (non‐IT) have positive effects on user 

satisfaction. However, the effect of focus of attention (IT) is higher than (non‐IT), at least during 

the early stage of technology use.  

H4a: Focus of attention (IT) is positively associated with user satisfaction.  
 
H4b: Focus of attention (non‐IT) is positively associated with user satisfaction. 
 
H4c: The positive association between focus of attention (IT) and user satisfaction is 
stronger than between focus of attention (non‐IT) and user satisfaction. 
 

2.6.5  Technology Anxiety and User Satisfaction 

  User satisfaction is a critical variable because it is considered a reflection of the success 

of the system (DeLone & McLean, 1992, 2003). Whereas users having positive cognitions about 

25 
 
a system leads to greater satisfaction with the system, technology anxiety leads to low user 

satisfaction (Igbaria & Nachman, 1990; Tarafdar et al., 2011). Anxiety is a result of mental 

pressure and has been associated with lower performance on cognitive‐demanding tasks. 

Limited use and dissatisfaction with the system may be associated, in part, with users’ lack of 

knowledge on how the system works (Igbaria & Nachman, 1990). When users do not have 

sufficient knowledge on the new technology, they tend to exhibit anxieties, leading to lower 

performance and less knowledge retention than those with low technology anxiety (Webster, 

Heian, & Michelman, 1990). Therefore, it is reasonable to argue that technology anxiety is one 

of the major factors influencing user satisfaction. Thus, I hypothesize 

  H5: Technology anxiety is negatively associated with user satisfaction. 

2.6.6.   Control Variables 

  The following variables are controlled for in this study: 

 Perceived Task Complexity. Technology tasks require necessary computer knowledge, 

skills, and extent of time and efforts needed to handle the task successfully (Chang, 

2005). Users’ perception of task complexity is described as “relative” because one 

may perceive a computer task to be of certain complexity, whereas others may or 

may not perceive the same degree of complexity (Chang, 2005). In this study, I 

predict that user perception of technology task complexity is positively correlated 

with technology anxiety. According to Rogers’ diffusion of innovation’s model (1995), 

perceived complexity is defined as the degree to which an innovation is perceived as 

26 
 
difficult to understand and use. Thus, the complexity of an innovation is negatively 

associated with adoption (Rogers, 1995). 

 Technology Multitasking. According to the attention control perspective, mind 

wandering and external distraction (e.g., technology multitasking) should be 

positively correlated such that individuals who experience more mind wandering 

should also experience more external distraction in situations where attention 

control is needed to maintain task goals (Forster & Lavie, 2014; Ralph, Thompson, 

Cheyne, & Smilek, 2014). Given technology multitasking consumes the limited 

cognitive resources, it will increase technology anxiety, especially when users are 

engaged in mind wandering. I control for the effect of technology multitasking on 

technology anxiety.  

 Demographic Information. Finally, I also control for some demographic variables, 

including age and gender.  

2.7   Research Method2 

  Mind wandering may have been overlooked by many researchers out of concern that it 

is too difficult to study (Schooler, Reichle, & Halpern, 2004). However, a number of studies have 

validated self‐report measures of mind wandering (e.g., Mrazek, Phillips, Franklin, Broadway, & 

Schooler, 2013) and demonstrated that they are reliable to predict various outcomes, such as 

performance error (e.g., Sarason et al., 2002; Smallwood et al., 2004) and emotional outcomes 

                                                       
2
 Parts of Sections 2.71, 2.7.2, and 2.8.1 are reproduced from Sullivan, Y. W., Davis, F. D., & Koh, C. (2015). 
Exploring mind wandering in a technological setting. Paper presented at the 36th International Conference on 
Information Systems (ICIS), Fort Worth, TX, December 13‐16. 

27 
 
(Killingsworth & Gilbert, 2010). Thus, in this current study, a self‐report technique is used to 

measure the occurrence of mind wandering and focus of attention during an actual IS task—a 

task that requires a certain degree of computer skills. 

2.7.1  Instrumental Development 

After constructing the conceptual framework, I developed and validated the measures 

for each thought category described in the model. When possible, the measurement items of 

the constructs were developed based on existing scales in extant literature that have been 

proven reliable; otherwise, I developed new measures by following the procedures proposed by 

MacKenzie et al. (2011). Although there has been a number of studies on mind wandering, this 

topic can be considered new in IS research and thus, a standardized metric to measure mind 

wandering (IT) and focus of attention (IT) is not available. To address this problem, I developed 

mind wandering and focus of attention metrics that can be used to measure causal mechanisms 

of attention and their potential outcomes more precisely. After I developed a conceptual 

definition of the new constructs (see the previous section), I generated the survey items from 

the theoretical definitions and a review of literature. Following this, I assessed the content 

validity of the measurement items. After that, I collected data to conduct a small pretest in 

order to purify and refine the items. I collected a new data set to test the convergent and 

discriminant validity of the scales. The same data set was used to test the structural model.  

Based on a review of previous relevant research in the IS discipline (e.g., Ortiz de Guinea 

& Webster, 2013), I generated seven items to measure focus of attention (IT) and seven items 

to measure mind wandering (IT). Consistent with the previous theories on mind wandering 

28 
 
(e.g., Sarason et al., 2002), all constructs were measured reflectively on seven‐point Likert 

scales. Items to measure focus of attention (non‐IT) were adapted from Kanfer, Ackerman, 

Murtha, Dugdale, and Nelson (1994) and items to measure mind wandering (non‐IT) were 

adopted from Sarason et al. (1986)3. Furthermore, items to measure technology anxiety were 

adapted from Kanfer et al. (1994) and Thatcher and Perrewe (2002) and satisfaction was 

measured using Nadkarni and Gupta’s (2007) satisfaction scale. Items used to measure 

perceived task complexity were adapted from Winter and Latham (1996), Seijts, Latham, Tasa, 

and Latham (1986), and Jiang and Punj (2010).  

Content validity of the items was obtained in two different stages. In the first, after I 

generated the items, I subjected these new items, along with the other items in the survey, to a 

card sorting exercise (Polites & Karahanna, 2012). Seven graduate students participated in this 

exercise. At the end of the exercise, I asked the participants to report any issues pertaining to 

the wording of the items. The sorting resulted in a satisfactory classification of items into my 

predefined categories. In the second stage, I collected data from a small sample size (n = 32) to 

                                                       
3
 Given longer surveys take more time to complete and tend to have more missing data than short surveys, I 
decided to shorten the CIQ questionnaire proposed by Sarason et al. (1986). First, items were chosen based on 
their factor loading scores reported in their original scales (Stanton, Sinar, Balzer, & Smith 2002). Second, items’ 
clarity of expression and their relevance to our target population were taken into account in deciding which items 
should remain in the survey (Stanton et al. 2002). For example, the first 10 items of the CIQ questionnaire were not 
included to measure mind wandering because these items refer to aspects of self‐evaluation (Sarason et al., 1986) 
and are more relevant to measure self‐regulatory activities (see Kanfer & Ackerman, 1989). Sarason et al. (1986) 
indicated that the remaining items of the CIQ questionnaire are the actual items used to measure a diversity of 
thoughts unrelated to the task. Given I specifically focus on the content of mind wandering, item 11 (i.e., I thought 
about other activities, such as assignment and work) was also dropped from the survey. Item 16 (i.e., I thought 
about something that made me feel tense) was also excluded from the actual survey due to its redundancy with 
item 15. Results indicate that the 9‐item version possesses acceptable psychometric properties when used to 
measure the state of mind wandering. I, therefore, conclude that the threat to content validity for using a short 
version of the CIQ questionnaire is not a significant factor in this current study. 

29 
 
test whether my survey items are consistent with the definitions of each thought. Based on the 

feedback, the measurement items were modified.  

2.7.2  Data Collection   

An important part of the research design in the current study is selecting an IS task that 

demands cognitive attention. I investigate whether mind wandering occurs when participants 

engage in novel tasks. I thus selected two basic IS courses that require students to use 

computer software in their projects (i.e., either Microsoft Excel or PowerPoint). An online 

questionnaire was administered to a sample of undergraduate students taking  an introductory 

IS course at a large university in the United States. Participants received extra credit as an 

exchange for their participation. An alternative assignment was provided to those not wishing 

to participate in the study. I was granted a permission to visit the classes and gave a short 

presentation about the study. Participants were instructed to fill out the questionnaire right 

after they did their computer assignment for the course. The course instructors were notified to 

send out the survey link by the assignment deadline. At the end of the study, a total of 406 valid 

responses were collected. Demographic information on the respondents is presented in Table 

2.3. 

2.8   Data Analysis and Results 

2.8.1  Assessment of Measurement Validation 

The measurement and structural models were tested using a structural equation 

modeling method. In order to establish the convergent and discriminant validity of the scales, 

30 
 
two separated steps were conducted. In the first step, I conducted an exploratory factor 

analysis (EFA) using a direct oblimin rotation method to identify the underlying relationships 

among 4 independent variables. This rotation method was selected because the factors in the 

analysis were expected to be correlated (Gorsuch, 1983; Thompson, 2004). Next, I examined 

the convergent and discriminant validity of the mind wandering and focus of attention 

measures at the item‐ and construct‐level using confirmatory factor analysis (CFA) procedures 

in Lisrel 8.8. 

Table 2.3 

 Sample Characteristics 

Variable  Category  Frequency 


Gender  Male   199 (49.0 %) 
Female  207 (51.0 %) 
Language  English  344 (84.8 %) 
Else  61 (15.0 %) 
Missing  1 (0.2%) 
Right‐Handed  Right‐handed  358 (88.2 %) 
Left‐handed  47 (11.6 %) 
Missing  1 (0.2 %) 
Age  Younger than 20  203 (50.0%) 
21 – 25   165 (40.6 %) 
26 – 30   16 (3.9 %) 
31 – 35   11 (2.7 %) 
Older than 35   9 (2.2 %) 
Missing  2 (0.5 %) 
 

The EFA result (the pattern matrix) is presented in Table 2.4. Based on the results of 

EFA, two items of mind wandering (IT) were dropped. As predicted, the final items of each 

category of thought loaded higher into their theoretical constructs than into other constructs. 

Together, four factors extracted accounted for 75.13 percent of the variance (see Table 2.4).   

31 
 
Table 2.4 
EFA Result 

Component 
Item  Mean  SD  1  2  3  4 
MW_IT1  3.83  1.79  0.13  0.00  ‐0.71  0.06 
MW_IT2  4.19  1.79  ‐0.02  0.02  ‐0.92  ‐0.02 
MW_IT3  4.33  1.70  0.03  ‐0.04  ‐0.92  0.02 
MW_IT4  4.67  1.63  ‐0.05  0.02  ‐0.90  ‐0.03 
MW_IT5  4.07  1.84  0.02  ‐0.06  ‐0.88  0.05 
MW_NON_IT1  3.91  1.63  0.82  0.06  ‐0.04  ‐0.05 
MW_NON_IT2  3.93  1.63  0.80  0.07  ‐0.15  ‐0.06 
MW_NON_IT3  3.30  1.85  0.89  ‐0.08  0.11  0.09 
MW_NON_IT4  4.09  1.75  0.82  0.05  ‐0.06  ‐0.12 
MW_NON_IT5  3.37  1.83  0.87  ‐0.08  0.04  0.07 
MW_NON_IT6  3.95  1.67  0.81  0.06  ‐0.11  ‐0.04 
MW_NON_IT7  3.81  1.73  0.90  0.04  ‐0.01  ‐0.02 
MW_NON_IT8  3.41  1.84  0.89  ‐0.09  0.07  0.10 
MW_NON_IT9  4.03  1.76  0.80  0.03  ‐0.09  ‐0.04 
FOA_IT1  4.95  1.47  0.02  0.87  0.02  ‐0.10 
FOA_IT2  4.86  1.37  ‐0.01  0.92  ‐0.04  ‐0.03 
FOA_IT3  4.78  1.42  0.06  0.82  ‐0.02  0.11 
FOA_IT4  4.80  1.41  0.03  0.82  0.02  0.11 
FOA_IT5  4.92  1.46  ‐0.03  0.78  ‐0.03  0.11 
FOA_IT6  5.00  1.40  ‐0.07  0.79  0.04  0.10 
FOA_IT7  5.14  1.39  0.00  0.84  0.06  ‐0.03 
FOA_NON_IT1  4.66  1.52  0.06  ‐0.07  0.01  0.84 
FOA_NON_IT2  4.74  1.49  0.02  0.12  0.00  0.84 
FOA_NON_IT3  4.83  1.46  ‐0.03  0.09  ‐0.03  0.85 
FOA_NON_IT4  4.72  1.63  ‐0.06  0.07  ‐0.05  0.78 
Notes: MW_IT = Mind Wandering (IT); MW_NON_IT = Mind Wandering (Non‐IT); FOA_IT = Focus of Attention (IT); 
FOA_NON_IT = Focus of Attention (Non‐IT). 
 

  Further, I conducted a CFA to examine the adequacy of the measurement model. The 

first model was a four‐factor model (see Table 2.5). Results indicate high levels of convergent 

and discriminant validity. As shown in Table 2.5, the hypothesized four‐factor model fit the data 

significantly better than other alternative measurement models. Finally, goodness‐of‐fit indices 

for the four‐factor model met the criteria suggested by Hu and Bentler (1999). Based on the 
32 
 
results of this validation process, all of the measurement items were considered sufficient, 

suggesting that categorizing mind wandering and focus of attention into IT and non‐IT (four‐

factor model) shows better goodness of fit than combining IT and non‐IT thoughts into one 

factor (two‐factor model or one‐factor model). 

After the validation of mind wandering and focus of attention measures was 

established, the CFA was constructed as a six‐factor model wherein the measures of technology 

anxiety and user satisfaction constructs were modeled as separate constructs. In the last CFA, I 

constructed an eight‐factor model by adding the control variables (i.e., perceived task 

complexity and media multitasking) in addition to the dependent variables. As hypothesized, 

the measurement model demonstrates a sufficient level of discriminant validity. 

Table 2.5 

Summary of Model Fit for All CFAs Examined 

Description  X2  df  X2/df  ΔX2/Δdf  CFI  IFI  SRMR  RMSEA 


4‐factor model (hypothesized 
615.63  262  2.34  ‐  .98  .98  .04  .06 
model) 
3‐factor model (mind 
1393.20  265  5.25  259.19***  .95  .96  .06  .10 
wandering combined) 
3‐factor model (on‐task 
2217.97  265  8.37  534.11***  .94  .94  .07  .13 
thought combined) 
2‐factor model (mind 
wandering and focus of  2994.94  267  11.21  475.86***  .91  .91  .09  .15 
attention combined) 
1‐factor model (all combined)  8060.40  268  30.07  1240.80***  .79  .79  .25  .27 
6‐factor model without control 
1192.72  535  2.23  ‐  .98  .98  .05  .05 
variables (hypothesized model) 
8‐factor model with control 
1554.97  666  2.33  ‐  .98  .98  .05  .06 
variables (hypothesized model) 
Notes: df = degree of freedom; CFI = Comparative Fit Index; SRMR = Standardized Root Mean Residual; IFI = 
Incremental Fit Index; RMSEA = Root Mean Square Error of Approximation; ***p<.001. 
   

33 
 
  The results of CFA of the eight‐factor model, reported in Table 2.6, indicate high levels 

of convergent validity. Each factor loading is strong and significant (all p values < .01), and the 

average variance extracted (AVE) value for each dimension of thought is well above Fornell and 

Larcker’s (1981) suggested cutoff of .50. Further, the composite reliability (CR) values for all of 

the constructs are greater than .70, demonstrating that all constructs have adequate reliability 

scores. The means, standard deviations, and correlations are reported in Table 2.7. As shown in 

Table 2.7, the square root of the AVE for each construct in the model is larger than the 

corresponding off‐diagonal correlations of the constructs to their latent variables. These results 

provide substantial evidence for the convergence/discriminant validity of the items and overall 

measurement model. 

Table 2.6 

Summary of Confirmatory Factor Analysis Result (8‐Factor Model) 

Factor 
Constructs/Items 
Loading 
Mind wandering (non‐IT) (CR; AVE)  (.96; .74)
1. I thought about members of my family. .79
2. I thought about friends.  .86
3. I thought about something that made me feel guilty. .75
4. I thought about personal worries.  .83
5. I thought about something that made me feel angry. .79
6. I thought about something that happened earlier today. .87
7. I thought about something that happened in the recent past (last few days, but not today).  .89
8. I thought about something that happened in the distant past. .81
9. I thought about something that might happen in the future.  .85
Mind wandering (IT) (CR; AVE)  (.94; .78)
1. I thought about checking my email.  .75
2. I thought about checking my social media (e.g., Facebook). .88
3. I thought about browsing other stuff. .94
4. I thought about checking my phone.   .82
5. I thought about doing other online activities (e.g., online shopping, online game).  .87
(table continues) 
 
 

34 
 
Table 2.6 (continued) 

Factor 
Constructs/Items 
Loading 
Focus of Attention (IT) (CR; AVE)  (.95; .73)
1. I focused on how to use the software to complete my task. .78
2. I paid attention on different features of the software to complete my task. .91
3. I focused my attention on different options provided by the software to format my task.   .89
4. I focused my attention on learning specific the software functions. .87
5. I paid close attention to the kind of errors I was making when I was using the software.   .80
6. I focused my total attention on making sure my computer outputs were correct.  .77
7. I focused my total attention on following the software’s instructions correctly. .74
Focus of Attention (non‐IT) (CR; AVE)  (.92; .74)
1.  I focused my total attention on how I could solve the problem. .69
2. I thought about strategies for solving the problem. .91
3. I thought ahead to what I would do next to solve the problem. .91
4. I focused my attention on correctly performing as many steps as I could. .74
Technology Anxiety (CR; AVE)  (.95; .77)
When you were doing your assignment, to what extent do you agree with these statements? 
1. I became frustrated with my inability to solve the problem using the software.  .80
2. I got mad at myself when I used the software.  .86
3. I was concerned about how poorly I was doing using this software. .87
4. I felt anxious about using this software to complete my task. .84
5. I was worried about how I was going to make a mistake by choosing the wrong options.  .84
6. I was worried about how I was going to make a mistake that I cannot correct when I used the  .82
software. 
User Satisfaction (CR; AVE) (.95; .86)
After doing your assignment using this software, how do you feel about your overall experience 
with the software?   
1. …I am very dissatisfied / very satisfied.  .95 
2. …I am very displeased / very pleased. .95
3. …I feel absolutely terrible / very delighted. .88
4. …I will strongly recommend this software to my friends. .80
Perceived Task Complexity (CR; AVE)  (.87; .68)
    1. This task required me to coordinate many different things at the same time. .66
2. This task required a lot of my effort.  .84
3. On scale 1 (extremely easy) to 7 (extremely difficult), how difficult was this task?  .71
Media Multitasking  ‐
A single item asking participants whether they used any electronic devices/media/software  1.00 
(besides their computer or laptop and the software they used to do their assignment). 
1. Survey Instructions: The following statements concern the kinds of thoughts that go through people’s heads at 
particular times. For example, while they are working on a task or assignment, they could think about something else, 
unrelated to their primary task. The following is a list of thoughts, some of which you might have had while doing 
your task. Please indicate approximately how often each thought occurred to you when you did your task. 
2. Scale to measure the thought content: 1 (never) – 7 (every Time); other measures: 1 (strongly disagree) – 7 
(strongly agree). 
3. CR = Composite Reliability; AVE = Average Variance Extracted. 

35 
 
Table 2.7  

Correlation Matrix and Descriptive Statistics 

   Construct  Mean  SD  1  2  3  4  5  6  7  8 


1  Mind Wandering (IT)  4.21  1.54  .88                  
2  Mind Wandering (Non‐IT)  3.76  1.49  .61 .86                
3  Focus of Attention (IT)  4.92  1.21  .04 ‐.06  .85             
4  Focus of Attention (Non‐IT)  4.73  1.31  .09 .07  .60 .86           
5  Tech. Anxiety  3.49  1.52  .45 .58  ‐.15 .10  .88        
6  Satisfaction  4.74  1.28  ‐.10 ‐.13  .46 .40  ‐.20  .93     
7  Complexity  4.37  1.13  .19 .22  .24 .24  .41  .15 .82   
8  Multitasking  3.19  0.95  .24 .17  ‐.00 .01  .02  .01 .00 1.00 
Note: Diagonal elements display the square root of AVE. 

  Further, common method bias can be a potential threat to the study if the independent 

and dependent variables are obtained from the same sources (Podsakoff, MacKenzie, Lee, & 

Podsakoff, 2003). In order to determine whether common method bias was a concern, I used 

several different methods. First, I performed the Harman’s single‐factor test suggested by 

Podsakoff et al. (2003). The result indicated that one factor only accounted for 28.6 percent of 

the variance, suggesting no common method bias. Next, I performed a common method 

variance test with a common latent factor (Podsakoff et al., 2003). I ran a CFA in AMOS that 

included a method construct. This allowed not only comparison of the loadings of each item on 

its own factor and the method factor, but also calculation of the amount of method bias 

present in the entire dataset (Polites and Karahanna, 2012). The estimated amount of method 

bias was only 2.8 percent. Further, all factor loadings remained significant in the presence of 

the method factor (see Appendix A). Lastly, I perform a marker variable technique suggested by 

Lindell and Whitney (2001). I examined the matrix of item‐to‐item correlations. The results 

showed that the correlation scores between the marker variable and items from construct 

36 
 
hypothesized range from ‐.01 to .12 (with only two significant correlations). This indicates the 

absence of common method bias.  

2.8.2   Structural Model 

The measurement of the structural model was estimated using the SEM approach. The 

overall fit indices for the structural model were acceptable (X2 =  1561.15, df = 668; X2/df  =  

2.36; CFI = .98; IFI = .98; SRMR = .05; RMSEA = .057). The results of the model estimates, 

including standardized path coefficients and t‐values, are presented in Figure 2.2. Both mind 

wandering (IT) and mind wandering (non‐IT) had significant, negative effect on technology 

anxiety, supporting H1a and H1b. Contrary to my hypotheses, only focus of attention (IT) had a 

negative effect on technology anxiety, whereas the effect of focus of attention (non‐IT) was not 

significant. Thus, H2a was supported and H2b was not supported. The effects of mind 

wandering (IT) and mind wandering (non‐IT) on user satisfaction were not significant, failing to 

support H3a and H3b. The effect of technology anxiety on user satisfaction was also significant, 

supporting H5.  

In order to test H1c, H2c, H3c, and H4c, I explore the relationships by performing a X2 

difference test using Lisrel between the baseline and constrained model. For instance, to test 

H1c, the path coefficients from mind wandering (IT) and mind wandering (non‐IT) to technology 

anxiety were constrained to .12 (the actual path coefficient of mind wandering (IT) to 

technology anxiety in the baseline model). If the X2 difference test is significant, then the 

baseline model would have a better fit in reflecting the hypothesized relationship. 

37 
 
 

Figure 2.2. Structural Model (t‐statistic in parentheses) 
 

The results of X2 difference test are presented in Table 2.8. As hypothesized, the positive 

association between mind wandering (non‐IT) and technology anxiety was greater than 

between mind wandering (IT) and technology anxiety. Thus, H1c was supported. Similarly, the 

negative association between focus of attention (IT) and technology anxiety was greater than 

between focus of attention (non‐IT) and technology anxiety, supporting H2c. The positive 

association between focus of attention (IT) and user satisfaction was also greater than between 

focus of attention (non‐IT) and user satisfaction. Thus, H4c was supported. The summary of the 

hypothesis results are presented in Table 2.9. 

   

38 
 
Table 2.8 

Chi‐Square, Goodness‐of‐Fit Values, and Model Comparison Tests 

Description  X2  df  ΔX2  Δdf  CFI  IFI  SRMR  RMSEA 


Baseline  1466.59  668  ‐  ‐  .98  .98  .05  .054 
Constrained Model 1 (H1c)  1491.68  670  25.09***  2  .98  .98  .06  .055 
Constrained Model 2 (H2c)  1496.16  670  29.57***  2  .98  .98  .06  .055 
Constrained Model 3 (H4c)  1483.38  670  16.79***  2  .98  .98  .06  .055 
Notes: *** p<.001; Because H3a and H3b were not significant, H3c was not tested in this analysis. 

Table 2.9 

Hypotheses Results  

  Hypothesis  Supported? 
H1a  Mind wandering (IT) is positively associated with technology anxiety.  Y 
H1b  Mind wandering (non‐IT) is positively associated with technology anxiety.  Y 
H1c  The association between mind wandering (non‐IT) and technology anxiety is  Y 
stronger than between mind wandering (IT) and technology anxiety. 
H2a  Focus of attention (IT) is negatively associated with technology anxiety.  Y 
H2b  Focus of attention (non‐IT) is negatively associated with technology anxiety.  N 
H2c  The association between focus of attention (IT) and technology anxiety is stronger  Y 
than between focus of attention (non‐IT) and technology anxiety. 
H3a  Mind wandering (IT) is negatively associated with user satisfaction.  N 
H3b  Mind wandering (non‐IT) is negatively associated with user satisfaction.  N 
H3c  The negative association between mind wandering (IT) and user satisfaction is  N 
stronger than between mind wandering (non‐IT) and user satisfaction. 
H4a  Focus of attention (IT) is positively associated with user satisfaction.  Y 
H4b  Focus of attention (non‐IT) is positively associated with user satisfaction.  Y 
H4c  The positive association between focus of attention (IT) and user satisfaction is  Y 
stronger than between focus of attention (non‐IT) and user satisfaction. 
H5  Technology anxiety is negatively associated with user satisfaction  Y 
 

2.9   Discussion 

This current study is intended to examine the relationship between mind wandering and 

emotional outcomes in a technological setting. Mind wandering scales were first developed and 

39 
 
the research model was empirically tested. The study provides an empirical support for the 

content regulation hypothesis by showing that (1) mind wandering and focus of attention are 

two different constructs that have distinct effects on emotional outcomes—mind wandering is 

the main predictor of technology anxiety, whereas focus of attention is the main predictor of 

user satisfaction; and (2) mind wandering and focus of attention can be differentiated into two 

dimensions: IT‐related and non‐IT related; the effect of mind wandering on emotional 

outcomes is dependent on these dimensions. By revealing that human cognition is a crucial 

factor in influencing emotional outcomes associated with technology usage, this research 

advances our understanding on the consequences of mind wandering in everyday life. 

2.9.1  Theoretical Contributions 

  In this study, a theoretical model based on the concept of mind wandering is formulated 

and empirically tested. My research framework provides an in‐depth understanding of the 

concept of mind wandering in a technological context. From an extensive literature review of 

the cognitive psychology, cognitive neuroscience, and IS disciplines, this study suggests that 

technology anxiety is primarily influenced by mind wandering, whereas user satisfaction is 

highly influenced by focus of attention. Specifically, mind wandering (non‐IT) has a greater 

effect on technology anxiety than mind wandering (IT) and focus of attention (IT) has a greater 

impact on user satisfaction than focus of attention (non‐IT). By demonstrating that the content 

of thought is associated with emotional outcomes, this study provides preliminary evidence 

that mind wandering covers a broad set of factors that potentially influence the outcomes of an 

interaction between users and technology.  

40 
 
This study resulted in a powerful model that explains 49 percent of the variance in 

technology anxiety. The small effect of mind wandering (IT) on technology anxiety can be 

explained by the notion that mind wandering (IT) reflects users’ positive attitude toward 

technology. Thus, its negative effect is not because users are worried about using the system, 

but because they shift their attention from the primary system used to complete the task to 

other irrelevant technologies (e.g., cellphone, social media, etc.). The insignificant effect of 

focus of attention (non‐IT) on technology anxiety could be due to the nature of task used in the 

study. The task was a technology task that required users to learn about a new system (e.g., 

MS. Excel, MS. PowerPoint) and apply what they have learned on the assigned task. In this case, 

focusing on technology features can be seen as a coping strategy that helps users reduce their 

degree of anxiety, whereas focusing on other subjects does not play a significant role in 

influencing technology anxiety.   

I also found that focus of attention is a major predictor of user satisfaction, whereas 

mind wandering did not have a direct effect on user satisfaction. These results suggest that 

mind wandering and focus of attention are two different constructs that have distinct effects 

on emotional outcomes. As one of the positive emotions, user satisfaction is the results of 

individuals’ positive experiences and situations that produce rewards (Bhattacherjee, 2001). 

When users focus on their task, their performance expectation increases as they become more 

confident in their ability to complete task. Thus, satisfaction can be seen as a reward for 

allocating their attention and effort on the task.  

Lastly, the results show that although people might argue that modern technology 

decreases the amount of time people mind wander because they always have electronic 

41 
 
devices in their hands to play with (Picciuto, 2015), the habit of using technologies can direct or 

stimulate our mind to wander about them. Given the unconscious mind is the storehouse of all 

memories and past experiences (Ritter & Dijksterhuis, 2014), the thought about this technology 

experience will remain in our memory and can emerge as our mind wanders. However, given 

that this study is a cross‐sectional study, the directional nature of the relationship between 

thought content and emotional outcomes cannot be confirmed. A complete understanding of 

the nature of this relationship requires overcoming several methodological and conceptual 

challenges that should be addressed in future research (Andrews‐Hanna et al., 2013). 

2.9.2  Practical Contributions 

  The results of this study offer practical contributions to technology users, education 

providers as well as business managers. First, the results suggest that mind wandering, 

especially about non‐technology related subjects, can increase technology anxiety, whereas 

focusing on relevant‐technology cues can slightly decrease such anxiety. Technology users are 

suggested to be aware of their mind wandering experience, at least during an early stage of 

technology usage when they are still in the process of learning a new technology. 

  This study also suggests that mind wandering can be an important issue for IS education 

in a number of ways. First, the occurrence of mind wandering during learning and retention of 

IS courses can make students become more anxious. Given that mind wandering involves a 

state of decoupled attention (Smallwood & Schooler, 2006), it prevents individuals from 

encoding information during the learning process. Interestingly, the thought contents of users 

determine the outcomes associated with mind wandering. Teachers and instructors who 

42 
 
prevent the use of electronic devices in classroom may be overlooking  the fact that individuals 

may engage in mind wandering (IT). In fact, efforts to focus on the primary technology or task 

and not to let distractions interfere with task performance may ironically increase the amount 

of mind wandering (Critcher & Gilovich, 2010). Based on the results of my current study, mind 

wandering can have a detrimental effect on technology learning by increasing the level of 

technology anxiety. Interestingly, the results suggest that mind wandering (non‐IT) has a 

stronger effect on technology anxiety than mind wandering (IT). Some intervention programs, 

such as integrating a mindfulness training in the curriculum will potentially help students be 

more aware of their own thought process. An alternative strategy is to focus on educating 

students about their attentional limitations and help them improve their engagement to the 

lecture (Risko, Buchanan, Medimorec, & Kingstone, 2013).   

  Third, although the data were collected from students, the results can be extended to 

understand mind wandering phenomena in work place. Given the results suggest that different 

dimensions of thought content are associated with different levels of technology anxiety as well 

as user satisfaction, knowledge workers who generally deal with technology tasks in their daily 

routine are encouraged to focus their attention on relevant cues while at the same time, be 

aware of their tendency to mind wander. By becoming aware of the nature of their internal 

thoughts, these knowledge workers may be able to reduce their degree of technology anxiety 

and increase their cognitive control ability. 

43 
 
2.10  Conclusions 

  This paper argues that mind wandering and focus of attention can be better understood 

by taking into account the content of thought. After developing and validating the scales used 

to measure mind wandering and focus of attention, I empirically examine the impact of mind 

wandering and focus of attention on emotional outcomes. Drawing upon and extending the 

content regulation hypothesis of mind wandering, the results reveal that mind wandering (non‐

IT) is the main source of technology anxiety, whereas focus of attention (IT) is the main 

antecedent of user satisfaction. This work sheds light on how mind wandering and focus of 

attention can influence emotional outcomes associated with technology use.  

2.11   References 

Addas, S., & Pinsonneault, A. (2015). The many faces of information technology interruptions: A 
taxonomy and preliminary investigation of their performance effects. Information Systems 
Journal, 25, 231‐273. 
Agarwal, R., & Karahanna, E. (2000). Time flies when you're having fun: Cognitive absorption 
and beliefs about information technology usage. MIS Quarterly, 24(4), 665‐694. 
Andrews‐Hanna, J. R., Smallwood, J., & Spreng, R. N. (2014). The default network and self‐
generated thought: Component processes, dynamic control, and clinical relevance. Annals 
of the New York Academy of Sciences, 1316(1), 29‐52. 
Andrews‐Hanna, J. R., Kaiser, R. H., Turner, A. E., Reineberg, A. E., Godinez, D., Dimidjian, S., & 
Banich, M. T. (2013). A penny for your thoughts: Dimensions of self‐generated thought 
content and relationships with individual differences in emotional wellbeing. Frontiers in 
Psychology, 4, 1‐13. 
Aral, S., Brynjolfsson, E., & Van Alstyne, M. (2012). Information, technology, and information 
worker productivity. Information Systems Research, 23(3), 849‐867. 
Bailey, J. E., & Pearson, S. W. (1983). Development of a tool for measuring and analyzing 
computer user satisfaction. Management Science, 29(5), 530‐545. 
Baird, B., Smallwood, J., & Schooler, J. W. (2010). I can shake that feeling: Positive mind 
wandering prevents the deterioration of mood. Poster Presented at Toward a Science of 
Consciousness Conference, Tucson, AZ. 

44 
 
Baird, B., Smallwood, J., & Schooler, J. W. (2011). Back to the future: Autobiographical planning 
and the functionality of mind‐wandering. Consciousness and Cognition, 20(4), 1604‐1611. 
Bagozzi, R. P. (2007). The legacy of the technology acceptance model and a proposal for a 
paradigm shift. Journal of the association for information systems, 8(4), 238‐245. 
Bagozzi, R. P., Gopinath, M., & Nyer, P. U. (1999). The role of emotions in marketing. Journal of 
the Academy Of Marketing Science, 27(2), 184‐206. 
Bagozzi, R. P., & Pieters, R. (1998). Goal‐directed emotions. Cognition & Emotion, 12(1), 1‐26. 
Barron, E., Riby, L. M., Greer, J., & Smallwood, J. (2011). Absorbed in thought: The effect of 
mind wandering on the processing of relevant and irrelevant events. Psychological Science, 
22(5), 596‐601. 
Beaudry, A., & Pinsonneault, A. (2010). The other side of acceptance: Studying the direct and 
indirect effects of emotions on information technology use. MIS Quarterly, 34(4), 689‐710. 
Bhattacherjee, A. (2001). Understanding information systems continuance: An expectation‐
confirmation model. MIS Quarterly, 25(3), 351‐370. 
Brockner, J. (1979). The effects of self‐esteem, success–failure, and self‐consciousness on task 
performance. Journal of Personality and Social Psychology, 37(10), 1732‐1741. 
Buckner, R. L., Andrews‐Hanna, J. R., & Schacter, D. L. (2008). The brain's default network. 
Annals of the New York Academy of Sciences, 1124(1), 1‐38. 
Butler, R. (1993). Effects of task‐and ego‐achievement goals on information seeking during task 
engagement. Journal of Personality and Social Psychology, 65(1), 18‐31. 
Carciofo, R., Du, F., Song, N., & Zhang, K. (2014). Mind wandering, sleep quality, affect and 
chronotype: An exploratory study. PloS one, 9(3), 1‐17. 
Chang, S. E. (2005). Computer anxiety and perception of task complexity in learning 
programming‐related skills. Computers in Human Behavior, 21(5), 713‐728. 
Chua, S. L., Chen, D. T., & Wong, A. F. (1999). Computer anxiety and its correlates: A meta‐
analysis. Computers in Human Behavior, 15(5), 609‐623. 
Christoff, K., Gordon, A. M., Smallwood, J., Smith, R., & Schooler, J. W. 2009. Experience 
sampling during fmri reveals default network and executive system contributions to mind 
wandering. Proceedings of the National Academy of Sciences, 106(21), 8719‐8724. 
Christoff, K., Gordon, A., & Smith, R. (2011). The role of spontaneous thought in human 
cognition. In O. Vartanian and D. R. Mardel (Eds.), Neuroscience of Decision Making (pp. 
259‐284), New York, NY: Psychology Press. 
Compeau, D. R., & Higgins, C. A. (1995). Computer self‐efficacy: Development of a measure and 
initial test. MIS Quarterly, 19(2), 189‐211. 
Compeau, D., Higgins, C. A., & Huff, S. (1999). Social cognitive theory and individual reactions to 
computing technology: A longitudinal study. MIS Quarterly, 23(2), 145‐158. 

45 
 
Critcher, C. R., & Gilovich, T. (2010). Inferring attitudes from mind wandering. Personality and 
Social Psychology Bulletin, 36(9), 1255‐1266. 
Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User acceptance of computer technology: A 
comparison of two theoretical models. Management Science, 35(8), 982‐1003. 
Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1992). Extrinsic and intrinsic motivation to use 
computers in the workplace. Journal of Applied Social Psychology, 22(14), 1111‐1132. 
Delaney, P. F., Sahakyan, L., Kelley, C. M., & Zimmerman, C. A. (2010). Remembering to forget 
the amnesic effect of daydreaming. Psychological Science, 21(7), 1036‐1042. 
DeLone, W. H., & McLean, E. R. (1992). Information systems success: The quest for the 
dependent variable. Information Systems Research, 3(1), 60‐95. 
DeLone, W. H., & McLean, E. R. (2003). The DeLone and McLean model of information systems 
success: A ten‐year update. Journal of Management Information Systems, 19(4), 9‐30. 
Dimitrova, N. G., van Dyck, C., van Hooft, E. A. J., & Groenewegen, P. (2015). Don’t fuss, focus: 
The mediating effect of on‐task thoughts on the relationship between error approach 
instructions and task performance. Applied Psychology: An International Review, 64(3), 
599‐624. 
Doll, W. J., Xia, W., & Torkzadeh, G. (1994). A confirmatory factor analysis of the end‐user 
computing satisfaction instrument. MIS Quarterly, 18(4), 453‐461. 
Fornell, C., & Larcker, D. F. (1981). Evaluating structural equations models with unobservable 
variables and measurement error. Journal of Marketing Research, 18(1), 39‐50. 
Forster, S., & Lavie, N. (2014). Distracted by your mind? Individual differences in distractibility 
predict mind wandering. Journal of Experimental Psychology: Learning, Memory, and 
Cognition, 40(1), 251‐260. 
Franklin, M. S., Mrazek, M. D., Anderson, C. L., Smallwood, J., Kingstone, A., & Schooler, J. W. 
(2013). The silver lining of a mind in the clouds: Interesting musings are associated with 
positive mood while mind wandering. Frontiers in Psychology, 4, 1‐5. 
Giambra, L. M. (1980). Sex differences in daydreaming and related mental activity from the late 
teens to the early nineties. International Journal of Aging and Human Development, 10(1), 
1‐34. 
Gorsuch, R. L. (1983). Factor analysis. Hillsdale, NJ: Erlbaum. 
Hollis, R. B. (2013). Mind wandering and online learning: A latent variable analysis. Unpublished 
doctoral dissertation, Kent State University College, Kent, OH.  
Howard,M .W.,& Kahana,M .J.( 2002). A distributed representation of temporal context. Journal 
of Mathematical Psychology, 46, 269‐299. 

46 
 
Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: 
Conventional criteria versus new alternatives. Structural Equation Modeling: A 
Multidisciplinary Journal, 6(1), 1‐55. 
Igbaria, M., & Nachman, S. A. (1990). Correlates of user satisfaction with end user computing: 
An exploratory study. Information & Management, 19(2), 73‐82. 
Jiang, Y., & Punj, G. N. (2010). The effects of attribute concreteness and prominence on 
selective processing, choice, and search experience. Journal of the Academy of Marketing 
Science, 38(4), 471‐489. 
Kanfer, R., & Ackerman, P. L. (1989). Motivation and cognitive abilities: An integrative/aptitude‐
treatment interaction approach to skill acquisition. Journal of Applied Psychology, 74(4), 
657‐690. 
Kanfer, R., Ackerman, P. L., Murtha, T. C., Dugdale, B., & Nelson, L. (1994). Goal setting, 
conditions of practice, and task performance: A resource allocation perspective. Journal of 
Applied Psychology, 79(6), 826‐835. 
Killingsworth, M. A., & Gilbert, D. T. (2010). A wandering mind is an unhappy mind. Science,  
330, 932. 
Klinger, E. (1984). A consciousness‐sampling analysis of test anxiety and performance. Journal 
of personality and Social Psychology, 47(6), 1376‐1390. 
Koufaris, M. (2002). Applying the technology acceptance model and flow theory to online 
consumer behavior. Information Systems Research, 13(2), 205‐223. 
Lee, F. K., Sheldon, K. M., & Turban, D. B. (2003). Personality and the goal‐striving process: The 
influence of achievement goal patterns, goal level, and mental focus on performance and 
enjoyment. Journal of Applied Psychology, 88(2), 256‐265. 
Lindell, M. K., and Whitney, D. J. (2001). Accounting for common method variance in cross‐
sectional research designs. Journal of Applied Psychology, 86(1), 114‐121. 
MacKenzie, S. B., Podsakoff, P. M., & Podsakoff, N. P. (2011). Construct measurement and 
validation procedures in MIS and behavioral research: Integrating new and existing 
techniques. MIS Quarterly, 35(2), 293‐334. 
Mahmood, M. A., Burn, J. M., Gemoets, L. A., & Jacquez, C. (2000). Variables affecting 
information technology end‐user satisfaction: A meta‐analysis of the empirical literature. 
International Journal of Human‐Computer Studies, 52(4), 751‐771. 
Mar, R. A., Mason, M. F., & Litvack, A. (2012). How daydreaming relates to life satisfaction, 
loneliness, and social support: The importance of gender and daydream content. 
Consciousness and Cognition, 21(1), 401‐407. 
McVay, J. C., & Kane, M. J. (2010). Does mind wandering reflect executive function or executive 
failure? Comment on Smallwood and Schooler (2006) and Watkins (2008). Psychological 
Bulletin, 136, 188‐197. 

47 
 
Moberly, N. J., & Watkins, E. R. (2008). Ruminative self‐focus and negative affect: An experience 
sampling study. Journal of Abnormal Psychology, 117(2), 314‐323. 
Mooneyham, B. W., & Schooler, J. W. (2013). The costs and benefits of mind wandering: A 
review. Canadian Journal of Experimental Psychology, 67(1), 11‐8. 
Mrazek, M. D., Phillips, D. T., Franklin, M. S., Broadway, J. M., & Schooler, J. W. (2013). Young 
and restless: Validation of the mind wandering questionnaire (MWQ) reveals disruptive 
impact of mind wandering for youth. Frontiers in Psychology, 4, 1‐7. 
Nadkarni, S., & Gupta, R. (2007). A task‐based model of perceived website complexity. MIS 
Quarterly, 31(3), 501‐524. 
Oettingen, G., & Schwörer, B. (2013). Mind wandering via mental contrasting as a tool for 
behavior change. Frontiers in Psychology, 4, 1‐5. 
Orlikowski, W. J., & Gash, D. C. (1994). Technological frames: Making sense of information 
technology in organizations. ACM Transactions on Information Systems, 12(2), 174‐207. 
Ortiz de Guinea, A. O., & Markus, M. L. (2009). Why break the habit of a lifetime? Rethinking 
the roles of intention, habit, and emotion in continuing information technology use. MIS 
Quarterly, 33(3), 433‐444. 
Ortiz de Guinea, O. A., & Webster, J. (2013). An investigation of information systems use 
patterns: Technological events as triggers, the effect of time, and consequences for 
performance. MIS Quarterly, 37(4), 1165‐1188. 
Petronzio, M. (2014). U.S. Adults spend 11 hours per day with digital media. Retrieved from 
Mashable website: http://mashable.com/2014/03/05/american‐digital‐media‐
hours/#TAE3TI5GwSqm 
Picciuto, E. (2015). Don’t stop daydreaming: Why a wandering mind is good for you. Retrieved 
from thedailybeast website: http://www.thedailybeast.com/articles/2015/03/02/don‐t‐
stop‐daydreaming‐why‐a‐wandering‐mind‐is‐good‐for‐you.html 
Podsakoff, P. M., MacKenzie, S. B., Lee, J.‐Y., & Podsakoff, N. P. (2003). Common method biases 
in behavioral research: a critical review of the literature and recommended remedies. 
Journal of Applied Psychology, 88(5), 879‐903. 
Poerio, G. L., Totterdell, P., & Miles, E. (2013). Mind wandering and negative mood: Does one 
thing really lead to another? Consciousness and Cognition, 22(4), 1412‐1421. 
Polites, G. L., & Karahanna, E. (2012). Shackled to the status quo: The inhibiting effects of 
incumbent system habit, switching costs, and inertia on new system acceptance. MIS 
Quarterly, 36(1), 21‐42. 
Raichle, M. E., MacLeod, A. M., Snyder, A. Z., Powers, W. J., Gusnard, D. A., & Shulman, G. L. 
(2001). A default mode of brain function. Proceedings of the National Academy of Sciences, 
98(2), 676‐682. 

48 
 
Ralph, B. C., Thomson, D. R., Cheyne, J. A., & Smilek, D. (2014). Media multitasking and failures 
of attention in everyday life. Psychological Research, 78(5), 661‐669. 
Randall, J. G., Oswald, F. L., & Beier, M. E. (2014). Mind wandering, cognition, and performance: 
A theory‐driven meta‐analysis of attention regulation. Psychological Bulletin, 140(6), 1411‐
1431. 
Risko, E. F., Buchanan, D., Medimorec, S., & Kingstone, A. (2013). Everyday attention: Mind 
wandering and computer use during lectures. Computers and Education, 68, 275‐283. 
Ritter, S. M., & Dijksterhuis, A. (2014). Creativity: The unconscious foundation of the incubation 
period. Frontiers in Human Neuroscience, 8, 1‐10. 
Robinson, M. D., & Tamir, M. (2011). A task‐focused mind is a happy and productive mind: A 
processing perspective. In K. M. Sheldon, T. B. Kashdan, & Steger, M. F. (Eds.), Designing 
positive psychology: Taking stock and moving forward (pp. 160‐174), New York, NY: Oxford 
University Press. 
Rogers, E. M. (1995). Diffusion of innovation, New York, NY: Free Press. 
Ruby, F. J., Smallwood, J., Engen, H., & Singer, T. (2013). How self‐generated thought shapes 
mood‐the relation between mind wandering and mood depends on the socio‐temporal 
content of thoughts. PLoS One, 8(10), 1‐7. 
Sarason, I. G., Pierce, G. R., & Sarason, B. R. (2002). Domains of cognitive interference. In I. G. 
Sarason, G. R. Pierce, & B. R. Sarason (Eds.), Cognitive interference: Theories, methods, and 
findings (pp. 139‐152), Mahwah, New Jersey: Lawrence Erlbaum. 
Sarason, I. G., Sarason, B. R., Keefe, D. E., Hayes, B. E., & Shearin, E. N. (1986). Cognitive 
interference: Situational determinants and traitlike characteristics. Journal of Personality 
and Social Psychology, 51(1), 215‐226. 
Schooler, J. W., Reichle, E. D., & Halpern D. V. (2004). Zoning Out while Reading: Evidence for 
Dissociations between Experience and Metaconsciousness. In D. T. Levin (Ed.), Thinking 
and seeing: Visual metacognition in adults and children (pp. 203‐226),  Cambridge, MA: 
MIT Press. 
Seijts, G. H., Latham, G. P., Tasa, K., & Latham, B. W. (2004). Goal setting and goal orientation: 
An integration of two different yet related literatures. Academy of Management Journal, 
47(2), 227‐239. 
Smallwood, J. M., & Andrews‐Hanna, J. A. (2013). Not all minds that wander are lost: The 
importance of a balanced perspective on the mind wandering state. Frontiers in 
Psychology, 4, 1‐6. 
Smallwood, J. M., Baracaia, S. F., Lowe, M., & Obonsawin, M. (2003). Task unrelated thought 
whilst encoding information. Consciousness and Cognition, 12(3), 452‐484. 

49 
 
Smallwood J. M., Davies J. B., Heim D., Finnigan F., Sudberry M., & O’Connor R. (2004). 
Subjective experience and the attentional lapse: Task engagement and disengagement 
during sustained attention. Consciousness and Cognition, 13, 657‐690. 
Smallwood, J. M, Fishman, D. J., & Schooler, J. W. 2007. Counting the cost of an absent mind: 
Mind wandering as an underrecognized influence on educational performance. 
Psychonomic Bulletin & Review, 14(2), 230‐236. 
Smallwood, J. M., Fitzgerald, A., Miles, L. K., & Phillips, L. H. (2009). Shifting moods, wandering 
minds: negative moods lead the mind to wander. Emotion, 9(2), 271. 
Smallwood, J. M., & O'Connor, R. C. (2011). Imprisoned by the past: Unhappy moods lead to a 
retrospective bias to mind wandering. Cognition & Emotion, 25(8), 1481‐1490. 
Smallwood, J. M., O'Connor, R. C., Sudbery, M. V., & Obonsawin, M. (2007). Mind wandering 
and dysphoria. Cognition and Emotion, 21(4), 816‐842. 
Smallwood, J. M., & Schooler, J. W. (2006). The restless mind. Psychological Bulletin, 132(6), 
946‐958. 
Smallwood, J. M., & Schooler, J. W. (2015). The science of mind wandering: Empirically 
navigating the stream of consciousness. Annual Review of Psychology, 66, 487‐518. 
Speier, C., Vessey, I., & Valacich, J. S. (2003). The effects of interruptions, task complexity, and 
information presentation on computer supported decision making performance. Decision 
Sciences, 34(4), 771‐797. 
Stanton, J. M., Sinar, E. F., Balzer, W. K., & Smith, P. C. (2002). Issues and strategies for reducing 
the length of self‐report scales. Personnel Psychology, 55(1), pp. 167‐194. 
Sullivan, Y. W., Davis, F. D., & Koh, C. (2015). Exploring mind wandering in a technological 
setting. Paper presented at the 36th International Conference on Information Systems 
(ICIS), Fort Worth, TX, December 13‐16. 
Tarafdar, M., Tu, Q., & Ragu‐Nathan, T. S. (2010). Impact of technostress on end‐user 
satisfaction and performance. Journal of Management Information Systems, 27(3), 303‐
334. 
Tarafdar, M., Tu, Q., Ragu‐Nathan, T. S., & Ragu‐Nathan, B. S. (2011). Crossing to the dark side: 
examining creators, outcomes, and inhibitors of technostress. Communications of the 
ACM, 54(9), 113‐120. 
Taylor, S., & Todd, P. A. (1995). Understanding information technology usage: A test of 
competing models. Information Systems Research, 6(2), 144‐176. 
Thatcher, J. B., & Perrewe, P. L. (2002). An empirical examination of individual traits as 
antecedents to computer anxiety and computer self‐efficacy. MIS Quarterly, 26(4), 381‐
396. 
Thompson, B. (2004). Exploratory and confirmatory factor analysis. Washington DC: American 
Psychological Association. 

50 
 
Trevino, L. K., & Webster, J. (1992). Flow in computer‐mediated communication electronic mail 
and voice mail evaluation and impacts. Communication Research, 19(5), 539‐573. 
Tu, W., Tarafdar, M., Ragu‐Nathan, B., & Ragu‐Nathan, T. (2007). How end‐user characteristics 
affect technostress: An exploratory investigation. In Proceedings of 38th Annual Meeting 
Decision Science Institute, November 17‐20, Phoenix, Arizona.  
Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of information 
technology: Toward a unified view. MIS Quarterly, 27(3), 425‐478. 
Venkatesh, V. (2000). Determinants of perceived ease of use: Integrating control, intrinsic 
motivation, and emotion into the technology acceptance model. Information Systems 
Research, 11(4), 342‐365. 
Watkins, E. R. (2008). Constructive and unconstructive repetitive thought. Psychological 
Bulletin, 134(2), 163‐206. 
Winters, D., & Latham, G. P. (1996). The effect of learning versus outcome goals on a simple 
versus a complex task. Group & Organization, 21(2), 236‐250. 
Wixom, B. H., & Todd, P. A. (2005). A theoretical integration of user satisfaction and technology 
acceptance. Information Systems Research, 16(1), 85‐102. 
Webster, J., Heian, J. B., & Michelman, J. E. (1990). Computer training and computer anxiety in 
the educational process: An experimental analysis. In Proceedings of International 
Conference in Information Systems, New York, December 16‐18. 
   

51 
 
CHAPTER 3 

MIND WANDERING AND COGNITIVE IS LEARNING 
 
3.1  Introduction 

In the last decade, mind wandering (or self‐generated thoughts or task‐unrelated 

thoughts) has become an important topic of research in the psychology and neuroscience 

disciplines (Smallwood & Andrews‐Hanna, 2013). Mind wandering represents a shift of 

attention away from a primary task to the processing of internal goals (Singer, 1966; Smallwood 

& Schooler, 2006). Originally, mind wandering is considered disruptive to accomplishing 

primary tasks and related to low task performance (e.g., McVay & Kane, 2009). Mind wandering 

is also negatively associated with an individual’s ability to retrieve information (Smallwood, 

Beach, Schooler, & Handy, 2008) and its occurrence has been linked to low performance of fluid 

intelligence test (Mrazek et al., 2012; Smallwood & Andrews‐Hanna, 2013). For that reason, 

mind wandering is considered to have significant implications within the education domain 

(Smallwood, Fishman, & Schooler, 2007a).  

In this study, I specifically focus on the effect of mind wandering on cognitive outcomes 

in the Information Systems (IS) learning context. IS learning involves making connections 

between the mechanisms of information and technology and particular applications of such 

mechanisms in work practices (Rawlings, White, & Stephens, 2005). Unlike other types of 

learning, IS learning requires students to pay attention not only to the task, but also to the 

technology associated with the task. Although learning is a critical part of technology adoption, 

research on the relationship between mind wandering and technology learning has been 

52 
 
overlooked in the IS literature. Thus, it is necessary to examine the impact of mind wandering 

on various IS learning outcomes because IS learning success depends, in part, on users’ ability 

to regulate their attentional mechanisms involved in the learning process (Smallwood et al., 

2007a).  

The interaction between user and technology has been shown to be a major predictor 

of task performance (e.g., Burton‐Jones & Straub, 2006; DeLone & McLean, 1992, 2003; Van der 

Heijden, 2004) (Sullivan, Davis, & Koh, 2015). Technology also provides a potential set of 

features that influence how users choose to perform a task (Fuller & Dennis, 2009). However, it 

is unclear how technology users maintain their attentional process mechanisms when they 

interact with technology, particularly during the early stage of technology learning (Sullivan, 

Davis, & Koh, 2015). At the same time, most research on mind wandering has primarily focused 

on many situations in which mind wandering has detrimental consequences on task 

performance (e.g., McVay & Kane, 2010; Schooler, Reichle, & Halpern, 2004). Although its 

benefits are less extensively studied, mind wandering may have some positive functions 

including contributing to future planning and learning (Ruby, Smallwood, Engen, & Singer, 2013; 

Smallwood & Schooler, 2015). It is also associated with certain cognitive skills, such as future 

planning, knowledge retention (Smallwood & Andrews‐Hanna, 2013), and consolidation of self‐

memory (Smallwood et al., 2011). Given many of the potential benefits of mind wandering is 

still unknown, the main objective of this study is to examine the cognitive costs and benefits of 

mind wandering on learning outcomes (e.g., software self‐efficacy). 

  This study addresses three research questions: (1) What are the factors that influence 

users’ mind wandering activities and focus of attention (at the early stage of learning phase)?; 

53 
 
(2) How does the interaction between mind wandering and focus of attention influence 

students’ cognitive learning outcome (i.e., software self‐efficacy)?;  and (3) Do mind wandering, 

focus of attention, and software self‐efficacy influence IS learning performance? Drawing on 

executive control failure theory of mind wandering, I postulate that a user’s mind wandering 

and focus of attention activities are influenced by executive functions (EFs)—the cognitive 

control processes that regulate thoughts and actions (Friedman et al., 2008)—and task 

complexity. Building on this theory, I trace the relationship between mind wandering, focus of 

attention, self‐efficacy and performance back to its underlying assumptions, which are rooted 

in resource allocation theory.  

  This work makes several contributions. It advances our understanding on the costs and 

benefits of mind wandering in the IS learning context. The interaction between the content of 

thought when users are engaged in mind wandering and focus on their task is the key indication 

to uncover the potential benefits of mind wandering. This work contributes to the literature by 

extending and integrating the theory of EFs and mind wandering in a technological setting. It 

links the objective measures of EFs to the subjective measures of mind wandering and focus of 

attention and empirically tests the influence of EFs, mind wandering, and focus of attention on 

cognitive learning outcomes.  

 
3.2   Executive Functions Theory 

Executive functions (EFs)—also known as executive control or cognitive control—refer 

to “a family of top‐down mental processes needed when you have to concentrate and pay 

attention, when going on automatic or relying on instinct or intuition would be ill‐advised, 

insufficient, or impossible” (Diamond, 2013, p. 136). EFs are crucial in the cognitive processes of 
54 
 
attentional control (i.e., top‐down control) (Diamond, 2013; Miyake et al., 2000). The theory of 

EFs is concerned with people’s ability to control their attentional or cognitive resources in the 

service of pursuing goals and accomplishing tasks, especially when faced with interference or 

distraction (Kane et al., 2007; Randall, Oswald, & Beier, 2014). There are three most frequently 

studied EFs: (1) inhibition (i.e., the ability to inhibit dominant or automatic responses), (2) 

updating working memory representations (i.e., the ability to monitor incoming information 

relevant to the task in hand and then appropriately update the working memory by replacing 

old, no longer relevant information with newer, more‐relevant information), and (3) set shifting 

(i.e., the ability to flexibly switch back and forth between tasks or mental sets) (Miyake et al., 

2000). According to Miyake and Friedman (2012), these three dimensions of EFs correlate with 

one another; thus, they share some common underlying ability, but they also show some 

separability.  

3.2.1   Inhibition 

Cognitive inhibition is defined as “a basic cognitive suppression that contributes to task 

performance by keeping task‐irrelevant information from entering and being maintained in 

working memory” (Harnishfeger, 1995, p. 178). The concept of inhibition is also used 

interchangeably with interference control—“suppressing a stimulus that pulls for a competing 

response so as to carry out a primary response, to suppressing distractors that might slow the 

primary response, or to suppressing internal stimuli that may interfere with the current 

operations of working memory” (Nigg, 2000, p. 222). Theories have proposed several forms of 

inhibition (Friedman & Miyake, 2004). Based on the nature of cognitive interference, inhibition 

55 
 
can be categorized into three different types—i.e., prepotent response inhibition, resistance to 

distractor interference, and resistance to proactive interference. Prepotent response inhibition 

is the ability to deliberately suppress dominant or automatic responses. Second, resistance to 

distractor interference is the ability to resist or resolve interference from information triggered 

by the external environment that is not relevant to the task at hand. It has been associated with 

individuals’ ability to ignore interruptions from the environment and focus on the task. Lastly, 

resistance to proactive interference is the ability to suppress information from memory that 

was previously relevant to the task but has since become irrelevant (Friedman & Miyake, 2004). 

Among these different types of inhibition, prepotent response inhibition is commonly 

measured in EFs studies because it represents an “internally generated act of control” (Logan, 

1994, p. 190) and linked to the frontal lobes—a part of the brain that plays a major role in 

regulating cognitive functions (Miyake et al., 2000). 

Inhibition mechanisms can also be distinguished based on psychological constructs (e.g., 

behavior, cognition) they act on (Harnishfeger, 1995). Given mind wandering is operationalized 

as a cognitive construct, the focus of this study is on cognitive inhibition. Cognitive inhibition 

can be defined as the suppression of cognitive interference from working memory (Friedman & 

Miyake, 2004). Cognitive inhibition enables individuals to control their thought contents or 

processes; it can be intentional and conscious, or unintentional or unconscious (Harnishfeger, 

1995). Hence, conceptualizing inhibition as a cognitive construct is relevant to the purpose of 

this study.  

56 
 
3.2.2   Switching 

  Switching ability refers to “any mechanism that reduces or dampens neuronal, mental, 

or behavioral activity” (Clark, 1996, p. 128). Task switching in everyday life is inevitable. The 

switching process involves the omission of irrelevant task sets (e.g., to be encountered stimuli, 

the required responses, and the cues used to indicate which information is the relevant to the 

current task) and the subsequent active engagement of a relevant task set (Eysench, 

Derakshan, Santon, & Calvo, 2007; Miyake et al., 2000). It reflects the ability of an individual to 

be flexible enough to go back and forth between one task to another (Dijksterhuis & Aarts, 

2010; Miyake et al., 2000). Switching involves active and intentional retrieval of task sets or 

instantiation of current task goals (Miyake et al., 2000). This retrieval of task sets is crucial for 

reducing the effects of interference from the irrelevant information of the task because it 

involves the activation of criteria for what information is currently relevant and what is 

currently irrelevant and should be ignored (Friedman & Miyake, 2004). Generally, switching 

ability is measured using a task‐switching paradigm, where subjects are presented with multiple 

stimuli. Subjects are first asked to respond to a set of stimuli and then after a certain amount of 

time they were asked to switch to attend another type of stimuli (Monsell, 2003).  

3.2.3   Updating or Working Memory Capacity 

  Updating or monitoring of working memory representations is closely linked to the 

notion of working memory capacity (WMC) (Miyake et al., 2000). Although researchers have a 

tendency to discuss attention and working memory in isolation of one another, it is suggested 

that they are mutually dependent functions because they share many of the same neural 

57 
 
substrates (Baddeley, 2000, 2007, 2012). For example, working memory maintains 

representations of current task demands that are necessary for the success of selecting task‐

relevant representations and actions (Milham et al., 2002). The function of working memory 

requires monitoring and coding incoming information by replacing old information with most 

recent and more relevant information (Baddeley, 2012).  

  According to the EFs perspective, WMC reflects, in part, an ability to control thought 

(McVay & Kane, 2010). WMC helps maintain an active goal and avoid distraction (Baddeley, 

2003; Kane & McVay, 2012; Unsworth, McMillan, Brewer, & Spillers, 2012). The concept of 

working memory suggests that a limited capacity system, which temporarily maintains and 

stores information, supports human thought processes by providing an interface between 

perception, long‐term memory, and action (Baddeley, 2003). In other words, normal variation 

in WMC is influenced by attentional processes that regulate thoughts and actions (Engle & 

Kane, 2004).  

3.3  Mind Wandering 

  Mind wandering has been conceptualized as “a shift of attention away from a primary 

task toward internal information, such as memories” (Smallwood & Schooler, 2006, p. 946). 

Mind wandering is off‐task thoughts that occur during an ongoing task or activity (McVay & 

Kane, 2010). Periods of mind wandering are generally associated with less accurate awareness 

of relevant information than periods of task focus because mind wandering involves a shift of 

attention away from the primary task to individuals’ internal thought processes (Smallwood & 

Schooler, 2006). Often the thought that occurs during mind wandering experience is described 

58 
 
as task unrelated or stimulus independent—terms that capture the independence of experience 

from external stimuli (Smallwood & Schooler, 2015). Research over the past 15 years has 

proposed different theories in understanding the occurrence of mind wandering (Smallwood & 

Schooler, 2015). For instance, Smallwood and Schooler (2006) proposed that mind wandering is 

operated under the same mechanisms of EFs. It occurs when information processing is 

decoupled from the primary task. In contrast, McVay and Kane (2009) suggested that mind 

wandering does not require resources from EFs. Mind wandering is able to enter consciousness 

because EFs fail to inhibit the irrelevant thoughts associated with mind wandering. This latter 

theory is known as executive‐control failure theory of mind wandering. Given we conceptualize 

EFs as one’s ability to deal with interference, this current study draws upon the executive‐

control failure theory of mind wandering to explain the relationship between EFs and mind 

wandering.  

3.3.1   Executive‐Control Failure Theory of Mind Wandering 

  According to the executive‐control failure perspective, mind wandering does not require 

executive control resources nor does it relies on the same cognitive mechanisms used for EFs 

(McVay & Kane, 2009, 2012a). Rather, mind wandering is “automatically and continually 

generated as part of the thought stream in response to internal and external cues, and 

executive‐control processes keep these thoughts out of the focus of attention during resource‐

demanding tasks.” (McVay and Kane 2009, p. 203). It is dually determined by the presence of 

automatically generated thoughts and the ability of EFs to deal with such automatic thoughts 

(McVay and Kane, 2010). Mind wandering is the result of executive‐control failure in that EFs 

59 
 
fail to adequately inhibit interfering thoughts that are generated and maintained automatically 

(i.e., unintentionally and without consuming executive resources) (McVay & Kane, 2012a).  

  Two functions of executive control: (1) goal maintenance processes (i.e., processes that 

allow for the sustained access to task‐relevant information in the face of distraction from 

environmental distractors or irrelevant thought) and (2) competition‐resolution mechanisms 

(i.e., functions that deal with in‐the‐moment interference from a stimulus) prevent mind 

wandering to enter consciousness (McVay & Kane, 2012a). Whereas the goal maintenance 

processes are initiated prior to the expected need for control in order to minimize experiences 

of distraction, the competition‐resolution mechanisms are initiated in‐the‐moment, as a 

response to any distraction (McVay & Kane, 2012a). These two functions have been shown to 

determine successful performance on many attention‐demanding tasks (Kane & Engle, 2003; 

McVay & Kane, 2012a). Mind wandering, on the other hand, disrupts goal maintenance 

processes, leading to performance errors in attention‐demanding tasks (McVay & Kane, 2009, 

2012a).  

In a set of experiments, McVay and Kane (2009) found that individuals with lower EFs 

mind wandered more during a demanding (primary) task than higher EFs subjects. These results 

suggest that individuals with lower EFs show frequent goal neglect because they fail to maintain 

the goals across different levels of complexity. Lower EFs subjects also seem less able than 

those with higher EFs to sustain focus of attention to the demands of the ongoing task (McVay 

& Kane, 2009). When EFs fail to maintain focus of attention, mind wandering will negatively 

influence task performance (Kane et al., 2007; McVay & Kane, 2009). In another study, Kane 

and McVay (2012a) examined the association between EFs (measured as WMC) and reading‐

60 
 
comprehension task performance, with mind wandering as a partial mediator. The results 

confirmed their hypotheses that: (1) EFs were correlated negatively with mind wandering and 

positively with task performance; (2) EFs had a significant direct effect on task performance; 

and (3) mind wandering mediated the relationship between EFs and reading comprehension. 

These findings suggest that the relationship between EFs and task performance arises partly 

because EFs reflect individuals’ executive ability to control thoughts (Engle & Kane, 2004; Kane 

& McVay, 2012). Hence, mind wandering is prevented when control is proactively initiated to 

inhibit mind wandering as they are activated in response to environmental or mental cues 

(McVay & Kane, 2010).    

In sum, executive‐control failure theory of mind wandering predicts that task demand 

(i.e., task complexity) is negatively associated with mind wandering (i.e., for high complexity 

tasks, mind wandering is diminished) in that individuals with high EFs will keep the unrelated 

thought out of the focus of attention. Consequently, individuals with low EFs, relative to those 

with high EFs, will experience more mind wandering, resulting in poor task performance. Given 

individuals with high EFs are able to maintain the task‐relevant information, they are likely to 

achieve higher task performance than individuals with low EFs. 

Figure 3.1. Graphical Illustration of Executive‐Control Failure Theory 

61 
 
3.3.2  Consequences of Mind Wandering 

  More than a decade of research has shown that mind wandering is associated with poor 

outcomes in a wide variety of tasks, such as reading (e.g.,  Franklin, Smallwood, & Schooler, 

2011; Schooler et al., 2004) and sustained attention tasks (e.g., Smallwood et al., 2004; McVay 

& Kane, 2012b). In addition to exploring the costs of mind wandering, researchers have started 

investigating the potential benefits of mind wandering (Mooneyham & Schooler, 2013; 

Smallwood & Andrews‐Hanna, 2013). Table 3.1 outlines a comprehensive list of studies that 

have demonstrated the consequences of mind wandering on cognitive outcomes in a number 

of settings, including reading comprehension, lecture, and sustained attention tasks. 

Table 3.1 

Literature Review of Costs and Benefits of Mind Wandering (In a Chronological Order) 

Study  Context  Cost  Benefit 


Smallwood, Baracaia,  Memory  Mind wandering was negatively  ‐ 
Lowe, and Obonsawin  associated with memory 
(2003)  retrieval.  
Smallwood et al. (2004)  Sustained  Mind wandering predicted  ‐ 
attention  performance deficits on 
attention‐demanding tasks. 
Smallwood, McSpadden,  Response  Mind wandering was associated  ‐ 
and Schooler (2007b)  inhibition  with a relative absence of the 
influence of recollection at 
retrieval.  
Smallwood, O’Connor,  Emotion and  Mind wandering was associated  ‐ 
Sudbery, and Obensawin  memory  with poor retrieval of 
(2007c)  information and it was used as a 
marker for depressive thinking. 
McVay, Kane, and Kwapil  Ecological  Mind wandering was negatively  ‐ 
(2009)  context  associated with performance in 
both laboratory and daily life 
settings. 
(table continues) 

62 
 
Table 3.1 (continued) 
 
Study  Context  Cost  Benefit 
McVay and Kane (2009)  Sustained  Mind wandering predicted goal‐ ‐ 
attention   neglect errors in the sustained 
attention task.  
Delaney, Sahakyan,  Memory  Mind wandering that was more   
Kelley, and Zimmerman  different from the current 
(2010)  moment (e.g., in distance, time, 
and circumstance) resulted in 
more forgetting than mind 
wandering that was less 
different from the current 
moment.  
Barron, Riby, Greer, and  Visual oddball  Mind wandering was associated  ‐ 
Smallwood (2011)  task  with low task performance. 
Franklin et al. (2011)  Reading  Mind wandering was negatively  ‐ 
comprehension  associated with reading 
comprehension.  
Lindquist and McLean  Academic  The frequency of mind  ‐ 
(2011)  learning  wandering was negatively 
associated with performance in 
course examinations. 
Smallwood (2011)  Reading  Mind wandering was associated   
comprehension  with low reading 
comprehension 
Baird et al. (2012)  Creativity  ‐  Mind wandering was 
problems  associated with 
higher performance 
on creative problem 
solving tasks. 
McVay and Kane (2012a)  Reading  Mind wandering was negatively  ‐ 
comprehension  associated with reading 
comprehension.  
McVay and Kane (2012b)  Sustained  Mind wandering had a negative  ‐ 
attention  effect on task performance. 
Risko et al. (2012)  Lecture  Mind wandering led to  ‐ 
decreased memory for lecture 
materials.  
(table continues) 

63 
 
Table 3.1 (continued) 

Study  Context  Cost  Benefit 


Mrazek et al. (2012)  Sustained  Mind wandering was negatively  ‐ 
attention  associated with task performance.  
Allen et al. (2013)  Error monitoring  Mind wandering was detrimental  Fluctuation between 
to task performance.  mind wandering and 
focus of attention was a 
characteristic of 
individuals with greater 
metacognitive capacity. 
Foulsham, Farley, and  Reading  Mind wandering led to slower   
Kingstone (2013)  comprehension  reading times and longer average 
fixation duration. 
McVay, Meier, Touron,  Sustained  Mind wandering was negatively  ‐ 
and Kane (2013)  attention  associated with task performance. 
Mrazek, Phillips,  Reading  Mind wandering was associated  ‐ 
Franklin Broadway, and  comprehension  with worse reading 
Schooler (2013)  comprehension.  
Oettigen and Schwörer  Problem solving  Solely mind wander about a  Mind wandering in the 
(2013)  desired future led to poor problem  form of mental 
solving and little behavior change.   contrasting led to skilled 
problem solving and 
substantial behavioral 
change.  
Ruby et al. (2013)  Problem solving   ‐  Although mind 
wandering was 
negatively associated 
with performance 
efficiency, it was 
associated with higher 
relevant means (higher 
performance accuracy). 
Smallwood, Ruby, and  Inter‐temporal  ‐  Mind wandering was 
Singer (2013)  discounting task  associated with reduced 
delay discounting.  
Unsworth and  Reading  Mind wandering was associated  ‐ 
McMillan (2013)  comprehension  with poor reading comprehension.  
Hollis and Was (2014)  Online course  Mind wandering was negatively  ‐ 
associated with learning 
performance. 
(table continues) 
 

64 
 
 

Table 3.1 (continued) 

Study  Context  Cost  Benefit 


Ritter and  Problem solving    Mind wandering 
Dijksterhuis (2014)  promoted creativity. 
Stawarczyk, Majerus,  Sustained  Mind wandering was associated  ‐ 
Catale, and  attention  with decreases in sustained 
D’Argembeau (2014)  attention task performance.  
Axelrod, Rees,  Sustained  ‐  Mind wandering did not 
Lavidor, and Bar  attention  impair external task 
(2015)  performance (but 
slightly improved it).  
   

As shown in Table 3.1, the negative relationship between mind wandering and 

performance is typically found in sustained attention tasks (e.g., Smallwood et al., 2004, 2009; 

McVay et al., 2013; Mrazek et al., 2012; Stawarczyk et al., 2014), whereas the positive 

relationship between mind wandering and performance is typically found in problem solving 

tasks (e.g., Baird et al., 2012; Ruby et al., 2013; Oettingen & Schwörer, 2013). This contradictive 

relationship between mind wandering and ongoing performance suggests that the costs and/or 

benefits of the experience can be better understood by taking into account the context in 

which it occurs (Andrews‐Hanna, Smallwood, & Spreng, 2014; Smallwood & Schooler, 2015) as 

well as the content of mind wandering experience itself (Andrews‐Hanna et al., 2014).   

3.4   Resource Allocation Theory 

  Resource allocation theory was proposed by Kanfer and Ackerman (1989). According to 

this theory, individuals’ attentional resources are generally indexed with measures of cognitive 

ability and a task is assumed to be resource‐limited when focus of attention is needed to 

65 
 
achieve high task performance (Randall et al., 2014). Resource allocation theory extends the 

cognitive resource perspective by explaining how attentional resources are allocated to task 

performance in terms of on‐task activities, off‐task activities or mind wandering, and self‐

regulatory activities4 (Randall et al., 2014). Whereas on‐task thought refers to the thought 

maintained on the primary task at hand (Randall et al., 2014), off‐task or mind wandering 

activities involve allocating attention to task‐unrelated thoughts (Smallwood et al., 2007a). Self‐

regulatory activities are essential mechanisms to maintain the resource allocation policy toward 

a task (Kanfer & Ackerman, 1989). Resource allocation theory views task performance as the 

outcome of person’s ability (i.e., the attentional resources devoted to the task), the person’s 

motivation to direct and regulate attention, and task characteristics, such as task complexity 

(Kanfer & Ackerman, 1989). 

  Resource allocation theory has been used to understand the occurrence of mind 

wandering during the process of skill acquisition and task execution (Randall et al., 2014). 

According to this theory, tasks that are relative novel and complex will require focused 

attentional resources at early stages of learning (Ackerman, 1988; Kanfer & Ackerman, 1989). 

Once the skills needed to execute the tasks become autonomous or automatic, fewer attention 

resources are necessary (Ackerman, 1988). For example, the first time users learn how to use 

keyboards, their attention resources are allocated to learn the positions of many of the keys. 

Once they master the typing skills, they could accurately locate the key and move their fingers 

with minimum effort. For tasks that are relatively complex, however, attention resources will 

                                                       
4
 Because the focus of this paper is to investigate the effect of mind wandering and focus of attention, self‐
regulation is entered as a covariate or control variable influencing both mind wandering and focus of attention.  

66 
 
remain necessary even at the later stages of performance. At the early stages of skill acquisition 

of relatively consistent tasks, self‐regulation is needed to maintain users’ focus of attention. 

However, self‐regulation remains important throughout skill acquisition phases for relatively 

complex and inconsistent tasks (Randall et al., 2014). When self‐regulation processes are 

engaged, mind wandering will be less likely to occur; thus, users are likely to experience mind 

wandering in later stages of performance for tasks that can be automated, when tasks are 

simple, or when they are very complex that are beyond the reach of a person executing them 

(Kanfer & Ackerman, 1989; Randall et al., 2014).  

3.5  Research Model 

  The research model, as illustrated in Figure 3.2, is proposed to understand and prescribe 

how EFs, perceived task complexity, mind wandering, focus of attention, and learning outcomes 

are related. Two dependent variables—software self‐efficacy and learning performance are 

selected. Building upon my previous study (see Chapter 2), I differentiate mind wandering (IT) 

from mind wandering (non‐IT), and focus of attention (IT) from focus of attention (non‐IT). 

Consistent with McVay and Kane’s executive control‐failure theory of mind wandering, I 

hypothesize that both mind wandering and focus of attention are determined by perceived 

complexity and EFs. In turn, learning performance is determined by EFs, mind wandering and 

focus of attention. Drawing upon the concept of attention, I also hypothesize that software self‐

efficacy is influenced by mind wandering, focus of attention, and the interactions between the 

two constructs. In turn, the effect of mind wandering and focus of attention on learning 

performance is mediated by software self‐efficacy. The content of mind wandering is 

67 
 
considered important when mind wandering is hypothesized as the determinants of both 

software self‐efficacy and learning performance. The construct definitions are presented in 

Table 3.2.   

 
 
Figure 3.2. Research Model 
 
Table 3.2 

Construct Definitions 

Construct  Definition 
Executive functions  Functions required for maintaining and monitoring goals in the 
face of distraction and interference and for suppressing distracting 
events (Engle & Kane, 2004; Engle, Tuholski, Laughlin, & Conway, 
1999; Kane et al., 2007). 
Perceived task complexity  The degree to which the task is perceived as difficult to 
understand and execute (Bala & Venkatesh, 2013). 
Mind wandering (IT)  Task‐unrelated thought which occurs spontaneously and the 
content is related to the aspects of computer systems. 
Mind wandering (Non‐IT)  Task‐unrelated thought which occurs spontaneously and the 
content is unrelated to the aspects of computer systems. 
Focus of attention (IT)  Task‐related thought which occurs when individuals think about 
the aspects of the computer systems used to complete the 
primary task at hand. 
(table continues) 

68 
 
Table 3.2 (continued) 

Construct  Definition 
Focus of attention (Non‐IT)  Task‐related thought which occurs when individuals think about 
the primary task, but this thought is not related the aspects of 
computer systems. 
Software self‐efficacy  “Individual’s feeling of self‐efficacy relative to a specific software 
package” (Agarwal, Sambamurthy, & Stair, 2000, p. 422).  
IS learning performance  The sum of correctly answered questions on a particular subject.  
 

3.6  Hypotheses Development 

3.6.1  Executive Functions, Mind Wandering, and Focus of Attention 

At its current stage, the mind wandering literature has proposed several hypotheses on 

the roles of EFs (Friedman et al., 2008) in influencing mind wandering. Although EFs have long 

been considered “a unitary, general‐purpose ability” consisting of at least three major 

functions—response inhibition, updating working memory, and shifting (Friedman et al., 2008; 

Miyake et al., 2000), previous studies in mind wandering have primarily focused on the 

functions of WMC as a predictor of mind wandering (McVay & Kane, 2010; Smallwood & 

Schooler, 2006). In this study, I investigate the role of EFs in influencing mind wandering 

experience. Ignoring any of these primary dimensions of EFs may prevent researchers to gain a 

complete picture of phenomena associated with mind wandering. 

EFs are skills essential for success in education and in life (Diamond, 2013). For example, 

Borella, Carretti, Riboldi, & De Beni (2010) showed that EFs predict both math and reading 

competence in academic learning. Greater EFs reflects greater attentional capacity used to 

avoid distraction and maintain a goal (Hollis & Was, 2014; McVay & Kane, 2010). Incorporating 

mind wandering into executive control model can be accomplished by recognizing that goal 

69 
 
maintenance and focus of attention may be disrupted by mind wandering. Prior literature has 

suggested that EFs may play an important role in mind wandering (e.g., Axelrod et al., 2015; 

McVay & Kane, 2009; Smallwood & Schooler, 2006). During mind wandering, individuals 

consciously attend to their thoughts and feelings rather than to information related to the task 

or to the external environment (Smallwood & Schooler, 2009). According to executive‐control 

failure theory of mind wandering, individuals with lower EFs, relative to those with higher EFs, 

show frequent goal neglect because goal maintenance depends, in part, on the ability to resist 

interference from mind wandering (McVay & Kane, 2009; Mrazek et al., 2012). Simply put, 

individuals with higher EFs will have a better ability to sustain attention to the ongoing task 

demands, relative to individuals with lower EFs (McVay & Kane, 2009). Consistent with this 

theory,  I hypothesize that 

H1a: EFs are negatively associated with mind wandering (IT). 
  
H1b: EFs are negatively associated with mind wandering (Non‐IT). 
 

There is substantial evidence that EFs play an important role in learning (e.g., Duncan et 

al., 2007; Gathercole, Pickering, Knight, & Stegmann, 2004; St Clair‐Thompson and Gathercole, 

2006). People with high EFs generally earn better school grades, attain more education, and 

less likely to be involved in automobile accidents (Kane et al., 2007). EFs determine individuals’ 

ability to monitor and encode incoming information (i.e., working memory function), 

deliberately inhibit dominant or automatic responses (i.e., inhibition), and move backwards and 

forwards between multiple cues or tasks (i.e., shifting) (Miyake et al., 2000). Individuals with 

high EFs are able to control their attention, behavior, and thoughts to override a strong internal 

interruption; EFs also enables individuals to selectively attend, focusing on what is important 

70 
 
and suppressing attention to other stimuli (Diamond, 2013). Individuals with high EFs are able 

to maintain their focus of attention and process relevant details to build a mental model of the 

learning materials for comprehension (McVay & Kane, 2012b). Thus, I hypothesize that EFs have 

a positive effect on both focus of attention (IT) and focus of attention (non‐IT). 

  H2a: EFs are positively associated with focus of attention (IT). 

  H2b: EFs are positively associated with focus of attention (non‐IT). 

   

3.6.2  Perceived Task Complexity, Mind Wandering, and Focus of Attention 

  Task complexity is considered one of the key factors of cognitive load (Xu, Benbasat, & 

Cenfetelli, 2014). Task complexity can increase demand for conceptual processing by increasing 

the number of information cues that must be processed (Wood, 1986; Zhu & Watts, 2010). 

Despite some evidence indicates that mind wandering varies with the task demands 

(Smallwood & Schooler, 2006), task demands have been shown to negatively associated with 

mind wandering. According to McVay and Kane’s executive‐control failure theory (2010), “the 

decrease in mind wandering that accompanies increases in task demands…reflects the initiation 

of controlled processing to block interfering, off‐task thoughts…” (p. 191). Thus, the negative 

relationship between task complexity and mind wandering is the result of a proactive initiation 

of EFs in response to task demands. A high complexity task initiates a greater degree of 

controlled processing than does a simple task. Executive control processes, then, are triggered 

to block irrelevant cues and shield the relevant information (McVay & Kane, 2010). Consistent 

with McVay and Kane (2010), I hypothesize that 

  H3a: Perceived task complexity is negatively associated with mind wandering (IT). 
 

71 
 
H3b: Perceived task complexity is negatively associated with mind wandering (non‐IT). 
 
 
Research suggests that increasing subject difficulties can increase learning effectiveness 

(Hughes et al., 2013; Roediger & Karpicke, 2006). Individuals who perceive the task to be 

complex are more likely to focus their attention on relevant cues and exclude some information 

cues that are perceived to be less relevant (Gupta, Li, & Sharda, 2013). Conversely, the 

perception of a task being simple may encourage users to process more available information 

cues, including the less relevant ones because a simple task is less demanding and therefore, is 

likely to lead to less overall mental workload. Thus, I hypothesize that  

H4a: Perceived task complexity is positively associated with focus of attention (IT). 

H4b: Perceived task complexity is positively associated with focus of attention (non‐IT). 
 

3.6.3  Focus of Attention, Mind Wandering, and Software Self‐Efficacy 

  In this study, I specifically focus on software‐specific self‐efficacy as a cognitive learning 

outcome. Software specific self‐efficacy has been primarily measured as the outcome of 

computer or software training, education, and technology acceptance (Marakas, Yi, & Johnson, 

1998). Software self‐efficacy is distinct from general computer self‐efficacy (Marakas et al. 

1998). Whereas computer self‐efficacy is a generalized individual trait, software self‐efficacy 

refers to “individual’s feeling of self‐efficacy relative to a specific software package” (Agarwal et 

al., 2000, p. 422). Thus, software self‐efficacy is a “particularized judgment” of efficacy related 

to a specific type of application or system (Agarwal et al., 2000).  

  Software self‐efficacy is associated with a specific task performance and thus, is “a 

cognition about that specific performance” (Marakas et al., 1998, p. 129). Software self‐efficacy 

72 
 
is derived from the combined determinants associated with the formation of a task‐specific 

self‐efficacy perception (Marakas et al., 1998). The underlying rationale for understanding 

software self‐efficacy comes from social cognitive theory (Bandura, 1986). According to this 

theory, individuals generally assess the availability of specific  resources and constraints for 

performing the task (Gist & Mitchell, 1992). The formation of self‐efficacy can involve extensive 

cognitive activities or very limited cognitive activities (often called automatic processing) (Gist & 

Mitchell, 1992). It is suggested that judgements about efficacy become more automatic as 

experience with tasks increase. When tasks are novel (e.g., learning on how to use a new 

system), individuals tend to assess the task demands and environmental constraints to perform 

self‐efficacy (Gist & Mitchell, 1992). Individuals can only learn about a specific subject (in this 

case, the subject is technology or application) if they allocate their attention to relevant 

information (Yi & Davis, 2003). In the context of technology learning, especially at the early 

stage of learning, relevant information is generally associated with technical cues. Given 

software self‐efficacy reflects individuals’ perception of efficacy in performing particular 

computer‐related tasks, I hypothesize that focus of attention (IT) is positively associated with 

software self‐efficacy, whereas focus of attention (non‐IT) is negatively associated with 

software self‐efficacy. Attention directed towards task strategies and task completion (e.g., 

thinking about the rules of the task) is less needed when software self‐efficacy is set as the end 

goal because it interferes with users’ focus of attention toward technology‐related cues.  

  H5a: Focus of attention (IT) is positively associated with software self‐efficacy. 
 
  H5b: Focus of attention (non‐IT) is negatively associated with software self‐
  efficacy. 
 

73 
 
  Whereas focus of attention is hypothesized as a positive antecedent of software self‐

efficacy, mind wandering is predicted to be a negative determinant of software self‐efficacy. In 

learning, people must rely on their knowledge to construct opinions and to weight and 

integrate predictive factors (Bandura, 1993). In order to construct an appropriate mental 

model, individuals need to allocate their attentional resources to on‐task activities associated 

with learning a new system. Engaging in mind wandering during learning distracts users and 

reduces the time available to practice new skills (Gravill & Compeau, 2008). Mind wandering 

suppresses the brain’s response to external stimuli that are either relevant or irrelevant to the 

primary task (Schooler et al., 2014; Smallwood et al., 2008). Thus, I expect that individuals who 

direct their effort toward task‐irrelevant stimuli and distracting activities would demonstrate 

lower levels of software self‐efficacy. In addition, although mind wandering about technology‐

related issues may be disruptive, it could be an indication that users have high levels of general 

computer self‐efficacy. In that sense, users use mind wandering as a ‘mental escape’ from the 

current task by thinking about unrelated technology. Thus, I hypothesize that   

H6: The negative association between mind wandering (non‐IT) and software self‐
efficacy is stronger than between mind wandering (IT) and software self‐efficacy.  
 
 
  Although the interaction between focus of attention and mind wandering has not been 

tested in the previous studies, research has shown that achieving a balance between focus of 

attention and mind wandering may aid individuals in optimizing their performance outcomes 

(Allen et al., 2013). In the learning context, successful learning requires individuals to integrate 

information from environment with their own internal representation (Smallwood et al., 

2007a). There is an indication that individuals’ ability to flexibly switch between and integrate 

74 
 
across different information sources contributes to this successful learning (Allen et al., 2013), 

which can be measured by estimating the relationship between self‐efficacy and performance 

(Marakas et al., 1998). 

  As described previously, by focusing attention on the application independent of the 

task being performed with it, we can assess an individual’s perception of ability to use the 

system (Marakas et al., 1998). Although empirical evidence concludes that mind wandering is 

negatively associated with self‐efficacy, mind wandering is an essential element of a healthy, 

satisfying mental life (McMillan, Kaufman, & Singer, 2013). Hence, mind wandering can 

potentially be beneficial for our cognitive learning process.  

  The interaction between mind wandering and focus of attention can be explained by the 

implicit learning perspective (Kaufman, 2013). According to this perspective, people are able to 

learn complex rules and relations between events that they encounter in the external 

environment without being consciously aware of them (Dijksterhuis & Aarts, 2010). This implicit 

learning process is goal dependent and requires at least some attentional resource (Dijksterhuis 

& Aarts, 2010). For example, Jimenez and Mendez (1999), as cited by Dijksterhuis and Aart 

(2010), conducted a number of experiments in which participants engaged in sequential 

learning under either a single‐ or dual‐task condition. Although they found that general 

attentional load hardly influenced learning, it was obvious that participants needed selective 

attention to the crucial information. Implicit learning theory speculates that intelligence is an 

adaptation process to task demands that are relevant to attaining personal goals (e.g., mind 

wandering), not just adaptation to the external goals (e.g., primary task) (Kaufman, 2013). 

Individuals’ controlled forms of cognition (e.g., attentional focus) and spontaneous forms of 

75 
 
cognition (e.g., intuition, affect, insights, and spontaneous triggering of episodic memories) are 

important potential contributors to the personal learning process (McMillan et al., 2013). Thus, 

the interaction effect between focus of attention and mind wandering relationship is worth 

investigating.  

  The interaction between mind wandering and focus of attention will be beneficial if the 

contents of both thought processes are closely related. According to the diversion thought 

paradigm, when individuals are engaged in diversionary thought (or mind wandering), their 

attention is shifted from current tasks to a different mental context induced by a mental 

diversion (Alves, Resende, & Pinho, 2015; Sahakyan & Kelley, 2002). This diversionary thought 

experience is associated with the difficulty of recalling information presented prior to the mind 

wandering episodes (Alves et al., 2015). However, Delaney et al. (2010) indicated that the 

magnitude of the mind wandering effect on forgetting information or memory lost depends on 

the mental distance (e.g., temporal, circumstantial, etc.) from the current moment. For 

instance, mind wandering about irrelevant technology while working on a technology task is 

likely to be less harmful to learning outcomes than mind wandering about something else.  

  This argument is in line with the implicit learning perspective discussed previously. As a 

human being, we have this ability to automatically learn, with or without our conscious 

awareness (Kaufman et al., 2010). This implicit learning is typically characterized by a set of 

automatic, associative, nonconscious, and unintentional thought processes (Kaufman et al., 

2010). In this context, learning is often termed as associative, when learning proceeds 

incidentally, because it describes the incidental formation of associations (Kaufman et al., 

2010). Given the fact that implicit learning requires the formation of associations, it is 

76 
 
reasonable to argue that the interaction between focus of attention (IT) and mind wandering 

(IT) will trigger learning cues that can potentially increase users’ software‐specific self‐efficacy, 

whereas the interaction between focus of attention (IT) and mind wandering (non‐IT) will slow 

down the learning process.  

  Further, prior studies on distraction have found that distracting features perceived to 

contain information related to the primary task elicit less irritation and avoidance (Gluck, bunt, 

& McGrenere, 2007; Tan, Yi, & Chan, 2015). For example, when faced with distracting website 

features, users will assess and be influenced by the congruence between the distracting 

content and the primary task (Tan et al., 2015). This argument can also be extended to 

understand the effect of mind wandering on learning outcomes. For example, for a student 

learning how to use Excel, mind wandering about other technologies, such as social media 

would probably be perceived as more task congruent and less disruptive than mind wandering 

about a family member. In contrast, letting our mind wander about other issues unrelated to 

technology (e.g., temporal content, personal worry, and self‐relevance) (i.e., mind wandering 

(non‐IT))  can be viewed as random thought and may be related to personal problem solving. In 

its relationship with the primary task, it can serve as a preparation function, motivating 

individuals to engage in problem solving behaviors and adopt adaptive behaviors that reduce 

potential threats (Watkins, 2008). In that sense, the outcomes of the interaction between focus 

of attention and mind wandering are determined by the thought content. In particular, the 

negative effect of focus of attention (non‐IT) on software self‐efficacy can be diminished if users 

experience high levels of mind wandering (non‐IT). To the same extent, the positive effect of 

focus of attention (IT) on software self‐efficacy is stronger for users who mind wander about 

77 
 
technology than for those who mind wander about non‐technology‐related content. Hence, I 

postulate that the content of mind wandering that is congruent with the context of the primary 

task to be beneficial, whereas mind wandering that is incongruent with the primary task to be 

destructive (see Table 3.3). I hypothesize the following: 

H7: Mind wandering moderates the relationship between focus of attention and software 
specific self‐efficacy, in which   
 
 H7a: The positive effect of focus of attention (IT) on software self‐efficacy is higher 
when users experience high mind wandering (IT) than when they experience low mind 
wandering (IT). 
 
 H7b: The positive effect of focus of attention (IT) on software self‐efficacy is higher 
when users experience low mind wandering (non‐IT) than when they experience high 
mind wandering (non‐IT). 
 
 H7c: The negative effect of focus of attention (non‐IT) on software self‐efficacy is higher 
when users experience low mind wandering (non‐IT) than when they experience high 
mind wandering (non‐IT). 
 
 H7d: The negative effect of focus of attention (non‐IT) on software self‐efficacy is higher 
when user experience high mind wandering (IT) than when they experience low mind 
wandering (IT). 
 

  From an evolutionary perspective, the temporal dynamics of mind wandering allows us 

to simulate plausible outcomes of alternative plans, including the emotional states of ourselves 

and other people in response to such plans (Gilbert & Wilson, 2007; Oettingen & Schwörer, 

2013). When mind wandering triggered by contextual factors, such as affective and temporal 

content, mind wandering can work as a vehicle for learning and problem solving. It is likely 

because when one focuses on his or her attention on the relevant cues, mental contrasting 

strengthens the associations between the problem and instrumental means to overcome the 

78 
 
problem (Oettingen & Schwörer 2013). Hence, I state a three‐way relationship between focus 

of attention, mind wandering, and software self‐efficacy.  

 H7e: The positive effect of focus of attention (IT) on software self‐efficacy is higher 
when users experience high mind wandering (IT) and low mind wandering (non‐IT).  
 
 H7f: The negative effect of focus of attention (non‐IT) on software self‐efficacy is higher 
when users experience high mind wandering (IT) and low mind wandering (non‐IT). 
 
Table 3.3 
Interactions Between Focus of Attention and Mind Wandering (Dependent Variable: Software 
Self‐Efficacy) 
 
    Mind Wandering (MW) 
    IT  Non‐IT 
IT 
Congruent  Incongruent 
 
Focus of Attention 
Non‐IT 
Incongruent  Congruent 
 
 

3.6.4  Focus of attention, Mind Wandering, and Learning Performance 

  The effect of focus of attention on learning has been discussed in the education 

literature (e.g., Fredricks, Blumenfeld, & Paris, 2004; Risko, Buchanan, Medimorec, & Kingstone, 

2013; Skinner, Furrer, Marchand, & Kindermann, 2008; Skinner, Kindermann, & Furrer, 2009). 

According to this literature, focus of attention is a strong predictor of school success (Risko et 

al., 2013). Focus of attention represents an important part of effective learning (e.g., Risko et 

al., 2013; Yi & Davis, 2003). Without focus of attention, information quickly fades and can 

hardly be remembered after time passes. Whereas focus of attention is positively associated 

with learning performance, mind wandering is negatively associated with learning performance 

(Smallwood et al., 2007a). For example, Lindquist and McLean (2011) conducted a correlation 

79 
 
study to investigate the relationship between mind wandering and course performance. They 

found that as mind wandering increased, the course performance decreased. In this study, I 

measure learning performance by summing of correctly answered questions on a particular 

subject (i.e., declarative knowledge or immediate performance). I postulate that the interaction 

effects between mind wandering and focus of attention are not substantial because the 

primary antecedent resources for immediate performance will be from users’ working memory 

consolidation.  

  Unlike learning in general, IS learning requires students to pay attention to not only the 

context of the task, but also to the technical aspects of the system being learned. One of the 

primary outcomes of technology learning is knowing how computer programs or other 

applications work and operate (Palvia and Palvia, 2007). In this case, focusing their attention on 

important features of software is critical in determining their learning performance. In contrast, 

although focus of attention (non‐IT) has been considered an important element of task 

performance, I argue that it is less salient at the early stage of technology learning. It may 

potentially impair task performance because it demands the limited resources that can be 

allocated to focus of attention (IT). Thus, I hypothesize that 

  H8: The positive effect of focus of attention (IT) on learning performance is higher than 
the positive effect of focus of attention (non‐IT) on learning performance. 
   

Mind wandering represents “the withdrawal of executive resources from one’s 

ostensible primary task toward mental pursuit of other personal goals, thus leaving fewer 

resources for the primary task” (Kane et al., 2007, p. 620). In other words, mind wandering 

competes with task performance on a limited capacity of attentional resources (Smallwood & 

80 
 
Schooler, 2006). Thus, whereas allocation of attentional resources to the (primary) task has a 

positive influence on performance, resources allocated to mind wandering exert a direct, 

negative influence on task performance (Kanfer & Ackerman, 1989). Mind wandering has been 

linked to poor outcomes in a wide range of tasks, such as those common in education (e.g., 

McVay & Kane, 2011; Schooler et al., 2004). When mind wandering intrudes during task 

completion, task performance impairs, especially if the task is complex and requires focused 

attention (Gruberger, Simon, Levkovitz, Zangen, & Hendler, 2011; McVay & Kane, 2009; McVay 

& Kane, 2012a; Mrazek et al., 2012; Unsworth et al., 2012). However, mind wandering can 

produce positive or negative outcomes, depending on the behavioral context (Allen et al., 

2013). Mind wandering can be used to infer one’s attitude toward technology (Critcher & 

Gilovich, 2010). To the extent that one needs to focus on tasks, yet his or her mind wander 

about other irrelevant technology, it could be a signal that the primary task is less enjoyable or 

he or she has a relatively high level of general computer self‐efficacy. Thus, although mind 

wandering itself has a negative effect on performance, mind wandering (IT) has less negative 

effect on performance than mind wandering (non‐IT).  

  H9: The negative effect of mind wandering (non‐IT) on IS learning performance is higher 
than the negative effect of mind wandering (IT) on IS learning performance. 
 

3.6.5  Software Self‐efficacy and Learning Performance 

  The effect of self‐efficacy on task performance has been well established in the 

literature (e.g., Locke, Frederick, Lee, & Bobko, 1984; Marakas et al., 1998; Yee, Hsieh‐Yee, 

Pierce, Grome, & Schantz, 2004). Individuals who have a high sense of efficacy are likely to 

develop strategies to maximize their performance (Bandura, 1993). In particular, at the early 

81 
 
stage of technology learning, it is hypothesized that as software self‐efficacy is increased, 

performance at the specific task level is improved (Marakas et al., 1998). Thus, I hypothesize 

that 

H10: Software self‐efficacy is positively associated with IS learning performance. 

3.6.6  EFs and Learning Performance  

  Kane and McVay (2012) argued that vulnerability to mind wandering may contribute to 

the role of EFs in predicting cognitive performance and abilities. Given the purpose of learning 

is to increase one’s particular knowledge, skill or behavior, immediate learning performance 

can be viewed as a cognitive outcome—“the knowledge users have about what a technology is 

and how to use it” (Marcolin, Compeau, Munro, & Huff, 2000, p. 39). When learning something 

new, those who are able to utilize their EFs most often perform best (Diamond, 2013). It is 

because individuals with high EFs have more cognitive resources to maintain task relevant 

information, thereby facilitating performance (Kanfer & Ackerman, 1989). Thus, I hypothesize   

  H11: EFs are positively associated with IS learning performance. 

3.6.6  Control Variables 

 Self‐regulation. Self‐regulation provides the very basic for purposeful action 

(Bandura, 1991). Self‐regulation enables individuals to set goals for themselves and 

then plan courses of action that are likely to produce desired outcomes (Bandura, 

1991). Given self‐regulation is an important mechanism for the allocation policy to 

82 
 
ensure limited resources are properly directed to a task, self‐regulation is included 

as a control variable on focus of attention and mind wandering.  

 User Demographics. Prior studies have shown that gender and age play a significant 

role on mind wandering. It is plausible that mind wandering decreases with age 

(Giambra, 1993; Smallwood et al., 2007b). Research also shows that computer self‐

efficacy varies across gender (Busch, 1995). Thus, gender and age are added as 

control variables on focus of attention, mind wandering, and learning outcomes.  

 General Computer Self‐efficacy. General computer self‐efficacy potentially 

influences software self‐efficacy (Agarwal et al., 2000; Yi & Davis, 2003). Thus, 

general self‐efficacy is added as a control variable on software self‐efficacy. 

3.7  Research Method 

3.7.1  Research Context and Sample 

  I selected Microsoft Access learning as the context in which the research model is 

empirically tested. The study consisted of two different stages administered in two sessions 

(i.e., online survey and online cognitive experiment) that lasted approximately 40 minutes in 

total. The study subjects – students enrolled in an introductory IS course – were required to 

learn Microsoft Access. Given students taking the course had very limited knowledge of 

Microsoft Access, it was considered an appropriate context for the study that measures 

learning outcomes.  

  Data were collected by administering a web‐based survey, which was considered 

appropriate since the target respondents were individuals who used IS as a part of their 

83 
 
profession. Undergraduate students majoring in business at a large university in the United 

States were recruited to participate in the study as an exchange for course credit. Despite 

concerns regarding students subjects (Compeau, Marcolin, Kelley, & Higgins, 2012), the 

selected sample did not introduce undue bias because research has shown that students used a 

number of technologies for both learning and socialization (e.g., Margaryan, Littlejohn, & Vojt, 

2011). Thus, using a student sample is relevant in the context of this study because students 

interact with technology when they learn about the computer concepts in their core IS courses 

(e.g., database management, systems analysis and design, etc.). 

In the first stage of the study, participants were instructed to complete a specific 

chapter of Microsoft Access as a part of their class assignment. They were instructed to fill out 

the questionnaire immediately after they completed their assignment. I worked with the course 

instructors to determine the deadline for both the assignment and survey. Participants were 

asked to report their thought experiences during their interaction with Microsoft Access. In the 

second stage, participants performed three EFs tasks—Stroop task, Color Shape, and N‐Back 

task. All three tasks were computerized using Inquisit Web Millisecond Version 4 software 

package (see www.millisecond.com for details).  

In total, a sample of 326 valid responses was included in the data analysis. Of the 326 

respondents, 53 percent were males and over 65 percent were in the 21 to 25 age range. Over 

85 percent of the respondents were English native speakers. 

84 
 
3.7.2  Operationalization of Research Constructs 

  The measurement items for focus of attention and mind wandering developed in 

Chapter 2 were used. Items to measure other principal constructs were adapted from existing 

measures (see Appendix B). In particular, items to measure task complexity were adapted from 

Seijts, Latham, Tasa, and Latham (2004), Winter and Latham (1996), and Jiang and Punj (2010); 

items to measure computer self‐efficacy were adapted from Kanfer, Ackerman, Murtha, 

Dugdale, and Nelson (1994) and Thatcher and Perrewe (2002). To develop the items for 

software self‐efficacy, the procedures used by Agarwal et al. (2000) were followed. First, a task 

associated with Ms. Access was selected on the basis of the key features and functionality of 

the software. Finally, students were asked to complete an assignment that required them to 

use specific features of the software. These features were included as items for measuring 

software self‐efficacy. Learning performance was measured using 10 multiple‐choice questions 

designed to assess participants’ understanding of the concepts and features needed to use the 

software appropriately (see Appendix C). Self‐regulatory activities measures were adapted from 

Kanfer and Ackerman (1989) and Grafill and Compeau (2008). All constructs were measured 

reflectively on seven‐point Likert scales.  

  To measure EFs, the recommendations made by Miyake et al. (2000) were closely 

followed. First, I selected three cognitive tasks—N‐Back, Color Shape, and Stroop Task. Stroop 

task was used to measure inhibition; color‐shape task was intended to measure shifting 

attention or flexibility; and N‐back was used to measure WMC. The order or stimuli within each 

task was the same for all participants. All tasks included additional practice trials to ensure that 

the participants understood the instructions and that they had mastered the rules for each task 

85 
 
(Friedman et al., 2008). The summary of the executive function tasks, including their 

references, is presented in Table 3.4.  

Table 3.4 

Executive Functions Measures 

Type of executive  Example 
Behavior  Key Dependent 
Description  functions often  References of 
Paradigm  Measure5 
invoked  Review 
Stroop task   On control trials, rectangle  The proportion  Inhibition  Friedman et al. 
shapes are printed in one of four  of correct  (2008); MacLeod 
colors (red, green, blue, or black).  responses on  (1991); Miyake et 
 On incongruent trials, the name  incongruent  al. (2000) 
of a word (such as green) is  trials.  
written in a different color.  
 On congruent trials, word and ink 
color are the same. 
Color‐  On pure task trials, only one task  The proportion  Shifting attention  Friedman et al. 
shape task  (either color or shape judgment) is  of correct  (2008); Miyake, 
performed in a given block of  responses on  Emerson, Padilla, 
trials.   mixed trials.  and Ahn (2004)  
 On mixed trials, participants 
performed several blocks of 
mixed‐task trials (i.e., repeat and 
switch trials) in which each trial is 
preceded by a cue indicating 
which task to perform. 
N‐Back  Participants are shown a sequence  The proportion  Updating/working  Jaeggi et al. 
of visual stimuli and they have to  of hits minus  memory capacity  (2010); Ophir, 
respond each time the current  false alarms.   Nass, and 
stimuli is identical to the one  Wagner (2009) 
presented ‘n’ positions back in the 
sequence.  

                                                       
5
 Reaction time (RT) differences were also collected in Stroop and color‐shape tasks. Although prior studies have 
primarily relied on RT, recent research has emphasized the importance of accuracy measures (Unsworth & Engle, 
2008). McElree (2001) suggested that whereas RT provides an index of the accessibility of an item, accuracy 
reflects the availability of the item. A further analysis also revealed that the score of N‐Back is highly correlated 
with the levels of accuracy of Stroop and color‐shape task (r = .43, p<.001 and r = .31, p<.001, respectively), 
whereas it does not correlate with the RT measures. These results suggest that are the unity and diversity of EFs 
are better represented using the accuracy measures. Thus, only the proportions of correct trials (after controlling 
for RT) were used as the dependent measures of Stroop and color‐shape task.   

86 
 
Stroop task was adopted from MacLeod (1991) and Miyake et al. (2000). It is intended 

to measure inhibition. Three trials were used in this study—control, congruent, and 

incongruent trials. On the incongruent trials, a color word was printed in a different color; on 

the congruent trials, a color word was printed in the same color; and on the control trials, 

rectangle shapes were printed in one of four colors (red, green, blue, or black). Color‐shape task 

was adopted from Miyake et al. (2004)  and Miyake and Friedman (2012). It is used to measure 

users’ switching ability. In the first part, participants performed several blocks of pure‐task trials 

in which only one task (either a color or shape task) was performed in a given block of trials. In 

the second part, participants performed several blocks of mixed‐task trials in which each trial 

was preceded by a cue indicating which task to perform (Miyake et al., 2004). N‐Back task was 

adopted from Jaeggie et al. (2010) and used to measure working memory capacity. In this task, 

participants were shown a sequence of visual stimuli and they were instructed to respond each 

time the current stimulus was identical to the one presented n positions back in the sequence. 

The detailed procedures of these tasks are presented in Appendix D. 

3.8  Data Analysis and Results 

3.8.1  Measurement Model 

  Before testing the hypothesized structural model, psychometric properties of the items 

for all the constructs measured by self‐report questionnaires as well as EFs measures were 

evaluated through confirmatory factor analysis (CFA). For the EFs measures, the trimming 

procedure was first conducted. This procedure helps us control for the RT measures. The upper 

and lower criteria were determined on the basis of overall, between‐subjects RT distributions. 

87 
 
Overall, only 2.2 percent of total observed values were affected by this trimming procedure. For 

the Stroop task, only correct trials longer than 200 ms were analyzed. Any extreme values (i.e., 

> 5000 ms) were removed. For the color‐shape task, any extreme RTs (i.e., < 100 ms or > 5000 

ms) were excluded from the computation of correct responses (Miyake et al. 2004). The 

proportion correct measures of color‐shape and Stroop task were computed and arcsine 

transformed to improve normality. This transformation is useful for creating more dispersion in 

ceiling and floor effects, while having little influence for accuracy scores in the range .20‐.80 

(Miyake et al., 2004). The key dependent measures of Stroop and color‐shape frequency 

achieved a satisfactory level of normality after the transformation process.  

Overall accuracy levels of Stroop task were high (M = .89, SE = 00), but the count for 

control trials (M = .95; SE = .00) was higher than the count for incongruent trials (M = .84, SE = 

.01; t322 = 11.63, p<.001). These results suggest that the incongruent trials required suppressing 

dominant or automatic responses, leading to higher levels of error, relative to the control trials. 

Further, the results suggest that overall accuracy of color‐shape task was high (M = .74, SE = 

.01), but the count for repeat trials was slightly more accurate (M = .76, SE = .01) than the count 

for switch trials (M = .71, SE = .01; t322 = 3.09, p<.05), suggesting that participants made more 

errors on the switch trials.  

For N‐Back task, the proportion of hits minus false alarm (i.e., the number of times a 

participant identifies a non‐target as a target) averaged over all n‐back levels was computed as 

the dependent measure (Jaeggie et al., 2010). The hit‐rates (the number of targets correctly 

identified) in the two‐back task (M = .60, SE= .01) was higher than in the three‐back task (M = 

88 
 
.46; SE= .01) and the four‐back task (M = .41; SE = .01). These results suggest that filtering 

irrelevant representations in memory becomes harder as the n‐back level increases.  

  The measurement model was assessed using the partial least squares (PLS) method to 

examine internal consistency, reliability, convergent, and discriminant validity. Internal 

consistencies (CR) of .70 or higher are considered adequate (Compeau, Higgins, & Huff, 1999; Yi 

& Davis, 2003). The composite reliability (CR) values for all of the constructs are greater than 

.70, demonstrating that all constructs have adequate reliability scores. Convergent and 

discriminant validity was assessed using two criteria: (1) the standardized item loadings should 

be at least .70, and items should load more highly on their intended constructs than on other 

constructs (Compeau et al., 1999) and (2) the square root of the average variance extracted 

(AVE) should be greater than .50 and should exceed the correlation scores with other 

constructs (Fornell & Larcker, 1981). The results of CFA, reported in Table 3.5, indicate high 

levels of convergent validity. Each factor loading is strong and significant (all p values < .01). The 

only exception is Stroop task, with the factor loading score of .58. I decided to retain this 

dimension for two reasons. First, it is a crucial part of EFs. Removing this dimension could lead 

to an erroneous conclusion about the findings. Second, the standardized factor loading of 

Stroop task reported in the previous studies (e.g., Miyake & Friedman, 2012; Miyake et al., 

2000) was relatively low (ranging from .46‐.50) and this low score can be explained by the 

unity/diversity framework of EFs—each EFs dimension shares a common variance while at the 

same time, has a unique variance by itself (Miyake & Friedman, 2012).  

89 
 
Table 3.5 

Confirmatory Factor Analysis (PLS Results) 

Factor 
Constructs/Items  Mean  SD 
Loading 
Mind Wandering (non‐IT) (CR; AVE)      (.95; .68) 
1. I thought about members of my family.  3.66  1.63  .81 
2. I thought about friends.  3.72  1.65  .84 
3. I thought about something that made me feel guilty.  2.91  1.75  .77 
4. I thought about personal worries.  3.72  1.72  .82 
5. I thought about something that made me feel angry.  2.98  1.74  .86 
6. I thought about something that happened earlier today.  3.73  1.71  .84 
7. I thought about something that happened in the recent past (last few days,  3.67  1.71  .86 
but not today). 
8. I thought about something that happened in the distant past.  3.15  1.74  .86 
9. I thought about something that might happen in the future.   3.85  1.76  .78 
Mind Wandering (IT) (CR; AVE)      (.94; .77) 
1. I thought about checking my email.  3.70  1.78  .83 
2. I thought about checking my social media (e.g., Facebook).  3.87  1.87  .91 
3. I thought about browsing other stuff.  4.02  1.77  .93 
4. I thought about checking my phone.   4.46  1.73  .82 
5. I thought about doing other online activities (e.g., online shopping, online  3.80  1.95  .89 
game). 
Focus of Attention (IT) (CR; AVE)      (.95; .72) 
1. I focused on how to use MS. Access to complete my task.  4.88  1.49  .87 
2. I paid attention on different features of MS. Access to complete my task.  4.88  1.42  .89 
3. I focused my attention on different options provided by MS. Access to  4.74  1.45  .88 
format my task.  
4. I focused my attention on learning specific MS. Access functions.  4.82  1.45  .89 
5. I paid close attention to the kind of errors I was making when I was using  4.92  1.41  .84 
MS. Access.  
6. I focused my total attention on making sure my computer outputs were  5.14  1.40  .81 
correct. 
7. I focused my total attention on following the MS Access’s instructions  5.40  1.42  .71 
correctly. 
Focus of Attention (non‐IT) (CR; AVE)      (.90; .70) 
1.  I focused my total attention on how I could solve the problem.  4.48  1.63  .74 
2. I thought about strategies for solving the problem.  4.60  1.52  .89 
3. I thought ahead to what I would do next to solve the problem.  4.66  1.50  .91 
4. I focused my attention on correctly performing as many steps as I could.  4.58  1.67  .79 
(table continues) 
 

90 
 
Table 3.5 (continued) 

Factor 
Constructs/Items  Mean  SD 
Loading 
Perceived Task Complexity (CR; AVE)      (.89; .72) 
    1. This task required me to coordinate many different things at the same  4.37  1.45  .82 
time. 
2. This task required a lot of my effort.  4.54  1.49  .89 
3. On scale 1 (extremely easy) to 7 (extremely difficult), how difficult was this  4.07  1.46  .84 
task? 
Executive Functions (EFs)      (.79; .55) 
1. N‐Back task  ‐0.93  2.08  .83 
2. Color‐shape task   0.87  0.15  .79 
3. Stroop task  0.83  0.19  .58 
Software Self‐Efficacy      (.97; .92) 
1. After I did my assignment, I believe I have the ability to export data from  5.40  1.24  .96 
MS. Access to Ms. Word format. 
2. …I believe I have the ability to export data from MS. Access to MS. Excel  5.43  1.19  .96 
format 
3. …I believe I have the ability to export data from MS. Access to PDF format.  5.38  1.25  .95 
4. …I believe I have the ability to export data from MS. Access to another  5.41  1.19  .94 
database.  
Self‐regulation (the composite score of each dimension was reported)      (.84; .64) 
1. Self‐monitoring  3.97  1.36  .88 
2. Performance evaluation  3.41  1.65  .73 
3. Self‐reaction  3.89  1.35  .78 
General Computer Self‐Efficacy      (.90; .76) 
1. I was able to use MS. Access if there was no one around me to tell me what  4.70  1.34  .89 
to do. 
2. I was able to use MS. Access if I had never used a package like it before.  4.07  1.55  .78 
3. I was able to use MS. Access if I had only the software manuals for  4.68  1.36  .93 
references. 
1. Survey Instructions: The following statements concern the kinds of thoughts that go through people’s heads at 
particular times. For example, while they are working on a task or assignment, they could think about something else, 
unrelated to their primary task. The following is a list of thoughts, some of which you might have had while doing your 
MS. Access assignment (CHAPTER 8). Please indicate approximately how often each thought occurred to you when you 
did your assignment.   
2. Scale to measure the thought content: 1 (never) – 7 (every Time); other measures: 1 (strongly disagree) – 7 (strongly 
agree) 
3. CR = Composite Reliability; AVE = Average Variance Extracted  
 

 
 

91 
 
Table 3.6 
Correlation Matrix 
   Mean  SD  1 2 3 4 5 6 7  8  9 10
1. General CSE  4.48  1.25  .87      
2. Self‐Regulation  3.75  1.16  .33** .80      
3. P. Complexity  4.51  1.12  .13* .31** .85      
4. EFsƗ  0.00  0.75  ‐.12* ‐.24** ‐.07 .75      
5. Mind Wandering (Non‐IT)  3.49  1.42  .03 .29** .25** ‐.22** .82      
6. Mind Wandering (IT)  3.97  1.60  .07 .17** .17** ‐.20** .64** .88      
7. Focus  (IT)  4.97  1.21  .28** .42** .20** .15** ‐.20** ‐.16**  .85    
8. Focus (Non‐IT)  4.58  1.32  .29** .60** .19** ‐.19** .07 .08 .46**  .84 
9. Software‐Specific SE  5.41  1.22  .23** .02 .06 .23** ‐.24** ‐.16**  .43**  .01  .96
10. Task Performance  0.70  0.17  .03 ‐.18** .01 .28** ‐.19** ‐.09 .12*  ‐.10 .24** 1.00
Notes: ** denotes significant correlations at the p < .01 level and * at p < .05 level. 
The diagonal elements (in bold) represent the square root of AVE. 
Ɨ
z‐score composite was used.  
   

The average variance extracted (AVE) value for all constructs is well above Fornell and 

Larcker’s (1981) suggested cutoff of .50. The means, standard deviations, reliabilities, AVE 

values, and factor loadings are reported in Table 3.5. The square root of the AVE for each 

construct in the model is larger than the corresponding off‐diagonal correlations of the 

constructs to their latent variables (Table 3.6). The correlations among all constructs are all well 

below the .90 threshold, suggesting that all constructs are distinct from each other. These 

results provide substantial evidence for the convergence/discriminant validity of the items and 

overall measurement model. 

Common method bias can be a potential threat to the study if the independent and 

dependent variables are obtained from the same sources (Podsakoff, MacKenzie, Lee, & 

Podsakoff, 2003). In order to determine whether common method bias was a concern, I 

performed several different tests. First, Harman’s single‐factor test was performed by including 

all items in a principal components factor analysis (Podsakoff et al., 2003). A single factor only 
92 
 
explained 28 percent of the variance, indicating no evidence of common method bias. Second, 

a partial correlation method was used (Podsakoff et al., 2003; Pavlou, Liang, & Xue, 2007). The 

highest factor from a principal component analysis was added into the PLS model as a control 

factor on all dependent factors. This factor did not significantly increase the variance explained 

in any of the dependent variables, suggesting common method bias is not a major issue. Next, I 

performed a marker variable technique suggested by Lindell and Whitney (2001). An unrelated 

construct (i.e., response bias) was measured in the survey. High correlations among any of the 

items used in the study and response bias will indicate common method bias. The average 

correlations between response bias and principal constructs was .07, suggesting lack of 

evidence of common method bias. Finally, EFs and the study’s dependent variables (i.e., 

software self‐efficacy and learning performance) were measured independently. All of these 

tests confirm that common method bias is not a major concern in this study.  

3.8.2  Structural Model   

Both regression analysis using SPSS and structural equation modeling using the PLS 

method were utilized to test the hypothesized paths in the model (Figure 3.2). PLS uses a 

component‐based approach to estimation (Chin, 1998) and thus, it places minimal demands on 

sample size and residual distributions (Chin, 1998; Fornell & Bookstein, 1982; Hair, Ringle, & 

Sarstedt, 2011). Paths are interpreted as standardized beta weights in a regression analysis. 

Although the results are not identical, both statistical packages (i.e., PLS and SPSS) yield very 

similar results. It is the author’s preference to report the standardized regression coefficients 

from SPSS. Given the dependent measures of EFs vary depending on the task paradigms, z‐

93 
 
score composite of the three EFs tasks after the arcsine transformation was used in the 

regression analysis (Cohen, Cohen, West, & Aiken, 2013). Although some scholars estimated the 

effect of each dimension of EFs on the outcome variables, it is also suggested  that each EFs 

dimension can be decomposed into one multidimensional construct consisting of three 

dimensions (i.e., common EFs). The path coefficients and explained variances for the model are 

shown in Figure 3.3. For clarity of presentation, only significant relationships are shown in 

Figure 3.3. 

  Variables involved in interactions were mean‐centered for an easy interpretation and 

reducing multicollinearity issues. In order to formally test for collinearity’s presence, I 

calculated the variable inflation factor (VIF) for focal constructs in the model. According to 

Tabachnik and Fidell (1996), the VIF greater than 10 indicates the existence of multicollinearity. 

Because the VIF’s did not exceed 3.00, the analysis indicates that collinearity did not bias the 

results. 

Figure 3.3. Structural Model 
94 
 
   

As hypothesized, EFs had a negative effect on both mind wandering (IT) and mind 

wandering (non‐IT), validating H1a and H1b. EFs had a significant effect on focus of attention 

(IT), but did not have a significant effect on focus of attention (non‐IT). Thus, H2a was 

supported, whereas H2b was not supported. Contrary to my hypothesis, perceived task 

complexity had a positive effect on mind wandering, failing to support H3. In addition, 

perceived task complexity had no significant effect on focus of attention. Thus, H4a and H4b 

were not supported. Next, whereas focus of attention (IT) had a positive effect on software self‐

efficacy, focus of attention (non‐IT) had a negative effect on software self‐efficacy, supporting 

H5a and H5b. The results also reveal that mind wandering (non‐IT) had a negative effect on 

software self‐efficacy, whereas the effect of mind wandering (IT) was not significant, validating 

H6. Three of four two‐way interaction effects between mind wandering and focus of attention 

were also significant, validating the moderating role of mind wandering on the relationship 

between focus of attention and software self‐efficacy.  

To test for the moderating effect, Cohen’s f2, which compares the R2 value of the 

interaction effect with the main effect, was performed using the following equation (Chin, 

Marcolin, & Newsted, 2003). 

  Cohen’s f2 = R2 (interaction mode) – R2 (main effects model) / [1 ‐ R2 (main effects 
  model)]  
 
Controlling for the covariates, the variance explained on software self‐efficacy was 30 

percent with two‐way interaction, 32 percent with three‐way interaction, whereas only 27 

percent with only the main effects (see Table 3.7). Cohen’s f2 for the two‐way interaction model 

was .04 and .07 for the three‐way interaction model, suggesting the three‐way interaction 

95 
 
model is a better model representation than the two‐way interaction model. Although Cohen 

(1988) suggested that f2 effect size of .03 to .10 is termed small, Kenny (2015) proposed that a 

more realistic standard for effect sizes should be .005, .01, and .025 for small, medium, and 

large respectively. Using Kenny’s standard for effect size, the effect size of .07 in this study is 

considered large. 

Table 3.7 

The Effect of Mind Wandering and Focus of Attention on the Dependent Variables 

  DV = Software‐Specific Self‐Efficacy  DV = Task Performance 
IVs  M1  M2  M3  M4  M1  M2 
R2   .07  .27  .30  .32  .02  .16 
2
(ΔR )   ‐  (.20***)  (.03*)  (.02*)  ‐  (.14***) 
Control Variables             
Gender    Ns  Ns  Ns  Ns  .14**  .14** 
Age  .12*  Ns  Ns  Ns  Ns  Ns 
General Self‐Efficacy  .23***  .17***  .17***  .18***    Ns  
Independent Variables             
Executive Functions            .19*** 
Software Self‐Efficacy             .15** 
Mind Wandering (MW)              
Technology‐related (IT)    Ns  Ns  Ns    Ns 
Nontechnology‐related (Non‐IT)    ‐.15**  ‐.16*  ‐.18**    ‐.16* 
Focus of Attention (FA)             
Technology‐related (IT)     .46***  .44***  .51***    Ns 
Nontechnology‐related (Non‐IT)    ‐.23***  ‐.21**  ‐.27**    Ns 
Two‐way Interaction             
FA  (IT) x MW (IT)      .20**  .16*     
FA (NonIT) x MW (IT)      Ns  Ns     
FA (IT) x MW (NonIT)      ‐.22**  ‐.23**     
FA (NonIT) x MW (NonIT)      .19**  .22**     
Three‐way Interaction             
FA (IT) x MW (IT) x MW (Non‐IT)        ‐.18*     
FA (Non‐IT) x MW (IT) x MW        .18*     
(Non‐IT) 
Notes: ***p<.001; **p<.01; *p<.05; Ns = Not Significant. 

   
I plot the moderating effects in Figure 3.4, 3.5, 3.6, and 3.7. As shown in Figure 3.4, the 

slopes for low mind wandering (IT) and high mind wandering (IT) were positive and significant 

96 
 
(b = 0.28, p<.001; b = 0.63, p<.001 respectively). However, as expected, at high levels of mind 

wandering (IT), software self‐efficacy increased rapidly as focus of attention (IT) increased. 

Thus, H7a was supported. As shown in Figure 3.5, at low levels of mind wandering (non‐IT), 

software self‐efficacy increased significantly as focus of attention (IT) increased (b = 0.66, 

p<.001). However, at high levels of mind wandering (non‐IT), software self‐efficacy increased 

marginally as focus of attention (IT) increased (b = 0.24, p<.001), validating H7b. Further, as 

shown in Figure 3.6, the negative effect of focus of attention (non‐IT) on software self‐efficacy 

was stronger at low levels of mind wandering (non‐IT) (b = ‐0.35, p<.001) than at high levels of 

mind wandering (non‐IT) (b = ‐0.05, p = .43), supporting H7c. As shown in Figure 3.7, the 

moderating effect of mind wandering (IT) on the relationship between focus of attention (non‐

IT) and software self‐efficacy was not significant (the slope difference between high and low 

mind wandering was not statistically significant). Thus, H7d was not supported.  

 
Figure 3.4. The Moderating Effect of Mind Wandering (IT) on the Relationship between Focus of 
Attention (IT) and Software Self‐Efficacy 

97 
 
 
Figure 3.5. The Moderating Effect of Mind Wandering (Non‐IT) on the Relationship between 
Focus of Attention (IT) and Software Self‐Efficacy 
 

 
Figure 3.6. The Moderating Effect of Mind Wandering (Non‐IT) on the Relationship between 
Focus of Attention (Non‐IT) and Software Self‐Efficacy 
 

98 
 
 
Figure 3.7. The Moderating Effect of Mind Wandering (IT) on the Relationship between Focus of 
Attention (Non‐IT) and Software Self‐Efficacy 
   

Hypothesis 7e suggested that the positive effect of focus of attention (IT) on software 

self‐efficacy was higher when users experienced high mind wandering (IT) and low mind 

wandering (non‐IT). In regression, the three‐way interaction term of mind wandering (IT), mind 

wandering (non‐IT), and focus of attention (IT) was negative and significant (b = ‐0.18, p<.05). 

The interaction effect was plotted in Figure 3.8; the results confirmed my hypothesis. When 

mind wandering (IT) was high and mind wandering (non‐IT) was low, the effect of focus of 

attention (IT) on software self‐efficacy was higher than when mind wandering (IT) and mind 

wandering (non‐IT) were low. Thus, H7e was supported. Next, I plot the interaction effect of 

mind wandering (IT), mind wandering (non‐IT), and focus of attention (non‐IT) on software self‐

efficacy. Hypothesis 7f proposes that the negative effect of focus of attention (non‐IT) on 

software self‐efficacy is stronger when users experience high mind wandering (non‐IT) and low 

mind wandering (IT). As shown in Figure 3.9, the slope for high mind wandering (IT) and low 

99 
 
mind wandering (non‐IT) was negative and significant (b = ‐0.67; p<.001). The coefficient for low 

mind wandering (IT)‐low mind wandering (non‐IT) was also significant (b = ‐0.17, p<.05). 

However the interaction effect was not significant for low mind wandering (IT)‐high mind 

wandering (non‐IT) and for high mind wandering (IT)‐high mind wandering (non‐IT) conditions. 

Figure 3.8. The Moderating Effect of Mind Wandering (IT) and Mind Wandering (Non‐IT) on the 
Relationship between Focus of Attention (IT) and Software Self‐Efficacy 
 

Figure 3.9. The Moderating Effect of Mind Wandering (IT) and Mind Wandering (Non‐IT) on the 
Relationship between Focus of Attention (Non‐IT) and Software Self‐Efficacy 
 

100 
 
In terms of the determinants of learning performance, both focus of attention (IT) and 

focus of attention (non‐IT) had not significant effect on learning performance. Thus, H8 was not 

supported. Mind wandering (non‐IT) had a negative and significant effect on learning 

performance, whereas the effect of mind wandering (IT) was not significant, validating H9. As 

expected, both software self‐efficacy and EFs had positive and significant effects on learning 

performance. Thus, H10 and H11 were supported.  The summary of all hypotheses tested in this 

study is presented in Table 3.8. 

Table 3.8 

Hypotheses Summary 

  Hypothesis  Supported? 
H1a  EFs are negatively associated with mind wandering (IT).  Yes 
H1b  EFs are negatively associated with mind wandering (Non‐IT).  Yes 
H2a  EFs are positively associated with focus of attention (IT).  Yes 
H2b  EFs are positively associated with focus of attention (Non‐IT).  No 
H3a  Perceived task complexity is negatively associated with mind wandering (IT).  No (opposite 
direction) 
H3b  Perceived task complexity is negatively associated with mind wandering (Non‐ No (opposite 
IT).  direction) 
H4a  Perceived task complexity is positively associated with focus of attention (IT).  No 
H4b  Perceived task complexity is positively associated with focus of attention (Non‐ No 
IT). 
H5a  Focus of attention (IT) is positively associated with software self‐efficacy.  Yes 
H5b  Focus of attention (non‐IT) is negatively associated with software self‐efficacy.  Yes 
H6  The negative association between mind wandering (non‐IT) and software self‐ Yes 
efficacy is stronger than between mind wandering (IT) and software self‐
efficacy. 
H7a  The positive effect of focus of attention (IT) on software‐specific self‐efficacy is  Yes 
higher when users experience high mind wandering (IT) than when they 
experience low mind wandering (IT). 
H7b  The positive effect of focus of attention (IT) on software self‐efficacy is higher  Yes 
when users experience low mind wandering (non‐IT) than when they 
experience high mind wandering (non‐IT). 
(table continues) 

101 
 
Table 3.8 (continued) 

  Hypothesis  Supported? 
H7c  The negative effect of focus of attention (non‐IT) on software self‐efficacy is  Yes 
stronger when users experience low mind wandering (non‐IT) than when they 
experience high mind wandering (non‐IT). 
H7d  The negative effect of focus of attention (non‐IT) on software self‐efficacy is  No 
stronger when user experience high mind wandering (IT) than when they 
experience low mind wandering (IT). 
H7e  The positive effect of focus of attention (IT) on software self‐efficacy is higher when  Yes 
users experience high mind wandering (IT) and low mind wandering (non‐IT). 
H7f  The negative effect of focus of attention (non‐IT) on software self‐efficacy is higher  Yes 
when users experience high mind wandering (IT) and low mind wandering (non‐IT). 
H8  The positive effect of focus of attention (IT) on learning performance is higher than  No 
focus of attention (non‐IT). 
H9  The negative effect of mind wandering (non‐IT) on IS learning performance is higher  Yes 
than the negative effect of mind wandering (IT) on IS learning performance. 
H10  Software self‐efficacy is positively associated with task performance.  Yes 
H11  EFs are positively associated with learning performance.  Yes 
 

3.8.3  Post Hoc Analysis: Moderated Mediation Model  

Edward and Lambert (2007) recommend against a simple mediation or moderation 

approach when both effects are considered in the same research model. Since software self‐

efficacy is influenced by focus of attention, mind wandering, and their interactions, software 

self‐efficacy potentially mediates the relationship between focus of attention, mind wandering, 

and learning performance. Thus, although the direct effect of focus of attention (IT) and focus 

of attention (non‐IT) were not significant, a formal test of moderated mediation based on the 

quantification of the relationship between the proposed moderators and the size of the indirect 

effect is required to determine whether the indirect effect depends on the moderators (Hayes, 

2013).  

102 
 
In order to test the indirect effect of focus of attention mediated by software self‐

efficacy on learning performance, I ran a conditional process model using PROCESS (Hayes, 

2013). The indirect effects of focus of attention at different levels of mind wandering are 

presented in Table 3.9 and 3.10. As shown in Table 3.9, the indirect effect of focus of attention 

(IT) on learning performance was significant at almost all levels of mind wandering. When mind 

wandering (non‐IT) was high and mind wandering (IT) was low, however, the indirect effect was 

not significant. As shown in Table 3.10, the indirect effect of focus of attention (non‐IT) on 

learning performance was only significant when mind wandering (non‐IT) was low.  

Table 3.9 
Direct and Indirect Effect of Focus of Attention (IT) Mediated by SSE on Learning Performance 

MW (Non‐IT)  MW (IT)  Direct Effect  95% CI  Indirect Effect  95% CI  Sig? 


High  High  .11  ‐.02, .24  .07  .01, .15  Y 
High  Low  .11  ‐.02, .24  .01  ‐.03, .08  N 
Low  High  .11  ‐.02, .24  .12  .03, .24  Y 
Low  Low  .11  ‐.02, .24  .07  .02, .15  Y 
Note: Number of bootstrap samples for bias corrected confidence intervals is 5000. 

Table 3.10 

Direct and Indirect Effect of Focus of Attention (Non‐IT) Mediated by SSE on Learning 
Performance 
MW (Non‐IT)  MW (IT)  Direct Effect  95% CI  Indirect Effect  95% CI  Sig? 
High  High  ‐.08  ‐.20,.04  ‐.02  ‐.07, .02  N 
High  Low  ‐.08  ‐.20,.04  .02  ‐.03, .08  N 
Low  High  ‐.08  ‐.20,.04  ‐.08  ‐.15, ‐.03  Y 
Low  Low  ‐.08  ‐.20,.04  ‐.04  ‐.09, ‐.01  Y 
Note: Number of bootstrap samples for bias corrected confidence intervals is 5000. 

 
3.9  Discussion 

  This study sought to address three research questions: (1) What are the factors that 

influence users’ mind wandering activities and focus of attention (at the early stage of IS 

103 
 
learning phase)?; (2) How do mind wandering and focus of attention interact with each other to 

influence technology learners’ cognitive learning outcomes?; and (3) Do mind wandering, focus 

of attention, and software self‐efficacy influence IS learning performance? To address these 

research questions, I draw upon executive‐control failure theory of mind wandering to develop 

a research framework that explains the relationship between the antecedents of mind 

wandering and focus of attention and their cognitive outcomes in the IS learning context. 

Results show that EFs are the main determinant of both focus of attention and mind 

wandering. The interaction effects between mind wandering and focus of attention are 

dependent on the content of users’ thought. Specifically, the findings indicate that (1) the 

positive effect of focus of attention (IT) on software self‐efficacy is higher at high levels of mind 

wandering (IT); (2) the positive effect of focus of attention (IT) on software self‐efficacy is 

higher at low levels of mind wandering (non‐IT); and (3) the negative effect of focus of attention 

(non‐IT) on software self‐efficacy is stronger at low levels of mind wandering (non‐IT). I found 

that mind wandering (non‐IT) had a negative and significant effect on learning performance, 

whereas EFs and software self‐efficacy had a positive and significant effect on IS learning 

performance.  

3.9.1  Theoretical Implications 

  This work offers several theoretical contributions. First, I present a research framework 

that integrates mind wandering theory with IS learning outcomes. The scope of this study is the 

early phase of learning that has been suggested to be the most critical phase of technology 

adoption (Yi & Davis, 2003; Yi & Hwang, 2003). Although prior research has investigated the 

104 
 
negative effect of distraction on task performance, there has been limited understanding of the 

influence of internal distraction (i.e., mind wandering) on learning outcomes. The findings 

reveal that mind wandering and focus of attention are the main predictors of software self‐

efficacy. These findings not only offer insights on the difference between mind wandering and 

focus of attention, but also shed light on the determinants and outcomes of both attentional 

states. Thus, by identifying and conceptualizing the relationships between EFs, mind wandering, 

focus of attention, and IS learning outcomes, this work complements the previous studies (e.g., 

Nelson & Cheney, 1987; Sun, 2012; Piccoli, Ahmad, & Ives, 2001; Yi & Davis, 2003; Yi & Huang, 

2003) that had studied the processes and outcomes of IS learning and/or training. Further, this 

work is a response to Smallwood et al.’s (2007) and Smallwood and Schooler’s (2015) call for a 

study on the functional benefits of mind wandering, especially in a learning context. Although 

previous studies (e.g., Allen et al., 2013; Hollis & Was, 2014; Oettingen & Schworer, 2013), 

including my study, had revealed some potential costs and benefits of mind wandering, there is 

still a need for extensively investigating the impacts of different contents of mind wandering on 

technology use outcomes. Such a study will be important in increasing our understanding on 

the role and functions of mind wandering in a technological environment.  

  Second, by testing and validating the influence of EFs on mind wandering and focus of 

attention, this study aims to bring the importance of attentional process mechanisms to the 

forefront of cognitive‐IS research. By identifying the key antecedent sources of mind wandering 

and focus of attention, it aims to not only test the theory of mind wandering (i.e., executive 

control failure theory), but also to guide future research on how to approach and apply a novel 

issue to IS research. One surprising finding I found in this current study is that the results show 

105 
 
perceived task complexity is positively associated with mind wandering. Although previous 

studies (e.g., McVay & Kane, 2009; 2010; Smallwood & Schooler, 2006) argued that task 

complexity is negatively associated with mind wandering, it is possible that in an actual learning 

environment, learners look for heuristics to reduce task demands (Payne, Bettman, & Johnson, 

1993). Thus, when the complexity is high, learners tend to minimize their cognitive effort by 

letting their mind to wander. Mind wandering has been shown to be useful in dealing with 

complex tasks by providing mental breaks (Smallwood & Schooler, 2015). Thus, it is likely that 

perceived task complexity increases the episodes of mind wandering by providing a break time 

to relieve the tension of cognitive demands.  

  Third, this work extends research in mind wandering by offering determinants of 

software self‐efficacy and learning performance. This work is the first study revealing the 

interaction effects of different contents of thought. Although previous studies have emphasized 

the importance of maintaining balance between focus of attention and mind wandering (e.g., 

Allen et al., 2013; Dijksterhuis & Aarts, 2010), there has been limited research, if any, that 

identified the interaction effects of the relationship between focus of attention and mind 

wandering. The current findings suggest that the costs and benefits of mind wandering depend, 

in part, upon its content. These findings also confirm the content regulation hypothesis of mind 

wandering (Andrews‐Hanna et al., 2013) and diversion thought paradigm (Delaney et al., 2010). 

Specifically, the results reveal that the congruence between the content of mind wandering and 

focus of attention determines the outcomes of the interaction between these two constructs. 

These findings offer important insights on how technology learners and/or users regulate their 

106 
 
content of thoughts. These findings also contribute to the technology design literature by 

highlighting the important aspects of mind wandering processes.  

  Fourth, I found that whereas mind wandering (non‐IT) has a negative and significant 

effect on learning performance, mind wandering (non‐IT), focus of attention (IT), and focus of 

attention (non‐IT) do not have significant effects on learning performance. The absence of 

significant effects of focus of attention on learning performance was surprising in light of the 

importance of this variable in learning (Kanfer & Ackerman, 1989). It is possible that its effect is 

attenuated by EFs and software self‐efficacy. This argument is also supported by the significant 

indirect effect of focus of attention on learning performance, mediated by software self‐

efficacy. The significant effect of EFs on learning performance suggests that at the early stage of 

technology learning, users who have higher EFs most often perform better than users with 

lower EFs (Diamond, 2013). Learning at this stage is a conscious experience (i.e., top‐down 

control) and consumes executive resources. If users fail to regulate their thought, then their 

learning performance will decrease (Kanfer & Ackerman, 1989; Smallwood et al., 2007a). Thus, 

early in learning, mind wandering can disrupt EFs functions, leading to performance impairment 

(Smallwood et al., 2007a).     

3.9.2  Limitations and Future Work 

  The findings of this current study should be interpreted in light of the limitations of this 

work. First, data were drawn from undergraduate students. Although this sample is relevant to 

the context of my study, future research can take this investigation further by drawing research 

subjects from a more diverse population, such as IT practitioners and knowledge workers. 

107 
 
Second, this study exclusively focuses on two dimensions of mind wandering—technology‐

related (i.e., IT) and nontechnology‐related (i.e., non‐IT). Future research should empirically 

measure different contents of thought in more details. Further, I only tested the moderating 

effect of mind wandering. Future research can investigate the effect of other potential 

moderators, such as self‐regulatory activities (Kanfer & Ackerman, 1989), working memory 

capacity (Kane et al., 2007), and task complexity (Randall et al., 2014). Fourth, although the 

validity of subjective self‐reports of mind wandering has been demonstrated in both 

experimental studies and everyday life experiences (e.g., McVay et al., 2009; Smallwood et al., 

2004), our understanding about mind wandering during users’ interaction with technology can 

be enhanced by utilizing more advanced neuropsychological tools (e.g., fMRI, eye‐tracking 

devices). Further, the context of this study is limited to a novel task among students (i.e., MS. 

Access learning). Future research should investigate the phenomenon of mind wandering under 

different types of tasks (e.g., simple tasks, problem tasks, decision tasks, and judgment tasks). 

3.9.3  Practical Contributions  

  The results of this study offer important practical implications for technology users and 

IS practitioners. My findings provide evidence of the significant moderating effect of mind 

wandering on software self‐efficacy. However, these interaction effects are depending on the 

thought content of mind wandering, whether it’s technology‐related or nontechnology‐related. 

Software self‐efficacy has been considered one of the most important factors in the technology 

learning and adoption literature (e.g., Agarwal et al., 2000; Compeau & Higgins, 1995; Marakas 

et al., 1998). The application of mind wandering, however, has not been introduced in the IS 

108 
 
field. The results of this study demonstrate that the costs and benefits of mind wandering 

depend, in part, on the content of such experience. Maintaining the congruency between the 

content of mind wandering and focus of attention contributes to the learning outcomes. In 

addition, mind wandering will only contribute to software self‐efficacy under the condition that 

the task‐relevant cues have been primarily activated. Applying these findings could, for 

example, motivate people to be aware of their mind wandering when they learn a new 

technology or application. 

  Further, given that we spend a lot of time engaging in mind wandering, I suggest 

business and IT managers to explore the beneficial aspects of this experience. When users learn 

how to use a system, mind wandering can be a major detriment to cognitive performance. 

However, mind wandering can also enhance software self‐efficacy with the presence of focus of 

attention. Drawing upon these results, I propose two practical challenges for technology users: 

how to optimally balancing mind wandering and focus of attention and how to noticing and 

correcting mind wandering when it occurs. These can be done by recognizing the content of our 

attention and determining whether a particular content is inconsistent with the intention of 

completing the task (Smallwood et al., 2008). These results can also be applied in the health, 

education, sport, and other domains, across all ages and cultures. Further, I also encourage 

technology users to engage in mind wandering activities when they work on relatively complex 

tasks. It can be achieved by simplifying or automatizing the tasks, while simultaneously 

providing time for breaks (Randall et al., 2014).   

  Given mind wandering can also be disruptive in nature, it is necessary to identify 

strategies that can be used to minimize its negative consequences. For example, business and 

109 
 
technology managers can encourage individuals to be mindful of their thought contents 

(Smallwood & Schooler, 2015). Meta‐awareness trainings may help individuals to curtail 

episodes of mind wandering. Offering mindfulness programs to IT or knowledge workers may 

encourage them to routinely take notice of the content of their thoughts and only focus on 

subject matters relevant to their current task context. Although more research is needed in this 

area, encouraging people to be aware of their own mind wandering experience may help to 

minimize its disruptive effects (Smallwood & Schooler, 2015). 

3.10  Conclusions 

  This current study extends our understanding on the role of mind wandering in a 

technological setting. To the best of my knowledge, this study is the first to demonstrate the 

interaction effects between mind wandering and focus of attention on software self‐efficacy. I 

specifically focus on the content of thought during mind wandering by differentiating the 

thought into two categories: IT and non‐IT. My findings suggest that the congruence between 

the content of mind wandering and focus of attention determines the costs and benefits of 

such experiences. The findings also demonstrate that EFs, mind wandering (IT), and software 

self‐efficacy are the main predictors of IS learning performance. In sum, this study advances our 

understanding on the mind wandering, particularly in the context of our interaction with IS.   

 
3.11  References 

Ackerman, P. L. (1988). Determinants of individual differences during skill acquisition: Cognitive 
abilities and information processing. Journal of Experimental Psychology: General, 117(3), 
288‐318. 

110 
 
Agarwal, R., Sambamurthy, V., & Stair, R. M. (2000). The evolving relationship between general 
and specific computer self‐efficacy: An empirical assessment. Information Systems 
Research, 11(4), 418‐430. 
Allen, M., Smallwood, J., Christensen, J., Gramm, D., Rasmussen, B., Jensen, C. G., Roepstorff, 
A., & Lutz, A. (2013). The balanced mind: The variability of task‐unrelated thoughts predicts 
error monitoring. Frontiers in Human Neuroscience, 7, 1‐15. 
Alves, F., Resende, F., & Pinho, M. S. (2015). Memory impairment in older adults’ diversionary 
thoughts. Frontiers in Aging Neuroscience, 7, 1‐5. 
Andrews‐Hanna, J. R., Smallwood, J., & Spreng, R. N. (2014). The default network and self‐
generated thought: Component processes, dynamic control, and clinical relevance. Annals 
of the New York Academy of Sciences, 1316(1), 29‐52. 
Axelrod, V., Rees, G., Lavidor, M., & Bar, M. 2015. Increasing propensity to mind‐wander with 
transcranial direct current stimulation. Proceedings of the National Academy of Sciences, 
112(11), 3314‐3319. 
Baddeley, A. (2000). The episodic buffer: A new component of working memory? Trends in 
Cognitive Sciences, 4(11), 417‐423. 
Baddeley, A. (2003). Working memory: Looking back and looking forward. Nature Reviews 
Neuroscience, 4(10), 829‐839. 
Baddeley, A. (2007). Working memory, thought, and action. New York, NY: Oxford University 
Press. 
Baddeley, A. (2012). Working memory: Theories, models, and controversies. Annual Review of 
Psychology, 63, 1‐29. 
Baird, B., Smallwood, J., Mrazek, M. D., Kam, J. W., Franklin, M. S., & Schooler, J. W. (2012). 
Inspired by distraction mind wandering facilitates creative incubation. Psychological Science, 
23(10), 1117‐1122. 
Bala, H., & Venkatesh, V. (2013). Changes in employees' job characteristics during an enterprise 
system implementation: A latent growth modeling perspective. MIS Quarterly, 37(4), 1113‐
1140. 
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. 
Englewood Cliffs, NJ: Prentice‐Hall. 
Bandura, A. (1991). Social cognitive theory of self‐regulation. Organizational Behavior and 
Human Decision Processes, 50(2), 248‐287. 
Bandura, A. (1993). Perceived self‐efficacy in cognitive development and functioning. 
Educational Psychologist, 28(2), 117‐148. 
Barron, E., Riby, L. M., Greer, J., & Smallwood, J. (2011). Absorbed in thought: The effect of 
mind wandering on the processing of relevant and irrelevant events. Psychological Science, 
22(5), 596‐601. 

111 
 
Borella, E., Carretti, B., Riboldi, F., & De Beni, R. (2010). Working memory training in older 
adults: evidence of transfer and maintenance effects. Psychology and Aging, 25(4), 767‐778. 
Burton‐Jones, A., & Straub, D. W. (2006). Reconceptualizing system usage: An approach and 
empirical test. Information Systems Research, 17(3), 228‐246. 
Busch, T. (1995). Gender differences in self‐efficacy and attitudes toward computers. Journal of 
Educational Computing Research, 12(2), 147‐158. 
Chin, W. W. (1998). The partial least squares approach to structural equation modeling. Modern 
Methods for Business Research, 295(2), 295‐336. 
Chin, W. W., Marcolin, B. L., & Newsted, P. R. (2003). A partial least squares latent variable 
modeling approach for measuring interaction effects: Results from a Monte Carlo 
simulation study and an electronic‐mail emotion/adoption study. Information Systems 
Research, 14(2), 189‐217. 
Clark, J. M. (1996). Contributions of inhibitory mechanisms to unified theory in neuroscience 
and psychology. Brain and Cognition, 30(1), 127‐152. 
Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Erlbaum. 
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2013). Applied multiple regression/correlation 
analysis for the behavioral sciences. Mahwah, New Jersey: Lawrence Erlbaum. 
Compeau, D., Higgins, C. A., & Huff, S. (1999). Social cognitive theory and individual reactions to 
computing technology: A longitudinal study. MIS Quarterly, 23(2), 145‐158. 
Compeau, D., Marcolin, B., Kelley, H., & Higgins, C. (2012). Generalizability of information 
systems research using student subjects‐a reflection on our practices and recommendations 
for future research. Information Systems Research, 23(4), 1093‐1109. 
Delaney, P. F., Sahakyan, L., Kelley, C. M., & Zimmerman, C. A. (2010). Remembering to forget 
the amnesic effect of daydreaming. Psychological Science, 21(7), 1036‐1042. 
DeLone, W. H., & McLean, E. R. (1992). Information systems success: The quest for the 
dependent variable. Information Systems Research, 3(1), 60‐95. 
DeLone, W. H., & McLean, E. R. (2003). The DeLone and McLean model of information systems 
success: A ten‐year update. Journal of Management Information Systems, 19(4), 9‐30. 
Diamond, A. (2013). Executive Functions. Annual Review of Psychology, 64, 135‐168. 
Dijksterhuis, A., & Aarts, H. (2010). Goals, attention, and (un)consciousness. Annual Review of 
Psychology, 61, 467‐490. 
Duncan, G. J., Dowsett, C. J., Claessens, A., Magnuson, K., Huston, A. C., Klebanov, P., ... & 
Sexton, H. (2007). School readiness and later achievement. Developmental Psychology, 
43(6), 1428‐1446. 

112 
 
Edwards, J. R., & Lambert, L. S. (2007). Methods for integrating moderation and mediation: A 
general analytical framework using moderated path analysis. Psychological Methods, 12(1), 
1‐22. 
Engle, R. W., & Kane, M. J. (2004). Executive attention, working memory capacity, and a two‐
factor theory of cognitive control. In B. H. Ross (Ed.), Psychology of learning and motivation 
(Vol. 44, pp. 145‐200), San Diego, CA: Elsavier Academic Press. 
Engle, R. W., Tuholski, S. W., Laughlin, J. E., & Conway, A. R. (1999). Working memory, short‐
term memory, and general fluid intelligence: a latent‐variable approach. Journal of 
Experimental Psychology: General, 128(3), 309‐331. 
Eysenck, M. W., Derakshan, N., Santos, R., & Calvo, M. G. (2007). Anxiety and cognitive 
performance: Attentional control theory. Emotion, 7(2), 336‐353. 
Fornell, C., & Bookstein, F. L. (1982). Two structural equation models: LISREL and PLS applied to 
consumer exit‐voice theory. Journal of Marketing Research, 19(4), 440‐452. 
Fornell, C., & Larcker, D. F. (1981). Evaluating structural equations models with unobservable 
variables and measurement error. Journal of Marketing Research, 18(1), 39‐50. 
Foulsham, T., Farley, J., & Kingstone, A. (2013). Mind wandering in sentence reading: 
Decoupling the link between mind and eye. Canadian Journal of Experimental 
Psychology/Revue Canadienne de Psychologie Expérimentale, 67(1), 51‐59. 
Franklin, M. S., Smallwood, J., & Schooler, J. W. (2011). Catching the mind in flight: Using 
behavioral indices to detect mindless reading in real time. Psychonomic Bulletin & 
Review, 18(5), 992‐997. 
Fredricks, J. A., Blumenfeld, P. C., & Paris, A. H. (2004). School engagement: Potential of the 
concept, state of the evidence. Review of Educational Research, 74(1), 59‐109. 
Friedman, N. P., & Miyake, A. (2004). The relations among inhibition and interference control 
functions: A latent‐variable analysis. Journal of Experimental Psychology: General, 133(1), 
101‐135. 
Friedman, N. P., Miyake, A., Young, S. E., DeFries, J. C., Corley, R. P., & Hewitt, J. K. (2008). 
Individual differences in executive functions are almost entirely genetic in origin. Journal of 
Experimental Psychology: General, 137(2), 201‐225. 
Fuller, R. M., & Dennis, A. R. (2009). Does fit matter? The impact of task‐technology fit and 
appropriation on team performance in repeated tasks. Information Systems Research, 20(1), 
2‐17. 
Gathercole, S. E., Pickering, S. J., Knight, C., & Stegmann, Z. (2004). Working memory skills and 
educational attainment: Evidence from national curriculum assessments at 7 and 14 years 
of age. Applied Cognitive Psychology, 18(1), 1‐16. 
Giambra, L. M. (1993). The influence of aging on spontaneous shifts of attention from external 
stimuli to the contents of consciousness. Experimental Gerontology, 28(4), 485‐492. 

113 
 
Gilbert, D. T., & Wilson, T. D. (2007). Prospection: Experiencing the future. Science, 317(5843), 
1351‐1354. 
Gist, M. E., & Mitchell, T. R. (1992). Self‐efficacy: A theoretical analysis of its determinants and 
malleability. Academy of Management Review, 17(2), 183‐211. 
Gluck, J., Bunt, A., & McGrenere, J. (2007). Matching attentional draw with utility in 
interruption. In Proceedings of the SIGCHI Conference on Human Factors in Computing 
Systems, 41‐50.  
Gravill, J., & Compeau, D. (2008). Self‐regulated learning strategies and software training. 
Information & Management, 45(5), 288‐296. 
Gruberger, M., Simon, E. B., Levkovitz, Y., Zangen, A., & Hendler, T. (2011). Towards a 
neuroscience of mind‐wandering. Frontiers in Human Neuroscience, 5(56), 1‐11. 
Gupta, A., Li, H., & Sharda, R. (2013). Should I send this message? Understanding the impact of 
interruptions, social hierarchy and perceived task complexity on user performance and 
perceived workload. Decision Support Systems, 55(1), 135‐145. 
Hair, J. F., Ringle, C. M., & Sarstedt, M. (2011). PLS‐SEM: Indeed a silver bullet. Journal of 
Marketing Theory and Practice, 19(2), 139‐152. 
Harnishfeger, K. K. (1995). The development of cognitive inhibition. In F. N. Dempster & C. J. 
Brainerd (Eds.), Interference and inhibition in cognition (pp. 175‐204), San Diego, CA: 
Academic Press. 
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A 
regression‐based approach. New York, London: The Guilford Press.  
Hollis, R. B., & Was, C. A. (2010). Mind wandering and online learning: How working memory, 
interest, and mind wandering impact learning from videos. In Proceedings of the 36th 
Annual Meeting of the Cognitive Science Society, Quebec City, Canada, Jul 23‐26. 
Hughes, M. G., Day, E. A., Wang, X., Schuelke, M. J., Arsenault, M. L., Harkrider, L. N., & Cooper, 
O. D. (2013). Learner‐controlled practice difficulty in the training of a complex task: 
Cognitive and motivational mechanisms. Journal of Applied Psychology, 98(1), 80‐98. 
Inquisit (Version 4.0.3) [Computer software]. (2014). Seattle, WA: Millisecond Software LLC. 
Jaeggi, S. M., Studer‐Luethi, B., Buschkuehl, M., Su, Y. F., Jonides, J., & Perrig, W. J. (2010). The 
relationship between N‐back performance and matrix reasoning‐implications for training 
and transfer. Intelligence, 38(6), 625‐635. 
Kane, M. J., Brown, L. H., McVay, J. C., Silvia, P. J., Myin‐Germeys, I., & Kwapil, T. R. (2007). For 
whom the mind wanders, and when: An experience‐sampling study of working memory and 
executive control in daily life. Psychological Science, 18, 614‐621. 
Kane, M. J., & Engle, R. W. (2003). Working‐memory capacity and the control of attention: The 
contributions of goal neglect, response competition, and task set to Stroop interference. 
Journal of Experimental Psychology: General, 132(1), 47‐70. 

114 
 
Kane, M. J., & McVay, J. C. (2012). What mind wandering reveals about executive‐control 
abilities and failures. Current Directions in Psychological Science, 21(5), 348‐354. 
Kanfer, R., & Ackerman, P. L. (1989). Motivation and cognitive abilities: An integrative/aptitude‐
treatment interaction approach to skill acquisition. Journal of Applied Psychology, 74(4), 
657‐690. 
Kanfer, R., Ackerman, P. L., Murtha, T. C., Dugdale, B., & Nelson, L. (1994). Goal setting, 
conditions of practice, and task performance: A resource allocation perspective. Journal of 
Applied Psychology, 79(6), 826‐835. 
Kaufman, S. B. (2013). Ungifted: Intelligence redefined. New York, NY: Basic Books.  
Kaufman, S. B., DeYoung, C. G., Gray, J. R., Jiménez, L., Brown, J., & Mackintosh, N. (2010). 
Implicit learning as an ability. Cognition, 116(3), 321‐340. 
Kenny, D. A. (2015). Moderator variables: Introduction. Retrieved from website: 
http://davidakenny.net/cm/moderation.htm. 
Lindell, M. K., & Whitney, D. J. (2001). Accounting for common method variance in cross‐
sectional research designs. Journal of Applied Psychology, 86(1), 114‐121. 
Lindquist, S. I., & McLean, J. P. (2011). Daydreaming and its correlates in an educational 
environment. Learning and Individual Differences, 21(2), 158‐167. 
Locke, E. A., Frederick, E., Lee, C., & Bobko, P. (1984). Effect of self‐efficacy, goals, and task 
strategies on task performance. Journal of Applied Psychology, 69(2), 241‐251. 
Logan, G. D. (1994). On the ability to inhibit thought and action: A user's guide to the stop signal 
paradigm. In D. Dagenbach & T. H. Carr (Eds.), Inhibitory Processes in Attention, Memory, 
and Language (pp. 189‐239). San Diego, CA: Academic Press. 
MacLeod, C. M. (1991). Half a century of research on the Stroop effect: an integrative review. 
Psychological Bulletin, 109(2), 163‐203. 
Marakas, G. M., Yi, M. Y., & Johnson, R. D. (1998). The multilevel and multifaceted character of 
computer self‐efficacy: Toward clarification of the construct and an integrative framework 
for research. Information Systems Research, 9(2), 126‐163. 
Marcolin, B. L., Compeau, D. R., Munro, M. C., & Huff, S. L. (2000). Assessing user competence: 
Conceptualization and measurement. Information Systems Research, 11(1), 37‐60. 
Margaryan, A., Littlejohn, A., & Vojt, G. (2011). Are digital natives a myth or reality? University 
students' use of digital technologies. Computers & Education, 56(2), 429‐440. 
McElree, B. (2001). Working memory and focal attention. Journal of Experimental Psychology: 
Learning, Memory, and Cognition, 27(3), 817‐835. 
McMillan, R. L., Kaufman, S. B., & Singer, J. L. (2013). Ode to positive constructive daydreaming. 
Frontiers in Psychology, 4, 1‐9. 

115 
 
McVay, J. C., & Kane, M. J. (2009). Conducting the train of thought: Working memory capacity, 
goal neglect, and mind wandering in an executive‐control task. Journal of Experimental 
Psychology: Learning, Memory, and Cognition, 35(1), 196‐204. 
McVay, J. C., & Kane, M. J. (2010). Does mind wandering reflect executive function or executive 
failure? Comment on Smallwood and Schooler (2006) and Watkins (2008). Psychological 
Bulletin, 136, 188‐197. 
McVay, J. C., & Kane, M. J. (2012a). Why does working memory capacity predict variation in 
reading comprehension? On the influence of mind wandering and executive attention. 
Journal of Experimental Psychology: General, 141(2), 302‐320. 
McVay, J. C., & Kane, M. J. (2012b). Drifting from slow to "d'oh!": Working memory capacity and 
mind wandering predict extreme reaction times and executive control errors. Journal of 
Experimental Psychology: Learning, Memory, and Cognition, 38(3), 525‐549. 
McVay, J. C., Kane, M. J., & Kwapil, T. R. (2009). Tracking the train of thought from the 
laboratory into everyday life: An experience‐sampling study of mind wandering across 
controlled and ecological contexts. Psychonomic Bulletin & Review, 16(5), 857‐863. 
McVay, J. C., Meier, M. E., Touron, D. R., & Kane, M. J. (2013). Aging ebbs the flow of thought: 
Adult age differences in mind wandering, executive control, and self‐evaluation. Acta 
Psychologica, 142(1), 136‐147. 
Milham, M. P., Erickson, K. I., Banich, M. T., Kramer, A. F., Webb, A., Wszalek, T., & Cohen, N. J. 
(2002). Attentional control in the aging brain: insights from an fMRI study of the stroop task. 
Brain and cognition, 49(3), 277‐296. 
Miyake, A., Emerson, M. J., Padilla, F., & Ahn, J. C. (2004). Inner speech as a retrieval aid for task 
goals: The effects of cue type and articulatory suppression in the random task cuing 
paradigm. Acta Psychologica, 115(2), 123‐142. 
Miyake, A., & Friedman, N. P. (2012). The nature and organization of individual differences in 
executive functions four general conclusions. Current Directions in Psychological Science, 
21(1), 8‐14. 
Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., Howerter, A., & Wager, T. D. (2000). 
The unity and diversity of executive functions and their contributions to complex "frontal 
lobe" tasks: A latent variable analysis. Cognitive Psychology, 41(1), 49‐100. 
Monsell, S. (2003). Task switching. Trends in Cognitive Sciences, 7(3), 134‐140. 
Mooneyham, B. W., & Schooler, J. W. (2013). The costs and benefits of mind wandering: A 
review. Canadian Journal of Experimental Psychology, 67(1), 11‐18. 
Mrazek, M. D., Phillips, D. T., Franklin, M. S., Broadway, J. M., & Schooler, J. W. (2013). Young 
and restless: Validation of the mind wandering questionnaire (MWQ) reveals disruptive 
impact of mind wandering for youth. Frontiers in Psychology, 4, 1‐7. 

116 
 
Mrazek, M. D., Smallwood, J., Franklin, M. S., Chin, J. M., Baird, B., & Schooler, J. W. (2012). The 
role of mind wandering in measurements of general aptitude. Journal of Experimental 
Psychology: General, 141(4), 788‐798. 
Nelson, R. R., & Cheney, P. H. (1987). Training end users: An exploratory study. MIS Quarterly, 
11(4), 547‐559. 
Nigg, J. T. (2000). On inhibition/disinhibition in developmental psychopathology: Views from 
cognitive and personality psychology and a working inhibition taxonomy. Psychological 
Bulletin, 126(2), 220‐246. 
Oettingen, G., & Schwörer, B. (2013). Mind wandering via mental contrasting as a tool for 
behavior change. Frontiers in Psychology, 4, 1‐5. 
Ophir, E., Nass, C., & Wagner, A. D. (2009). Cognitive control in media multitaskers. Proceedings 
of the National Academy of Sciences of the United States of America, 106(37), 15583‐15587. 
Palvia, S. C. J., & Palvia, P. C. (2007). The effectiveness of using computers for software training: 
An exploratory study. Journal of Information Systems Education, 18(4), 479‐489. 
Pavlou, P. A., Liang, H., & Xue, Y. (2006). Understanding and mitigating uncertainty in online 
environments: A principal‐agent perspective. MIS Quarterly, 31(1), 105‐136. 
Payne, J. W., Bettman, J. R., & Johnson, E. J. (1993). The adaptive decision maker. Victoria, AU: 
Cambridge University Press. 
Piccoli, G., Ahmad, R., & Ives, B. (2001). Web‐based virtual learning environments: A research 
framework and a preliminary assessment of effectiveness in basic IT skills training. MIS 
Quarterly, 25(4), 401‐426. 
Podsakoff, P. M., MacKenzie, S. B., Lee, J.‐Y., & Podsakoff, N. P. (2003). Common method biases 
in behavioral research: A critical review of the literature and recommended remedies. 
Journal of Applied Psychology, 88(5), 879‐903. 
Randall, J. G., Oswald, F. L., & Beier, M. E. (2014). Mind‐wandering, cognition, and performance: 
A theory‐driven meta‐analysis of attention regulation. Psychological Bulletin, 140(6), 1411‐
1431. 
Rawlings, P., White, P., & Stephens, R. (2005). Practice‐based learning in information systems: 
The advantages for students. Journal of Information Systems Education, 16(4), 455‐464. 
Risko, E. F., Anderson, N., Sarwal, A., Engelhardt, M., & Kingstone, A. (2012). Everyday 
attention: Variation in mind wandering and memory in a lecture. Applied Cognitive 
Psychology, 26(2), 234‐242. 
Risko, E. F., Buchanan, D., Medimorec, S., & Kingstone, A. (2013). Everyday attention: Mind 
wandering and computer use during lectures. Computers & Education, 68, 275‐283. 
Ritter, S. M., & Dijksterhuis, A. (2014). Creativity: The unconscious foundation of the incubation 
period. Frontiers in Human Neuroscience, 8, 1‐10. 

117 
 
Roediger, H. L., & Karpicke, J. D. (2006). The power of testing memory: Basic research and 
implications for educational practice. Perspectives on Psychological Science, 1(3), 181‐210. 
Ruby, F. J., Smallwood, J., Engen, H., & Singer, T. (2013). How self‐generated thought shapes 
mood‐the relation between mind wandering and mood depends on the socio‐temporal 
content of thoughts. PLoS One, 8(10), 1‐7. 
Sahakyan, L., & Kelley, C. M. (2002). A contextual change account of the directed forgetting 
effect. Journal of Experimental Psychology: Learning, Memory, and Cognition, 28(6), 1064‐
1072. 
Schooler, J. W., Mrazek, M. D., Franklin, M. S., Baird, B., Mooneyham, B. W., Zedelius, C., & 
Broadway, J. M. (2014). The middle way: Finding the balance between mindfulness and 
mind wandering. Psychology of Learning and Motivation, 60, 1‐33. 
Schooler, J. W., Reichle, E. D., & Halpern D. V. (2004). Zoning out while reading: Evidence for 
dissociations between experience and metaconsciousness. In D. T. Levin (Ed.), Thinking and 
seeing: Visual metacognition in adults and children (pp. 203‐226), Cambridge, MA: MIT 
Press. 
Seijts, G. H., Latham, G. P., Tasa, K., & Latham, B. W. (2004). Goal setting and goal orientation: 
An integration of two different yet related literatures. Academy of Management Journal, 
47(2), 227‐239. 
Singer, J. L. (1966). Daydreaming: An introduction to the experimental study of inner experience. 
New York, NY: Random House. 
Skinner, E. A., Furrer, C., Marchand, G., & Kindermann, T. (2008). Engagement and disaffection 
in the classroom: Part of a larger motivational dynamic? Journal of Educational Psychology, 
100, 765‐781. 
Skinner, E. A., Kindermann, T. A., & Furrer, C. J. (2009). A motivational perspective on 
engagement and disaffection: Conceptualization and assessment of children's behavioral 
and emotional participation in academic activities in the classroom. Educational and 
Psychological Measurement, 69, 493‐525. 
Smallwood, J. M. (2011). Mind‐wandering while reading: Attentional decoupling, mindless 
reading and the cascade model of inattention. Language and Linguistics Compass, 5(2), 63‐
77. 
Smallwood, J. M., & Andrews‐Hanna, J. (2013). Not all minds that wander are lost: The 
importance of a balanced perspective on the mind wandering state. Frontiers in Psychology, 
4, 1‐6. 
Smallwood, J. M., Baracaia, S. F., Lowe, M., & Obonsawin, M. (2003). Task unrelated thought 
whilst encoding information. Consciousness and Cognition, 12(3), 452‐484. 
Smallwood, J. M., Beach, E., Schooler, J. W., & Handy, T. C. (2008). Going AWOL in the brain: 
Mind wandering reduces cortical analysis of external events. Journal of Cognitive 
Neuroscience, 20(3), 458‐469. 

118 
 
Smallwood J. M., Davies J. B., Heim D., Finnigan F., Sudberry M., & O'Connor R. (2004). 
Subjective experience and the attentional lapse: Task engagement and disengagement 
during sustained attention. Consciousness and Cognition, 13, 657‐690. 
Smallwood, J. M., Fishman, D. J., & Schooler, J. W. (2007a). Counting the cost of an absent 
mind: mind wandering as an underrecognized influence on educational performance. 
Psychonomic Bulletin & Review, 14(2), 230‐236. 
Smallwood, J. M., McSpadden, M., & Schooler, J. W. (2007b). The lights are on but no one's 
home: Meta‐awareness and the decoupling of attention when the mind wanders. 
Psychonomic Bulletin & Review, 14(3), 527‐533. 
Smallwood, J. M., O'Connor, R. C., Sudbery, M. V., & Obonsawin, M. (2007c). Mind wandering 
and dysphoria. Cognition and Emotion, 21(4), 816‐842. 
Smallwood, J. M., Ruby, F. J., & Singer, T. (2013). Letting go of the present: Mind wandering is 
associated with reduced delay discounting. Consciousness and Cognition, 22(1), 1‐7. 
Smallwood, J. M., & Schooler, J. W. (2006). The restless mind. Psychological Bulletin, 132(6), 
946‐958. 
Smallwood, J. M., & Schooler, J. W. (2015). The science of mind wandering: Empirically 
navigating the stream of consciousness. Annual Review of Psychology, 66, 487‐518. 
Smallwood, J. M., Schooler, J. W., Turk, D. J., Cunningham, S. J., Burns, P., & Macrae, C. N. 
(2011). Self‐reflection and the temporal focus of the wandering mind. Consciousness and 
Cognition, 20(4), 1120‐1126. 
St Clair‐Thompson, H. L., & Gathercole, S. E. (2006). Executive functions and achievements in 
school: Shifting, updating, inhibition, and working memory. The Quarterly Journal of 
Experimental Psychology, 59(4), 745‐759. 
Stawarczyk, D., Majerus, S., Catale, C., & D'Argembeau, A. (2014). Relationships between mind 
wandering and attentional control abilities in young adults and adolescents. Acta 
Psychologica, 148, 25‐36. 
Sullivan, Y. W., Davis, F. D., & Koh, C. (2015). Exploring mind wandering in a technological 
setting. Paper presented at the 36th International Conference on Information Systems (ICIS), 
Fort Worth, TX, December 13‐16. 
Sun, H. (2012). Understanding user revisions when using information system features: Adaptive 
system use and triggers. MIS Quarterly, 36(2), 453‐478. 
Tabachnick, B. G., & L. S. Fidell. (1996). Using multivariate statistics. New York, NY: 
HarperCollins. 
Tan, B., Yi, C., & Chan, H. C. (2015). Deliberation without attention: The latent benefits of 
distracting website features for online purchase decisions. Information Systems Research, 
26(2), 437‐455. 

119 
 
Thatcher, J. B., & Perrewe, P. L. (2002). An empirical examination of individual traits as 
antecedents to computer anxiety and computer self‐efficacy. MIS Quarterly, 26(4), 381‐396. 
Unsworth, N., & Engle, R. W. (2008). Speed and accuracy of accessing information in working 
memory: An individual differences investigation of focus switching. Journal of Experimental 
Psychology: Learning, Memory, and Cognition, 34(3), 616‐630. 
Unsworth, N., & McMillan, B. D. (2013). Mind wandering and reading comprehension: 
Examining the roles of working memory capacity, interest, motivation, and topic 
experience. Journal of Experimental Psychology: Learning, Memory, and Cognition, 39(3), 
832. 
Unsworth, N., McMillan, B. D., Brewer, G. A., & Spillers, G. J. (2012). Everyday attention failures: 
An individual differences investigation. Journal of Experimental Psychology: Learning, 
Memory, and Cognition, 38(6), 1765‐1772. 
Van der Heijden, H. (2004). User acceptance of hedonic information systems. MIS Quarterly, 
28(4), 695‐704. 
Watkins, E. R. (2008). Constructive and unconstructive repetitive thought. Psychological 
Bulletin, 134(2), 163‐206. 
Wood, R. E. (1986). Task complexity: Definition of the construct. Organizational Behavior and 
Human Decision Processes, 37(1), 60‐82. 
Xu, J. D., Benbasat, I., & Cenfetelli, R. T. (2014). The influences of online service technologies 
and task complexity on efficiency and personalization. Information Systems Research, 25(2), 
420‐436. 
Yee, P. L., Hsieh‐Yee, I., Pierce, G. R., Grome, R., & Schantz, L. (2004). Self‐evaluative intrusive 
thoughts impede successful searching on the Internet. Computers in Human Behavior, 20(1), 
85‐101. 
Yi, M. Y., & Davis, F. D. (2003). Developing and validating an observational learning model of 
computer software training and skill acquisition. Information Systems Research, 14(2), 146‐
169. 
Yi, M. Y., & Hwang, Y. (2003). Predicting the use of web‐based information systems: Self‐
efficacy, enjoyment, learning goal orientation, and the technology acceptance model. 
International Journal of Human‐Computer Studies, 59(4), 431‐449. 
Zhu, B., & Watts, S. A. (2010). Visualization of network concepts: The impact of working 
memory capacity differences. Information Systems Research, 21(2), 327‐344. 

120 
 
CHAPTER 4 

MIND WANDERING AND IS DECISION MAKING PERFORMANCE6 

4.1  Introduction 

Studies have suggested that the excessive volume of information and frequent 

interruptions of work can significantly degrade not only personal well‐being but also decision 

making, innovation, and productivity (Hemp, 2009). Given that distractions caused by 

technologies are inevitable, information systems (IS) researchers have investigated how 

distractions may decrease users’ task performance (Dabbish & Kraut, 2008). However, until 

recently, studies on technology related distractions were restricted to those caused by external 

stimuli. In daily life, our task attention may be distracted not only by external stimuli, but also 

by internally generated stimuli, known as mind wandering (Forster, 2013). Yet, our 

understanding on how mind wandering may influence our performance when we use 

technology is limited.  

Further, while researchers in the areas of IS and human‐computer interaction (HCI) have 

highlighted the negative effect of interruptions (e.g., Speier, Vessey, & Valacich, 2003), none of 

the existing studies investigated individual variability in mind wandering and focus of attention. 

In fact, the reference discipline has acknowledged that a core aspect of human cognition 

involves “overcoming the constraints of the present environment by mentally simulating 

another time, place, and perspective” (Andrews‐Hanna et al., 2013, p. 1). Although the early 

6
 Parts of this chapter are reproduced from Wati [Sullivan], Y., Koh, C., & Davis, F. D. (2014). Can you increase your 
performance  in  a  technology‐driven  society  full  of  distraction?.  Poster  presented  at  the 35th International
Conference on Information Systems, Auckland, NZ, December 14‐17. 

121 
mind wandering literature generally suggests that mind wandering is negatively associated with 

task performance (e.g., Killingsworth & Gilbert, 2010; McVay & Kane, 2010; Moberly & Watkins, 

2008; Smallwood & Schooler, 2006), recent research from the unconscious thought theory 

(UTT) perspective has shown that mind wandering facilitates future planning and productivity—

two important components for working effectively (Davis, 2015; Ruby, Smallwood, Engen, & 

Singer, 2013; Smallwood & Schooler, 2015). As digital technology has become more ubiquitous 

in many people’s daily routines (Ophir, Nass, & Wagner, 2009), an effort to understand the role 

of individuals’ focus of attention and mind wandering in this highly connected environment is 

important. This current study recognizes the importance of individual cognitive differences in 

influencing behavioral performances in users. Therefore, drawing on UTT, this research seeks to 

understand the effect of individuals’ focus of attention and mind wandering on task 

performance within the context of IS decision making (i.e., B2C e‐commerce). Two research 

questions are addressed: (1) How do the different levels of IS task complexity influence the 

relationships between focus of attention and IS decision making performance? and (2) How do 

the different levels of IS task complexity influence the relationships between mind wandering 

and IS decision making performance?  

The main goal of the current study is to contribute to the cognitive IS literature by 

examining the effect of individuals’ focus of attention and mind wandering in influencing IS 

decision making performance at different levels of task complexity. To achieve the objective of 

this research, a laboratory experiment was conducted. During the experiment, subjects were 

instructed to do an online shopping task and at the end of the experiment, they were asked to 

complete self‐report measures of focus of attention and mind wandering. This work makes 

122 
 
several contributions. First, through studying mind wandering and its consequences in IS 

research, this work sheds light on the important function of internal distraction for technology 

users by explicitly acknowledging the role of mind wandering in complex decision making 

environments. Second, this study enhances the findings of study 1 and study 2 discussed in 

Chapter 2 and 3, respectively, by showing that distraction can come not only from external 

environments (voluntary attraction), but also from users’ own thoughts (involuntary attraction). 

Third, this study applies UTT principles and their association with mind wandering to study 

performance variability among technology users. This study also contributes to the literature by 

empirically testing the role of task complexity in influencing the effect of focus of attention and 

mind wandering on both performance efficiency and effectiveness in a technological context. 

4.2  Executive Control Theory of Mind Wandering 

Drawing on executive functions (EFs) theory, Smallwood and Schooler (2006) proposed 

that mind wandering is a mental activity that requires the (limited) resources typically used for 

cognitive control—functions required for maintaining and monitoring goals in the face of 

distraction and interference and for suppressing distracting events (Engle & Kane, 2004; Engle, 

Tuholski, Laughlin, & Conway, 1999; Kane et al., 2007). Despite the spontaneous and lack of 

deliberate intent characteristics of mind wandering, the executive control perspective indicates 

that mind wandering shares certain similarities with the mechanisms of executive control 

(Smallwood & Schooler, 2006). Specifically, Smallwood and Schooler (2006) argued that  

“…it is likely that executive control is not involved in the initiation of mind wandering 
episodes. Rather, the activation of goal‐relevant information may occur automatically through a 
process that does not require conscious intention. The absence of explicit and deliberate intent 
associated with mind wandering may be enabled by the simple fact that we often lack explicit 
123 
 
awareness of the current contents of our own experience…Thus, mind wandering may entail 
situations in which individuals temporarily fail to notice that their immediate goal of task 
completion has been temporarily displaced by another concern.” (p. 946).  
 
If mind wandering is associated with controlled processing, then tasks that rely on the same 

resources should suppress mind wandering. In other words, if the primary task requires 

individuals to maintain task‐relevant information, then few resources will be available to 

coordinate mind wandering.  

Because both the primary task and mind wandering require cognitive resources, this 

theory predicts that mind wandering varies as a function of task complexity—it should be less 

likely to occur when the primary task is complex and more likely to occur when the task is 

simple (Smallwood & Schooler, 2006). When the processing demands are high, mind wandering 

should be associated with low performance because it involves a shift of attention away from 

the outside world. This shift of attention from the primary task to one’s personal goals suggests 

that “mind wandering is a state when information processing is decoupled from the primary 

task” (Smallwood & Schooler, 2006, p. 947). There are two possible consequences of this 

perceptual decoupling. First, it is associated with a reorganization of cognition to focus on 

internal goals rather than external environments. Second, it is associated with reduced 

attention to the external environments. Consequently, under high levels of task complexity, 

mind wandering should be associated with poor performance because fewer resources are 

available to complete the primary task. However, when the task is simple, the effect of mind 

wandering on task performance is not harmful. It is because resources are simultaneously 

available for mind wandering and monitoring of the primary task (Smallwood & Schooler, 

2006). 

124 
 
4.3  Unconscious Thought Theory 

  Decision making research has long been studied in the IS field to explain how technology 

users make decisions in a digital environment (e.g., Agarwal & Singh, 2013; Armbruster & 

Delage, 2015; Barkhi, 2001; Biros, George, & Zmud, 2002; Dennis, 1996; Dennis & Carte, 1998). 

Despite the fact that IS research in decision making has progressed, it has been argued that 

most of the existing research is conducted under the assumption of a rational choice 

perspective (Frisk, Lindgren, & Mathiassen, 2014; Ortiz de Guinea & Markus, 2009). According 

to this perspective, “conscious, deliberate, goal‐directed thought processes” is the best way to 

make a decision (Christoff, Gordon, & Smith,2011; Tan, Yi, & Chan 2015). However, with the 

excessive amount of information generated by technology, it is impossible for users to take into 

account all alternatives needed to make an optimal decision (Hong, Thong, & Tam, 2004). It is 

possible that users can engage in unconscious thought (i.e., cognitive or affective task‐relevant 

processes that take place outside individuals’ conscious awareness) and that conscious 

deliberation is not necessary to arrive at the ideal decision (Dijksterhuis & Aarts, 2010; Ortiz de 

Guinea & Markus, 2009).  

  According to UTT, a simple task can be better solved using conscious thought, whereas a 

complex task can be better approached using unconscious thought (Dijksterhuis & Nordgren, 

2006). In a series of experiments, Dijksterhuis and his colleagues highlighted the importance of 

spontaneous, unconscious thought in a decision making process, especially in a complex 

decision making environment. In these experiments, participants were presented with a 

number of alternatives (e.g., apartments, roommates) with the goal to decide what alternative 

is the most attractive. Participants in the conscious thought condition were given a few minutes 

125 
 
to think about the information before they made a decision. Participants in the unconscious 

thought condition were asked to perform a distracting task for a few minutes before they were 

asked to make a decision. During the distraction period, participants were assumed to engage 

in unconscious, spontaneous thought (Christoff et al., 2011). The results showed that when 

making complex decisions, a brief period of unconscious thought leads to a better decision 

relative to conditions under which unconscious thought is prevented.  

  One major principle of UTT is the capacity principle (Dijksterhuis & Nordgren, 2006). 

According to this principle, conscious thought is constrained by the low capacity of cognitive 

resources. Unconscious thought, however, does not have this limitation. During unconscious 

thought, a large amounts of information can be processed (Dijksterhuis & Aarts, 2010; 

Dijksterhuis, Bos, Nordgren, & Van Baaren, 2006; Dijksterhuis & Nordgren, 2006). Using 

unconscious thought, users are able to integrate huge amounts of information and consistently 

give the pieces of information appropriate weights depending on their relative importance. 

Thus, unconscious thought may lead to better choices than conscious thought when the 

amount of information to be processed is large (Dijksterhuis & Nordgren, 2006).  

4.4  Research Model  

The research model is illustrated in Figure 4.1. In this model, I test the interaction 

effects of task complexity in influencing the relationships between mind wandering, focus of 

attention, and decision making performance. The ultimate goal of testing these interaction 

effects is to improve our understanding on how mind wandering and focus of attention 

influence IS decision making performance, depending on levels of task complexity. The study 

126 
 
focuses on contexts where IS are used to perform specific tasks. For example, workers use a 

web browser to help them search for information that is relevant to their task at hand. The 

definitions of focal constructs are presented in Table 4.1. 

Figure 4.1. Research Model 

Table 4.1 

Construct Definition 

Construct  Definition  Source(s) 


Users’ Performance  “The degree to which the results of the task meet  Ortiz de Guinea and 
task goals.”  Webster (2013, p. 1168) 
Focus of Attention  The degree to which thought is maintained on the  Randall, Oswald, and Beier 
primary task at hand.  (2014) 
Mind Wandering  Self‐generated mental activity that occurs when  Smallwood and Schooler 
attention shifts (spontaneously, and largely  (2006); Unsworth and 
autonomously) from the current task to the  McMillan (2014); Ritter and 
processing of personal goals.  Dijksterhuis (2014) 
Task Complexity  A multi‐dimensional construct as a direct function of  Mennecke, Crossland, and 
the number of information resources, number of  Killingsworth (2000); 
alternative available, and number of dimensions of  Woods (1986) 
information available per alternative that need to be 
processed in the performance of a task.  

127 
 
4.5  Hypotheses Development 

4.5.1  Task Complexity and IS Decision Making Performance 

  Performance refers to “the degree to which the results of the task meet task goals” 

(Ortiz de Guinea & Webster, 2013, p. 1168). In the IS domain, IS task complexity has been 

shown to affect individuals’ performance (e.g., Jiang & Benbasat, 2007; Mennecke et al., 2000; 

Xiao & Benbasat, 2007). As task complexity increases, task difficulty increases and at the same 

time, decision makers take more time to complete the task and produce more faulty results 

(Campbell, 1988; Mennecke et al., 2000; Xu, Benbasat, & Cenfetelli, 2014). High task complexity 

is also associated with high cognitive demands on individuals (Campbell & Gingrich, 1986). 

Simple tasks generally require processing fewer cues (pieces of information) than complex tasks 

(Payne, 1982; Speier et al., 2003). Thus, decision makers need less cognitive resources to 

process simple tasks than when they process complex tasks. One of the research implications 

by these previous studies is the greater the number of information cues to be processed, the 

more difficult it will be for the decision makers to reach the best decision. I apply this 

perspective to study the complexity of IS task and hypothesize that: 

H1: The degree of IS task complexity will have a negative effect on performance 
accuracy. 
 
H2: The degree of IS task complexity will have a negative effect on performance 
efficiency. 
 

4.5.2  Focus of Attention and Performance Efficiency 

According to cognitive resource theory, individuals’ cognitive processes and task 

demands significantly predict variance in task performance (Randall et al., 2014). A number of 

studies have investigated the role of focus of attention in predicting individual differences in job 

128 
 
performance (e.g., Delaney‐Klinger, & Hemingway, 2005; Hunter, 1986; Morgeson, Schmidt, 

2002). Focus of attention can be considered a specific cognitive skill that controls the 

mechanism of distinguishing relevant from irrelevant stimuli and to filter out information that is 

irrelevant to the task at hand (Kaufman, 2011). In particular, when people interact with IT, they 

are easily distracted because multiple applications or features are available in one device to 

simultaneously grab users’ attention. Distractions are not just frustrating; they can be 

exhausting. Staying focused on a task, even when the task is simple, will increase performance 

efficiency because when the primary task is disrupted, users will spend an extra time to switch 

back to the primary task (i.e., higher switching cost is negatively associated with performance 

efficiency). By the time workers get back to where they were, their ability to stay focused goes 

down even further (Rock, 2009). Therefore, individuals’ ability to concentrate or focus on key 

performance related factors while effectively blocking out distractions will influence their 

performance efficiency.  

H3: Under all levels of task complexity, users’ focus of attention has a positive effect on 
performance efficiency. 
 

4.5.3  Mind Wandering and Performance Accuracy 

Mind wandering is referred to as task‐unrelated thought, occurred when individuals are 

not engaged in the task (Smallwood & Schooler 2015). According to Smallwood and Schooler 

(2006), mind wandering recruits executive resources and potentially affects task performance. 

Others believe that mind wandering is the result of executive‐control failure and does not 

require cognitive resources (McVay & Kane, 2010). However, both perspectives agree that mind 

wandering could lead to performance errors (i.e., low performance efficiency) when task 

129 
 
demand for monitoring is high. When our interaction with technology becomes automatic, our 

behaviors and thoughts are often focused on the task (or something else, such as mind 

wandering) rather than on the computer itself (Ortiz de Guinea & Webster, 2013; Louis & 

Sutton, 1991). Because individuals are often unaware of this mind wandering process, it often 

occurs unintentionally and interferes with performance (Forster & Lavie, 2009). Although 

research measuring the impact of mind wandering on performance efficiency in a real life 

setting is limited, recent findings on the association between longer reaction times in a go/no 

go task (a task commonly used to measure EFs) and a higher rate of mind wandering (e.g., 

Smallwood et al., 2004; Forster and Lavie, 2009) may suggest that mind wandering is negatively 

associated with performance efficiency. It is likely because the process of shifting from the main 

task at hand to the state of mind wandering, then bringing back attention to the main task is a 

time consuming process. Therefore, no matter how complex the task is, the completion time 

will actually get longer with the presence of mind wandering.  

H4: Under all levels of task complexity, mind wandering has a negative effect on 
performance efficiency. 
 

4.5.4  The Moderating Effect of IS Task Complexity and Focus of Attention 

The notion of focus, complexity, and performance relationship is supported by 

distraction conflict theory (DCT). According to this theory, simple tasks require processing fewer 

cues than complex tasks (Payne, 1982). On the other hand, when processing complex tasks, 

decision makers minimize their expenditure of scarce cognitive resources, uncritically 

examining both relevant and irrelevant cues (Baron, 1986). Under a simple task environment, 

individuals may have plenty of processing capacity to perform the task. Under this situation, 

130 
 
any distractions may not lower performance because decision makers have sufficient cognitive 

resources to process simple tasks when interruptions occur and do not need to change the way 

in which they process information. However, good performance of complex tasks require more 

of the person’s cognitive resources (Mayer & Gavin, 2005). Under this condition, individuals’ 

effort to focus is more likely to ensure what action is required to keep the goal active and to 

shield it from distraction (Dijksterhuis & Aarts, 2010). Thus, I hypothesize  

H5: The effect of focus of attention on performance accuracy is positively moderated by 
IS task complexity, in which focus of attention will have a minimal effect under low task 
complexity, but will have an increasing effect under medium to high complexity. 
 

4.5.5  The Moderating Effect of IS Task Complexity and Mind Wandering 

Many researchers have found mixed results in studying the relationship between mind 

wandering and task performance accuracy. Some researchers believe that there is a negative 

relationship between mind wandering and performance accuracy. For example, according to 

the cognitive failure hypothesis, a significant proportion of mind wandering episodes in highly 

demanding tasks is caused by a failure of cognitive control to inhibit the irrelevant thought 

(McVay & Kane, 2010). This failure can lead to performance deficiency (McVay & Kane, 2010). 

However, some recent studies in this area offer a contrasting explanation. They suggest that 

mind wandering is actually beneficial. According to this new finding, for individuals who are 

engaged in a highly complex task, mind wandering offers the possibility of very real, personal 

rewards. These rewards include self‐awareness, creative incubation, memory consolidation, 

goal driven thought, future planning, etc. (Kaufman, 2013).  

In order to reconcile these contradictory findings, the characteristics of the task itself 

need to be taken into consideration before we debate which paradigm is more accurate. 

131 
 
According to the deliberation‐without‐attention hypothesis of UTT (Dijksterhuis & Nordgren 

2006), conscious thought does not lead to good decisions in very complex situations. The 

quality of decisions made using conscious thoughts is negatively associated with the complexity 

of the tasks, suggesting that conscious thought may lead to better choices than unconscious 

thought only when the amount of information to be processed is limited (Dijksterhuis and 

Nordgren, 2006).  

Moreover, research has shown that more mind wandering is associated with a greater 

use of unconscious resources to make a decision, leading to more accurate decision making 

under complex situations (albeit with low performance efficiency due to mind wandering) 

(Zhong, Dijksterhuis, & Galinsky, 2008). Therefore, it is reasonable to argue that individuals who 

experience more mind wandering during a complex task tend to use their unconscious 

resources. This argument is also supported by a new theory of personal intelligence by Kaufman 

(2011). According to Kaufman, decreased inhibition, especially under a highly cognitive 

demanding task can have its advantages. This decreased inhibition can make an individual more 

likely to mind wander and be creative in solving the problems. 

However, mind wandering could actually be harmful to performance accuracy when 

individuals use their conscious thought. Conscious thought is more suitable under low to 

medium level task complexity because the standard is easy to determine and amount of 

information is limited. When the task is simple, users are free to direct their thoughts to other 

things (Ortiz de Guinea & Webster, 2013), and thus, mind wandering is less likely to impair task 

performance. Under a moderate complexity condition, however, mind wandering could 

compete with task demands for conscious thought. Consequently, irrelevant items that are not 

132 
 
directed to the intended target of concentration should become more accessible to 

consciousness as the result of mental load (Wegner, 1997). Based on this literature review, I 

hypothesized that mind wandering will have no effect under low task complexity; a negative 

effect under the medium complexity task condition; and it will have a positive effect under high 

complexity task condition. 

H6: The effect of users’ experience of mind wandering on performance accuracy is 
moderated by IS task complexity, in which mind wandering will have no effect under low 
task complexity, negative effect under medium task complexity, but it will have a 
positive effect under high task complexity. 
 

4.6  Research Method 

4.6.1  Experimental Design 

The experiment utilized a Completely Randomized Design (CRD) with three treatment 

groups. The variable manipulated in the experiment is task complexity (high, medium, and low). 

The other two independent variables (i.e., focus of attention and mind wandering) are 

measured as continuous variables. Although mind wandering cannot be directly induced by 

altering the complexity of an ongoing task, it is possible to gain some degree of control over the 

experiment (Smallwood & Schooler, 2015). This manipulation is critical to understand the 

nature of wandering mind because it provides boundary conditions that inform our 

understanding on the functions of mind wandering.  

To operationalize the manipulation of task complexity, the number of features and 

criteria were systematically varied (Wood, 1986). Following Wood’s (1986) concept of task 

complexity, I investigated the choice strategies used by individuals as a function of the three 

task determinants: number of information resources, number of alternatives available, and 

133 
 
number of dimensions of information available per alternative (Mennecke et al., 2000; Payne, 

1976; Speier et al., 2003; Wood, 1986). The task is an online shopping task in the B2C e‐

commerce setting that requires the subjects to evaluate products based on the criteria 

provided by the researchers. This task was designed to represent a familiar IS task that many IS 

users as well as most knowledge workers would do in their job.  

For the purpose of this experiment, the subject was presented with a scenario of 

purchasing a product—a computer and/or laptop. Each product contains a number of features 

that must be considered in the process of decision making. Each subject solved a multi‐criteria 

online decision making task that asked subjects to assume that they had to buy a laptop. Before 

they decided which laptop they were going to buy, they were advised to search for information 

about the products. To help them make the best decision, subjects prioritized potential 

products against the criteria that defined the features of the products. In the low complexity 

condition, subjects were asked to go to an online shopping website and evaluate five product 

alternatives (e.g., MacBook Pro, Asus, etc.) based on three criteria (i.e., operating systems, 

memory and battery life, and price). In the medium task complexity, there were 10 alternatives 

with 7 criteria. Subjects were asked to go to two different websites—5 alternatives for each 

website. In the high task complexity condition, there were 15 alternatives, with 10 criteria for 

each alternative. Subjects were informed to go to three different websites. The decision criteria 

associated with each level of complexity are presented in Table 4.2. The sample experiment 

materials are presented in Appendix E.  

A scoring sheet was provided so that the subjects were able to assign points for each 

criterion described in the tasks (see Appendix E). Participants then were asked to make a 

134 
 
decision based on those criteria. Accuracy was measured by comparing the answers on the 

scoring sheet with the optimal solution prepared by the researcher prior to the experiment. To 

measure the completion time, subjects were asked to record the time when they began and 

finished completing the task. Students received extra credit points from their instructors for 

participation in the study. Gift card prizes were also offered as an incentive for participants who 

scored top performance. 

Table 4.2 

Website and Product Criteria 

Complexity Level  Number of  Number of Product  Criteria 


Source/Website  Alternative 
Low Complexity  1  5   Operating systems 
 Memory and battery life 
 Price 
Medium Complexity  2  10   Operating systems 
 Memory and battery life 
 Price 
 Laptop size 
 Hard drive 
 Processor 
 Item weight 
High Complexity  3  15   Operating systems 
 Memory and battery life 
 Price 
 Laptop size 
 Hard drive 
 Processor 
 Item weight 
 Wi‐Fi and Bluetooth 
 Screen Resolution 
 Number of USB 3 Port 
 

4.6.2  Experimental Procedure 

Those subjects assigned in both low and medium conditions were required to complete 

filler tasks so that all subjects completed relatively the same amount of task time. After 
135 
 
completing the main task, participants completed a paper‐based questionnaire to measure the 

degree of their focus on the task as well as the tendency to engage in mind wandering during 

the task.  

The measurement items for both variables were adapted from the measures established 

and fully validated by previous studies. Items for focus of attention were adapted from 

Nadkarni and Gupta (2007). Study subjects showed a high degree of familiarity with the task 

online shopping, and their interaction with the technology (i.e., shopping websites) is almost 

automatic (Ortiz de Guinea & Markus, 2009; Ortiz de Guinea & Webster, 2013). Thus, their 

focus of attention was measured based on how well they concentrate on solving the problem 

(e.g., strategies to find correct information) rather than on the technical aspect of the computer 

itself (Ortiz de Guinea & Webster, 2013). Items for mind wandering were adapted from the 

measurement developed by Mrazek et al. (2013) and McVay and Kane (2009). Each item was 

worded in accordance with mind wandering in a technological environment. The final items 

were presented in Table 4.3. As expected, a correlation analysis shows that focus is negatively 

correlated with mind wandering (r = ‐0.281; p = 0.013). The exploratory factor analysis (EFA) 

results showed that all of the items loaded on their intended factors (see Table 4.3). 

4.6.3  Manipulation Check 

As a manipulation check, I examined subjects’ perceptions of task complexity using a 

single measure (on a 7‐point Likert scale) “working on the task did not require a lot of effort.” 

The results show a significant difference (F = 16.27, p<.001) of perceived task complexity among 

low (M = 5.69; SD = 1.44; n = 26), medium (M = 4.68; SD = 1.52; n = 25) and high (M = 3.48; SD = 

136 
 
1.28; n = 27), confirming a successful manipulation of task complexity. Furthermore, according 

to Miller’s (1956) information processing theory, people can hold 7 + 2 chunks of information in 

their working memory concurrently. Therefore, the first treatment, with the 5 product 

alternatives and 3 criteria (within the lower limit of Miller’s prediction) represents a low level of 

complexity; the second treatment, with the 10 alternatives and 7 criteria (within the higher 

limit of Miller’s prediction) represents a medium level of complexity; and the third treatment, 

with the 15 alternatives and 10 criteria represents a high level of complexity. Jiang and 

Benbasat (2007) suggested that based on the findings reported in the cognitive psychology 

literature, this manipulation of task complexity can be considered objective. Therefore, it can 

be generalized into other situations as this measure is not necessarily influenced by specific 

experimental conditions.  

Table 4.3 

Measurement Items and EFA Results  

Component  Cronbach’s 
Construct   Items 
1  2  Alpha 
Focus of  (When I was using the website to complete the task), I 
‐0.20  0.74 
attention  had a distinct identifiable purpose. 
…, I was able to look up specific information.  ‐0.20  0.81  0.81 
…, I was able to focus on finding specific information.   ‐0.16  0.77 
…, I was absorbed in finding specific information.  0.09  0.87 
Mind  (When I was using the website to complete the task), 
0.72  ‐0.35 
Wandering  sometimes I cannot easily pay attention when I should. 
…, I found myself always distracted by other things in 
0.85  ‐0.03 
mind. 
…, I had so many things in mind.  0.80  ‐0.12  0.93 
…, my mind wandered.  0.90  ‐0.16 
…, I was easily distracted by unnecessary information in 
0.93  ‐0.23 
mind. 
…, sometimes I had trouble concentrating on the tasks.  0.85  0.01 
 
 
137 
 
4.6.4  Sample 

Student subjects were recruited from a college of business at a large public university in 

the U.S. A total of 84 subjects participated in the experiment. Among them, 6 responses were 

regarded invalid, resulting in 78 valid responses (26, 25, and 27 responses in treatment 1, 2, and 

3, respectively). According to Hair, Black, Babin, and Anderson (2010), the number of subjects in 

each treatment group should be at least 20. Although the sample size is relatively small, it 

satisfies the requirement of conducting an experiment. Furthermore, a power analysis showed 

that the sample is large enough to test the main effect and major interaction effects (d = .82) 

(Cohen, 1988). 

Among the participants, 40 (51%) were female. 21 (27%) of the participants were 

between 17‐20 years old; 43 (55%) were from 21‐25 years old; and the rest of the participants 

were above 25 years old, but no more than 35 years old. In general, the participants were 

familiar with the internet shopping (M = 5.96/7) and computer (M = 5.95/7) and had a high 

degree of computer self‐efficacy (M = 5.59/7). There were no significant differences across 

three complexity conditions with respect to gender, age, prior internet and computer 

experience, and computer self‐efficacy. 

4.7  Data Analysis and Results 

The results of ANOVA test on task performance are reported in Table 4.4. Levine test for 

inequality of variance was statistically significant for both decision accuracy and decision time. 

For that reason, a square root transformation was performed on decision time, and a log 

transformation was performed on decision accuracy (Dennis & Carte, 1998). The inequality of 

138 
 
variance test on the transformed data found no significant differences, indicating that both 

transformations were successful in addressing the issue of inequality of variance. The results 

showed that there were significant differences among the three complexity levels on both 

performance accuracy and performance efficiency. 

Table 4.4 

Group Comparison (ANOVA) Results 

Task Performance  Mean  SD  SS  MS  Df  F  Sig. 


Performance Accuracy               
    Low   .84  .09  11.12  5.56  2  70.74  .000 
    Medium  .66  .66 
    High  .36  .36 
Performance Efficiency*               
    Low  10.97  3.45  83.65  41.83  2  83.47  .000 
    Medium  20.48  7.00 
    High  33.68  5.55 
Notes: *In minutes; SS = Sum of Squares; MS = Mean Squares; Means and SDs reported in this table are the actual 
descriptive statistics before the transformation. 
 

4.7.1  Performance Accuracy 

The Generalized Linear Model (GLM) procedure in SAS was used to analyze the data 

using task complexity as a categorical variable and both focus and mind wandering as 

continuous variables. To test the main effect hypotheses, the additive model was observed 

(Table 4.5, Model 1). As predicted, the effect of task complexity on performance accuracy was 

significant. Therefore, H1 was supported. The results also show that the main effect of focus of 

attention on decision accuracy was significant. Model 2 (Table 4.5) shows that the interaction 

term between task complexity and focus of attention was significant (F‐stat = 4.30; p = 0.02). 

Similarly, the interaction between task complexity and mind wandering was also significant (F‐

stat = 3.30; p = 0.048). Thus, H5 and H6 were supported. Together, all of these factors 

139 
 
accounted for 73% of the variance in performance accuracy. To examine the predictive power 

of the proposed model, I compare the R2 of Model 1 with Model 2 using Cohen’s f2 formula. The 

result shows that the effect size is high (f2 = .19). 

Table 4.5 

GLM Results for Performance Accuracy 

  Model 1  Model 2 
IVs  F‐stat  p‐value  F‐stat  p‐value 
Task complexity  61.91  <.0001  10.21  0.0001 
Focus of attention  18.00  <0.001  22.34  <.0001 
Mind wandering  2.44  0.123  3.80  0.0553 
Task complexity x Focus of attention    4.30  0.0174 
Task complexity x Mind wandering  3.30  0.0427 
R2  0.675  0.728 
Δ R2                                                         0.053*** (f2 = 0.19) 
Note: Test statistics reported for performance accuracy are based on a log transformation.  

In order to better understand the direction of the effects of independent variables at 

each level of task complexity (when the interaction terms are significant), I ran a regression 

analysis with dummy coding (high complexity group was used as a reference group). The results 

are presented in Table 4.6. Because the interaction terms were significant, the model needs to 

be interpreted in the presence of other variables in the interaction model. In the non‐additive 

model (Model 2), low complexity estimate (β = 2.41; t = 4.15; p<.0001) was the estimated 

difference in the mean of performance accuracy between low and high complexity conditions. 

The positive sign indicated that performance accuracy of low complexity was higher than high 

complexity. 

  Similarly, medium complexity estimate was 1.195 (t = 2.12; p<.05), indicating that 

performance accuracy of a medium complexity condition was higher than a high complexity 

140 
 
condition. These results provide additional support for H1. For a high complexity condition, the 

influence of focus of attention on performance was significant (β = 0.256; t = 4.18; p<.001). 

However, there was no difference in slopes for medium and high complexity conditions. 

Furthermore, the interaction between high complexity and mind wandering was significant (β = 

0.115; t = 3.15; p<.01). 

Table 4.6 

Regression Analysis with Dummy Coding for Task Complexity 

  Model 1  Model 2 
Parameter  Estimate  t‐value  Estimate  t‐value  VIF 
Intercept (b0)  ‐1.908  ‐8.27***  ‐2.750  ‐6.86***  ‐ 
Low (b1)  0.775  11.05***  2.415  4.51***  1.353 
Medium (b2)  0.464  6.42***  1.195  2.12*  1.436 
Focus of attention (b3)  0.148  4.24***  0.256  4.18***  3.924 
Mind wandering (b4)  0.034  1.79+  0.115  3.15**  3.464 
Low x Focus of attention (b5)      ‐0.230  ‐2.80**  2.353 
Medium x Focus of attention (b6)  ‐0.07  ‐0.83 (Ns)  2.24 
Low x Mind wandering (b7)  ‐0.108  ‐2.19*  2.250 
Medium x Mind wandering (b8)  ‐0.116  ‐2.29*  2.318 
***p<.001; **p<0.01; *p<.05; Ns= Not significant. 

Next, based on Table 4.6, I computed the regression equations and tested the 

significance of slope coefficients for each group (see Table 4.7). To test whether the slopes are 

significant in each group, I re‐ran multiple regression analysis with different reference groups. 

As shown in Table 4.7, focus of attention had no significant effect under low task complexity. 

However, its effect was significant under both low and high task complexity. Thus, H5 was 

partially supported. As I hypothesized, mind wandering had a positive effect under high task 

complexity. However, mind wandering did not have any significant effect under both low and 

medium complexity. Thus, H6 was supported. 

141 
 
Table 4.7. 

Regression Equation for Each Condition (DV: Performance Accuracy) 

t‐value 
Task  Regression Equation 
FA  MW 
Low  Y = (b0+b1) + (b3+b5)*FA + (b4+b7)*MW  Y = ‐.335 + .026*FA + .007*MW  Ns  Ns 
Med  Y = (b0+b2) + (b3+b6)*FA + (b4+b8)*MW  Y = ‐1.555 + .256*FA ‐ .001*MW  3.33***  Ns 
High  Y = b0 + b3*FA + b4*MW  Y = ‐2.750 + .256*FA + .115*MW  4.18***  3.15** 
Notes: Y= Decision Accuracy; FA= Focus of attention; MW= Mind Wandering; ***p<.001; **p<0.01; Ns= Not significant. 

 
Furthermore, I conducted a follow‐up contrast on the two‐way interactions to test 

significant differences between two treatment groups. I found that four of six contrasts were 

significant, confirming the significance of interaction effects (see Table 4.8). The first contrast 

analysis indicated that under high task complexity, the effect of focus of attention was higher 

than when task complexity is low (the negative parameter estimate indicates that the 

interaction term of high complexity is higher than low complexity). However, the effect of focus 

of attention on performance accuracy did not differ under medium and high complexity, 

suggesting a monotonic increase in the effect of focus of attention and task complexity on 

performance accuracy (low < medium/high). Furthermore, the second contrast indicated that 

the strength of the interaction between mind wandering and task complexity under the low 

and medium conditions was not significantly different. However, mind wandering had a 

stronger effect on performance accuracy when task complexity was high than when it was 

either low or medium.  

The interaction effects can be explained by inspecting Figure 4.2 and 4.3. As shown in 

Figure 4.2, the effect of mind wandering on performance accuracy was not significant under the 

low and medium condition, whereas its effect was positive and significant under the high 

142 
 
condition. As presented in Figure 4.3, the effect of focus of attention on performance accuracy 

was positive and significant under the medium and high condition, but not under the low 

condition. 

Table 4.8 

Interaction Contrast Analysis Results 

Interaction  Contrast  Estimate  t‐value 


Task complexity x Focus of attention  Low vs. High  ‐.231  ‐2.82** 
Medium vs. High   ‐.067  ‐0.80 (Ns) 
Low vs. Medium   ‐.164  ‐2.09* 
Task complexity x Mind wandering  Low vs. High  ‐.109  ‐2.22* 
Medium vs. High   ‐.112  ‐2.14* 
Low vs. Medium   .002  0.05 (Ns) 
**p<.01; *p<.05; Ns= not significant. 

Figure 4.2. The Moderating Effect of Task Complexity between Mind Wandering and 
Performance Accuracy 
 

143 
 
 

Figure 4.3. The Moderating Effect of Task Complexity between Focus of Attention and 
Performance Accuracy 
 

4.7.2  Performance Efficiency 

Next, I tested the effects of task complexity, mind wandering, and focus of attention on 

performance efficiency. For an easy interpretation, the scale of decision time was reverse‐ 

coded; thus, high scores represent high performance efficiency. The results of regression with 

dummy coding analysis are presented in Table 4.9.  

As hypothesized, task complexity had a negative effect on performance efficiency (β = 

2.49, p<.001 indicated that low complexity had a higher efficiency than high complexity and β = 

1.43, p<.001 indicated that medium complexity had a higher efficiency than high complexity). 

Thus, H2 was supported. The effect of focus of attention on performance efficiency was not 

significant, failing to support H3. As hypothesized, the main effect of mind wandering on 

performance efficiency was significant, supporting H4. Because I only hypothesized the main 

144 
 
effect of both mind wandering and focused attention, I expected the three complexity 

conditions would have identical slopes, b3 and b4. This means the three regression lines will be 

parallel to one another (see Table 4.10). I calculated the effect size of the proposed model by 

comparing Model 1 (dummy variables only) and Model 2 (dummy variables and continuous 

variables) (Aiken & West, 1991). The Cohen’s f2 effect size is .14, indicating the effect size of 

mind wandering is relatively high. I summarize the hypotheses results in Table 4.11. 

Table 4.9 

Regression Analysis with Dummy Coding for Task Complexity (DV: Performance Efficiency) 
 
  Model 1  Model 2 
IVs  Estimate  t‐value  Estimate  t‐value 
Intercept (b0)  4.21  30.94***  4.63  7.47*** 
Low (b1)  2.51  12.89***  2.49  13.19*** 
Medium (b2)  1.37  6.99***  1.43  7.36*** 
Focus of attention (b3)      .00  .01 (ns) 
Mind wandering (b4)      ‐0.14  ‐2.58** 
R2  .68    .72 
ΔR2       0.04*** (f2 = 0.14) 
Notes: Test statistics reported for performance efficiency are based on a square root transformation of the 
raw data; ***p<.001; **p<0.01; *p<.05; High Complexity was used as a reference group; ns= not significant. 

 
Table 4.10 

Regression Equation for Each Condition (DV: Performance Efficiency) 

t‐value 
Condition  Regression Equation 
FA  MW 
Low  Y = b1 + b0 + b3*FA + b4*MW  Y = 7.12 + .00*FA ‐ .14*MW  Ns  ‐2.58** 
Medium  Y = b2 + b0 + b3*FA + b4*MW  Y = 6.06 + .00*FA ‐ .14*MW  Ns  ‐2.58** 
High  Y = b0 + b3*FA + b4*MW  Y = 4.63 + .00*FA ‐ .14*MW  Ns  ‐2.58** 
Notes: Y= Decision Efficiency; FA= Focus of Attention; MW= Mind Wandering; Ns= Not Significant; **p<0.01. 

   

145 
 
Table 4.11 

Hypotheses Summary 

  Hypothesis  Supported? 
H1  The degree of IS task complexity will have a negative effect on performance  Yes 
accuracy. 
H2  The degree of IS task complexity will have a negative effect on performance  Yes 
efficiency. 
H3  Under all levels of task complexity, users’ focus of attention has a positive effect  No 
on performance efficiency. 
H4  Under all levels of task complexity, mind wandering has a negative effect on  Yes 
performance efficiency. 
H5  The effect of focus of attention on performance accuracy is positively moderated  Yes 
by IS task complexity, in which focus of attention will have a minimal effect 
under low task complexity, but will have an increasing effect under medium to 
high complexity. 
H6  The effect of users’ experience of mind wandering on performance accuracy is  Partially 
moderated by IS task complexity, in which mind wandering will have no effect  supported 
under low task complexity, negative effect under medium task complexity, but it 
will have a positive effect under high task complexity. 
 

4.8  Discussion and Implications 

Drawing on the mind wandering and IS literature, this work empirically demonstrates 

the importance of focus of attention and mind wandering in influencing users’ performance at 

different levels of task complexity. Specifically, the results demonstrate that users’ performance 

is influenced by individuals’ focus of attention and mind wandering. In relatively difficult tasks, 

focus of attention represents a potent predictor of performance accuracy. In a medium and low 

task environment, mind wandering does not have a significant effect on performance accuracy. 

However, in a high complexity task environment, mind wandering can bring a positive effect on 

users’ performance.  

146 
 
4.8.1  Theoretical Implications 

This research makes several contributions. Responding to the calls for richer 

understanding of unconscious processes to study phenomena in the IS discipline (e.g., Ortiz de 

Guinea & Markus, 2009), I integrated a cognition‐oriented construct (i.e., focus of attention) 

and a more spontaneous cognition construct (i.e., mind wandering) and test their influence on 

IS task performance. The mind wandering construct enhances our understanding on the effect 

of internal distraction on performance. It shows that distraction can come not only from an 

external environment, but also from users’ own thoughts. I introduce the role of focus of 

attention as an individual cognitive factor that should be considered an important factor in IS 

environments. 

Furthermore, this current study integrates UTT and mind wandering theory to study 

performance variability among technology users. There is an ongoing debate on whether mind 

wandering requires cognitive resources (Smallwood & Schooler, 2006) or it is a spontaneous 

cognition that can be beneficial for individuals’ performance (Baird et al., 2012). My study 

found that under low task complexity, mind wandering is not a main determinant of 

performance. It is probably because task automaticity is likely to take place and users’ minds 

are free to wander without hurting their performance. Contrary to my hypotheses, mind 

wandering does not have a significant effect on performance accuracy under medium task 

complexity. Under this condition, focus of attention is the main predictor of performance 

accuracy. However, under high complexity condition, mind wandering has a positive effect on 

performance accuracy. It is likely because individuals who experience more mind wandering 

during the complex task tend to use their unconscious resources; thus, are more creative in 

147 
 
solving the problems. Future research is needed to explore the relationship between 

performance and mind wandering under various settings. Furthermore, more advanced 

techniques, such as eye tracking or brain imaging could be used to enhance the theoretical 

implications of mind wandering in IS research.  

This study also contributes to the literature by empirically testing the role of task 

complexity in influencing the effect of focus of attention and mind wandering on IS decision 

making performance. Task and individual characteristics have been found to affect users’ 

performance (Campbell, 1988). This study analyzed task characteristics by manipulating the 

complexity of decision making under a B2C e‐commerce environment. The results show how 

different levels of task complexity can moderate the impact of cognitive constructs on 

performance. However, I did not find that focus of attention significantly influences 

performance efficiency. This is probably because subjects’ primary goal was to achieve high 

accuracy; thus, they were likely to optimize their time resource to increase their output 

accuracy. Another possibility is that its effect was attenuated by mind wandering regardless the 

complexity level of the task. Further research is needed to investigate how individuals deal with 

both internal and external distractions and how focus of attention influences their 

performance.  

4.8.2  Practical Implications 

This research has several practical implications, such as training programs to improve 

focus of attention of technology users and knowledge workers. Because the results show that 

focus of attention will significantly impact the quality of the work of IS users, training should be 

148 
 
designed such that individuals can learn how to pay attention to their tasks. Research has 

shown that focus of attention can become stronger with practice and it demonstrates 

significant improvements over a relatively brief period of time (Baime, 2011). The results show 

that mind wandering can be beneficial for technology users. With IS that are fundamentally 

complex, organizations can incorporate mindfulness program in their work environment. 

Mindfulness program will help IS users to transform their mind wandering experience into an 

innovative thought process. 

This research also has implications for IS and HCI designs. As the results indicate that 

task complexity will influence the effect of focus of attention and mind wandering on 

performance, I suggest that systems be designed in such way to maintain the balance between 

individuals’ focus attention and mind wandering, particularly if the task is relatively complex. 

One possible design alternative is by creating a system that can estimate the complexity of the 

decision making task. The system should be designed to facilitate mind wandering by 

deliberately distracting individuals’ decision making process before a final decision is made. For 

example, users can be instructed to take a break and engage in an undemanding task or 

technology, and then, they can be directed back to the primary task.     

4.8.3  Limitations 

This study presents some limitations due to the use of a self‐report method to measure 

mind wandering. Although the validity of this technique to measure mind wandering has been 

well established (e.g., McVay & Kane, 2009; Mrazek et al., 2013; Smallwood & Schooler, 2015), 

it is still difficult to verify the results objectively. Future research is needed to explore the 

149 
 
possibility of using different techniques (e.g., eye‐tracking technique, direct observation of 

brain activity, etc.) to objectively measure mind wandering. Further, to overcome the 

disadvantages of every single method used and enhance the findings, a triangulation between 

different methods can be used (Smallwood & Schooler, 2015). 

In this study, three actual shopping websites were used to represent a real shopping 

experience. It should be noted that these three websites are different in their specific contents 

and complexity. Thus, strictly speaking, user performance differences may not be exclusively 

attributed to the complexity manipulation. However, website complexity is generally used as a 

proxy of task complexity (Nadkarni & Gupta, 2007), and it can be viewed as an important 

element of task complexity. Furthermore, given that the tasks performed using all three 

websites are similar in nature—despite it is practically impossible to identify and select three 

websites that contain the same amount of information and have the same level of complexity—

I believe my task complexity manipulation is acceptable.  

The experiment task (i.e., e‐commerce decision making) was designed to represent the 

most familiar IS task. The complexity was manipulated by varying the number of sources, 

number of products, and number of attributes to be processed by the participants. It is possible 

that other tasks, such as managerial decision making, may contain different complexity 

attributes (e.g., uncertainty and decision immediacy). Thus, it may be necessary to test the 

theoretical framework in different contexts.  

150 
 
4.9  Conclusions 

  The main goal of this study is to contribute to the cognitive Information Systems (IS) 

literature by examining the effect of individuals’ focus of attention and mind wandering in 

influencing IS decision making performance at different levels of task complexity. Drawing upon 

UTT and executive control theory of mind wandering, the results show that (1) higher task 

complexity leads to lower IS task performance; (2) under low task complexity, focus of attention 

and mind wandering do not have any significant effects on task performance; (3) under 

medium task complexity, focus of attention leads to higher performance accuracy, whereas 

mind wandering does not have a significant effect on performance accuracy; (4) under high task 

complexity, both focus of attention and mind wandering lead to higher performance accuracy. 

Furthermore, mind wandering negatively influences performance efficiency under all levels of 

IS task complexity. 

4.10  References 

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. 
Thousand Oaks, CA: Sage Publications. 
Agarwal, R., & Singh, H. (2013). Differential influence of blogs across different stages of decision 
making: The case of venture capitalists. MIS Quarterly, 37(4), 1093‐1112. 
Andrews‐Hanna, J. R., Kaiser, R. H., Turner, A. E., Reineberg, A. E., Godinez, D., Dimidjian, S., & 
Banich, M. T. (2013). A penny for your thoughts: Dimensions of self‐generated thought 
content and relationships with individual differences in emotional wellbeing. Frontiers in 
Psychology, 4, 1‐13. 
Armbruster, B., & Delage, E. (2015). Decision making under uncertainty when preference 
information is incomplete. Management Science, 61(1), 111‐128. 
Baird, B., Smallwood, J., Mrazek, M. D., Kam, J. W., Franklin, M. S., & Schooler, J. W. (2012). 
Inspired by distraction mind wandering facilitates creative incubation. Psychological Science, 
23(10), 1117‐1122. 

151 
 
Baime, M. (2011). This is your brain on mindfulness. Shambhala Sun (July), 45‐49. 
Barkhi, R. (2001). The effects of decision guidance and problem modeling on group decision‐
making. Journal of Management Information Systems, 18(3), 259‐282. 
Baron, R. S. (1986). Distraction‐conflict theory: Progress and problems. Advances in 
Experimental Social Psychology, 19, 1‐39. 
Biros, D. P., George, J. F., & Zmud, R. W. (2002). Inducing sensitivity to deception in order to 
improve decision making performance: A field study. MIS Quarterly, 26(2), 119–144. 
Campbell, D. J. (1988). Task complexity: A review and analysis. Academy of Management 
Review, 13(1), 40‐52. 
Campbell, D. J., & Gingrich, K. (1986). The interactive effects of task complexity and 
participation on task performance: A field experiment. Organizational Behavior and Human 
Decision Processes, 38(2), 162‐180. 
Christoff, K., Gordon, A., & Smith, R. (2011). The role of spontaneous thought in human 
cognition,” In O. Vartanian & D. R. Mardel (Eds.), Neuroscience of decision making (pp. 259‐
284), New York, NY: Psychology Press. 
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd Ed.). Hillsdale, NJ: 
Lawrence Earlbaum Associates. 
Dabbish, L., & Kraut, R. (2008). Awareness displays and social motivation for coordinating 
communication. Information Systems Research, 19(2), 221‐238. 
Davis, J. (2015). Zoning out can make you more productive. Harvard Business Review, Retrieved 
from https://hbr.org/2015/06/zoning‐out‐can‐make‐you‐more‐productive. 
Dennis, A. R. (1996). Information exchange and use in group decision making: You can lead a 
group to information, but you can’t make it think. MIS Quarterly, 20(4), 433‐457. 
Dennis, A. R., & Carte, T. A. (1998). Using geographical information systems for decision making: 
Extending cognitive fit theory to map‐based presentations. Information Systems Research, 
9(2), 194‐203. 
Dijksterhuis, A., & Aarts, H. (2010). Goals, attention, and (un)consciousness. Annual Review of 
Psychology, 61, 467‐490. 
Dijksterhuis, A., Bos, M. W., Nordgren, L. F., & Van Baaren, R. B. (2006). On making the right 
choice: The deliberation‐without‐attention effect. Science, 311(5763), 1005‐1007. 
Dijksterhuis, A., & Nordgren, L. F. (2006). A theory of unconscious thought. Perspectives on 
Psychological Science, 1, 95‐109. 
Engle, R. W., & Kane, M. J. (2004). Executive attention, working memory capacity, and a two‐
factor theory of cognitive control. In B. H. Ross (Ed.), Psychology of learning and motivation 
(Vol. 44, pp. 145‐200), San Diego, CA: Elsavier Academic Press. 

152 
 
Engle, R. W., Tuholski, S. W., Laughlin, J. E., & Conway, A. R. (1999). Working memory, short‐
term memory, and general fluid intelligence: A latent‐variable approach. Journal of 
Experimental Psychology: General, 128(3), 309‐331. 
Forster, S. (2013). Distraction and mind wandering under load. Frontiers in Psychology, 4, 1‐6. 
Forster, S., & Lavie, N. (2009). Harnessing the wandering mind: The role of perceptual load. 
Cognition, 111(3), 345‐355. 
Frisk, J. E., Lindgren, R., & Mathiassen, L. 2014. Design matters for decision makers: Discovering 
IT investment alternatives. European Journal of Information Systems, 23(4), 442‐461. 
Hair, J. F. Jr., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Multivariate data analysis (7th 
ed.). Upper Saddle River, NJ: Prentice Hall.   
Hemp, P. (2009). Death by information overload. Harvard Business Review (September), 1‐10. 
Hong, W., Thong, J. Y. L., & Tam, K. Y. (2004). Does animation attract online users’ attention? 
The effects of flash on information search performance and perceptions. Information 
Systems Research, 15(1), 60‐86. 
Hunter, J. E. (1986). Cognitive ability, cognitive aptitudes, job knowledge, and job performance. 
Journal of Vocational Behavior, 29(3), 340‐362. 
Jiang, Z., & Benbasat, I. (2007). The effects of presentation formats and task complexity on 
online consumers’ product understanding. MIS Quarterly, 31(3), 475‐500. 
Jiang, Y., & Punj, G. N. (2010). The effects of attribute concreteness and prominence on 
selective processing, choice, and search experience. Journal of the Academy of Marketing 
Science, 38(4), 471‐489. 
Kane, M. J., Brown, L. H., McVay, J. C., Silvia, P. J., Myin‐Germeys, I., & Kwapil, T. R. (2007). For 
whom the mind wanders, and when: An experience‐sampling study of working memory and 
executive control in daily life. Psychological Science, 18, 614‐621. 
Kaufman, S. B. (2011). Intelligence and the cognitive unconscious. In R. J. Sternberg & S.B. 
Kaufman (Eds.), Cambridge handbook of intelligence (pp. 442‐467), Cambridge, UK: 
Cambridge University Press. 
Kaufman, S. B. (2013). Mind wandering: A new personal intelligence perspective. Scientific 
America, Retrieved from http://blogs.scientificamerican.com/beautiful‐
minds/2013/09/25/mind wandering‐a‐new‐personal‐intelligence‐perspective/.  
Killingsworth, M. A., & Gilbert, D. T. (2010). A wandering mind is an unhappy mind. Science, 
330, 932. 
Louis, M. R., & Sutton, R. I. (1991). Switching cognitive gears: From habits of mind to active 
thinking. Human Relations, 44(1), 55‐76. 

153 
 
Mayer, R. C., & Gavin, M. B. (2005). Trust in management and performance: Who minds the 
shop while the employees watch the boss? Academy of Management Journal, 48(5), 874‐
888. 
McVay, J. C., & Kane, M. J. (2009). Conducting the train of thought: Working memory capacity, 
goal neglect, and mind wandering in an executive‐control task. Journal of Experimental 
Psychology: Learning, Memory, and Cognition, 35(1), 196‐204. 
McVay, J. C., & Kane, M. J. (2010). Does mind wandering reflect executive function or executive 
failure? Comment on Smallwood and Schooler (2006) and Watkins (2008). Psychological 
Bulletin, 136, 188‐197. 
Mennecke, B. E., Crossland, M. D., & Killingsworth, B. L. (2000). Is a map more than a picture? 
The role of SDSS technology, subject characteristics, and problem complexity on map 
reading and problem solving. MIS Quarterly, 24(4), 601‐629. 
Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity 
for processing information. Psychological Review, 63, 81‐97. 
Moberly, N. J., & Watkins, E. R. (2008). Ruminative self‐focus and negative affect: An experience 
sampling study. Journal of Abnormal Psychology, 117(2), 314‐323. 
Morgeson, F. P., Delaney‐Klinger, K., & Hemingway, M. A. (2005). The importance of job 
autonomy, cognitive ability, and job‐related skill for predicting role breadth and job 
performance. Journal of Applied Psychology, 90(2), 399‐406. 
Mrazek, M. D., Phillips, D. T., Franklin, M. S., Brodway, J. M., & Schooler, J. W. (2013). Young 
and restless: Validation of the mind wandering questionnaire (MWQ) reveals disruptive 
impact of mind wandering for youth. Frontiers in Psychology, 4, 1‐7. 
Nadkarni, S., & Gupta, R. (2007). A task‐based model of perceived website complexity. MIS 
Quarterly, 31(3), 501‐524. 
Ophir, E., Nass, C., & Wagner, A. D. (2009). Cognitive control in media multitaskers. Proceedings 
of the National Academy of Sciences of the United States of America, 106(37), 15583‐15587. 
Ortiz de Guinea, A. O., & Markus, M. L. (2009). Why break the habit of a lifetime? Rethinking 
the roles of intention, habit, and emotion in continuing information technology use. MIS 
Quarterly, 33(3), 433‐444. 
Ortiz de Guinea, A. O., & Webster, J. (2013). An investigation of information systems use 
patterns: Technological events as triggers, the effect of time, and consequences for 
performance. MIS Quarterly, 37(4), 1165‐1188. 
Payne, J. W. (1976). Task complexity and contingent processing in decision making: An 
information search and protocol analysis. Organizational Behavior and Human Performance, 
16(2), 366‐387. 
Payne, J. W. (1982). Contingent decision behavior. Psychological Bulletin, 92(2), 382–402. 

154 
 
Randall, J. G., Oswald, F. L., & Beier, M. E. (2014). Mind wandering, cognition, and performance: 
A theory‐driven meta‐analysis of attention regulation. Psychological Bulletin, 140(6), 1411‐
1431. 
Ritter, S. M., & Dijksterhuis, A. (2014). Creativity: The unconscious foundation of the incubation 
period. Frontiers in Human Neuroscience, 8, 1‐10. 
Rock, D. (2009). Easily distracted: Why it's hard to focus, and what to do about it. In Your Brain 
at Work, Psychology Today, Retrieved from http://www.psychologytoday.com/blog/your‐
brain‐work/200910/easily‐distracted‐why‐its‐hard‐focus‐and‐what‐do‐about‐it. 
Ruby, F. J., Smallwood, J., Engen, H., & Singer, T. (2013). How self‐generated thought shapes 
mood‐the relation between mind wandering and mood depends on the socio‐temporal 
content of thoughts. PLoS One, 8(10), 1‐7. 
Schmidt, F. L. (2002). The role of general cognitive ability and job performance: Why there 
cannot be a debate. Human Performance, 15(1‐2), 187‐210. 
Smallwood J. M., Davies J. B., Heim D., Finnigan F., Sudberry M., & O’Connor R. (2004). 
Subjective experience and the attentional lapse: Task engagement and disengagement 
during sustained attention. Consciousness and Cognition, 13, 657‐690. 
Smallwood, J. M., & Schooler, J. W. (2006). The restless mind. Psychological Bulletin, 132(6), 
946‐958. 
Smallwood, J. M., & Schooler, J. W. (2015). The science of mind wandering: Empirically 
navigating the stream of consciousness. Annual Review of Psychology, 66, 487‐518. 
Speier, C., Vessey, I., & Valacich, J. S. (2003). The effects of interruptions, task complexity, and 
information presentation on computer‐supported decision‐making performance. Decision 
Sciences, 34(4), 771‐797. 
Tan, B., Yi, C., & Chan, H. C. (2015). Deliberation without attention: The latent benefits of 
distracting website features for online purchase decisions. Information Systems Research, 
26(2), 437‐455. 
Unsworth, N., & McMillan, B. D. (2014). Similarities and differences between mind wandering 
and external distraction: A latent variable analysis of lapses of attention and their relation 
to cognitive abilities. Acta Psychologica, 150, 14‐25. 
Wati, Y., Koh, C., & Davis, F. D. (2014). Can you increase your performance in a technology‐
driven society full of distraction?. Poster presented at the 35th International Conference on 
Information Systems, Auckland, NZ, December 14‐17. 
Wegner, D. M. (1997). Why the mind wanders. In J. D. Cohen & J. W. Schooler (Eds.), Scientific 
approaches to consciousness (pp. 295‐315), Mahwah, NJ: Erlbaum. 
Winters, D., & Latham, G. P. (1996). The effect of learning versus outcome goals on a simple 
versus a complex task. Group & Organization, 21(2), 236‐250. 

155 
 
Wood, R. E. (1986). Task complexity: Definition of the construct. Organizational Behavior and 
Human Decision Processes, 37(1), 60‐82. 
Xiao, B., & Benbasat, I. (2007). E‐commerce product recommendation agents: Use, 
characteristics, and impact. MIS Quarterly, 31(1), 137‐209. 
Xu, J. D., Benbasat, I., & Cenfetelli, R. T. (2014). The influences of online service technologies 
and task complexity on efficiency and personalization. Information Systems Research, 25(2), 
420‐436. 
Zhong, C. B., Dijksterhuis, A., & Galinsky, A. D. (2008). The merits of unconscious thought in 
creativity. Psychological Science, 19(9), 912‐918. 
   

156 
 
 
 
 
 
 
 
 
 
 
 
 
 
APPENDIX A 

TESTING FOR COMMON METHOD BIAS (WITH A COMMON LATENT FACTOR) 

157 
 
Relationship  Without CLF*  With CLF*  Delta 
Anx1  <‐‐‐  comp_anxiety  0.72  0.62  0.10 
Anx2  <‐‐‐  comp_anxiety  0.77  0.68  0.09 
Anx3  <‐‐‐  comp_anxiety  0.80  0.72  0.08 
Anx4  <‐‐‐  comp_anxiety  0.89  0.83  0.06 
Anx5  <‐‐‐  comp_anxiety  0.93  0.88  0.05 
Anx6  <‐‐‐  comp_anxiety  0.93  0.88  0.05 
Comp1  <‐‐‐  Task_Complex  0.66  0.55  0.11 
Comp3  <‐‐‐  Task_Complex  0.85  0.80  0.05 
Comp4  <‐‐‐  Task_Complex  0.70  0.58  0.12 
MW_IT1  <‐‐‐  MW_IT  0.75  0.66  0.09 
MW_IT2  <‐‐‐  MW_IT  0.88  0.82  0.06 
MW_IT3  <‐‐‐  MW_IT  0.94  0.88  0.06 
MW_IT4  <‐‐‐  MW_IT  0.82  0.74  0.08 
MW_IT5  <‐‐‐  MW_IT  0.87  0.80  0.06 
FA_IT1  <‐‐‐  FA_IT  0.77  0.68  0.09 
FA_IT2  <‐‐‐  FA_IT  0.90  0.80  0.10 
FA_IT3  <‐‐‐  FA_IT  0.87  0.77  0.10 
FA_IT4  <‐‐‐  FA_IT  0.87  0.77  0.10 
FA_IT5  <‐‐‐  FA_IT  0.82  0.73  0.09 
FA_IT6  <‐‐‐  FA_IT  0.80  0.72  0.09 
FA_IT7  <‐‐‐  FA_IT  0.77  0.68  0.09 
MW_NonIT1  <‐‐‐  MW_NonIT  0.82  0.73  0.09 
MW_NonIT2  <‐‐‐  MW_NonIT  0.88  0.79  0.09 
MW_NonIT3  <‐‐‐  MW_NonIT  0.79  0.71  0.08 
MW_NonIT4  <‐‐‐  MW_NonIT  0.83  0.75  0.08 
MW_NonIT5  <‐‐‐  MW_NonIT  0.82  0.74  0.07 
MW_NonIT6  <‐‐‐  MW_NonIT  0.87  0.79  0.08 
MW_NonIT7  <‐‐‐  MW_NonIT  0.90  0.82  0.08 
MW_NonIT8  <‐‐‐  MW_NonIT  0.83  0.76  0.07 
MW_NonIT9  <‐‐‐  MW_NonIT  0.83  0.76  0.08 
FA_NonIT1  <‐‐‐  FA_NonIT  0.70  0.60  0.10 
FA_NonIT2  <‐‐‐  FA_NonIT  0.91  0.82  0.09 
FA_NonIT3  <‐‐‐  FA_NonIT  0.90  0.82  0.09 
FA_NonIT4  <‐‐‐  FA_NonIT  0.74  0.65  0.10 
Sat1  <‐‐‐  Satisfaction  0.95  0.86  0.10 
Sat2  <‐‐‐  Satisfaction  0.95  0.87  0.09 
Sat3  <‐‐‐  Satisfaction  0.88  0.78  0.10 
Sat5  <‐‐‐  Satisfaction  0.80  0.70  0.10 
* Standardized Regression Weights 

   

158 
 
 
 
 
 
 
 
 
 
 
 
 
 
APPENDIX B 

MEASUREMENT ITEMS 

   

159 
 
Constructs/Items  Source 
Mind wandering (non‐IT)   
1. I thought about members of my family.  Sarason et al. (1986) 
2. I thought about friends.  Sarason et al. (1986) 
3. I thought about something that made me feel guilty.  Sarason et al. (1986) 
4. I thought about personal worries.  Sarason et al. (1986) 
5. I thought about something that made me feel angry.  Sarason et al. (1986) 
6. I thought about something that happened earlier today.  Sarason et al. (1986) 
7. I thought about something that happened in the recent past (last few  Sarason et al. (1986) 
days, but not today). 
8. I thought about something that happened in the distant past.  Sarason et al. (1986) 
9. I thought about something that might happen in the future.   Sarason et al. (1986) 
Mind wandering (IT)    
1. I thought about checking my email.  Self‐developed 
2. I thought about using different software. (dropped)  Self‐developed 
3. I thought about checking my social media (e.g., Facebook).  Self‐developed 
4. I thought about browsing other stuff.  Self‐developed 
5. I thought about checking my phone.   Self‐developed 
6. I thought about doing other online activities (e.g., online shopping,  Self‐developed 
online game). 
7. I thought about how complicated the software was. (dropped)  Self‐developed 
Focus of attention (IT)   
1. I focused on how to use the software to complete my task.  Self‐developed 
2. I paid attention on different features of the software to complete my  Self‐developed 
task. 
3. I focused my attention on different options provided by the software to  Self‐developed 
format my task.  
4. I focused my attention on learning specific the software functions.  Self‐developed 
5. I paid close attention to the kind of errors I was making when I was using  Self‐developed 
the software.  
6. I focused my total attention on making sure my computer outputs were  Self‐developed 
correct. 
7. I focused my total attention on following the software’s instructions  Self‐developed 
correctly. 
Focus of attention (non‐IT)   
1.  I focused my total attention on how I could solve the problem.  Kanfer et al. (1994) 
2. I thought about strategies for solving the problem.  Kanfer et al. (1994) 
3. I thought ahead to what I would do next to solve the problem.  Kanfer et al. (1994) 
4. I focused my attention on correctly performing as many steps as I could.  Kanfer et al. (1994) 
Technology anxiety    
1. I became frustrated with my inability to solve the problem using the  Kanfer et al. (1994) 
software. 
(table continues) 

160 
 
(continued) 
Constructs/Items  Source 
2. I got mad at myself when I used the software.  Kanfer et al. (1994) 
3. I was concerned about how poorly I was doing using this software.  Kanfer et al. (1994) 
4. I felt anxious about using this software to complete my task.  Thatcher and Perrewe (2002) 
5. I was worried about how I was going to make a mistake by choosing the  Thatcher and Perrewe (2002) 
wrong options. 
6. I was worried about how I was going to make a mistake that I cannot  Thatcher and Perrewe (2002) 
correct when I used the software. 
User satisfaction   
After doing your assignment using this software, how do you feel about your  Nadkarni and Gupta (2007) 
overall experience with the software? 
1. …I am very dissatisfied / very satisfied. 
2. …I am very displeased / very pleased.  Nadkarni and Gupta (2007) 
3. …I feel absolutely terrible / very delighted.  Nadkarni and Gupta (2007) 
4. …I will strongly recommend this software to my friends.  Nadkarni and Gupta (2007) 
Perceived task complexity    
Seijts et al. (2004); Winter 
1. This task required me to coordinate many different things at the same time. 
and Latham (1996);  
2. This task required me to check one thing before I can move to work on  Winter and Latham (1996) 
something else. (dropped) 
2. This task required a lot of my effort.  Jiang and Punj (2010) 
3. On scale 1 (extremely easy) to 7 (extremely difficult), how difficult was this  Jiang and Punj (2010) 
task? 
Media multitasking  Ophir et al. (2009) 
Did you use any electronic devices/media/software (besides your 
computer/laptop and MS. Access) when you did your assignment? If yes, please 
specify:  
__ Television 
__ Computer based video (such as YouTube or online television episodes) 
__ Music 
__ Non‐music audio 
__ Video/Computer game 
__ Telephone or mobile phones 
__ Instant messaging 
__ Email  
__ Web surfing 
__ Other computer based applications.
Software self‐efficacy   
1. After I did my assignment, I believe I have the ability to export data from  Self‐developed following 
MS. Access to Ms. Word format.  Agarwal et al.’s (2000) 
2. …I believe I have the ability to export data from MS. Access to MS. Excel  procedures.  
format 
(table continues) 
161 
 
(continued) 
3. …I believe I have the ability to export data from MS. Access to PDF format.   
4. …I believe I have the ability to export data from MS. Access to another 
database.  
Self‐regulation (performance evaluation)   
1. I thought about how others would have known better about this software.  Kanfer and Ackerman (1989) 
2. I thought about how I was doing compared to others.  Kanfer and Ackerman (1989) 
3. I thought about others’ ability compare to mine needed to accomplish this  Kanfer and Ackerman (1989) 
task. 
Self‐regulation (self‐reactions)   
1. I thought about how much my knowledge about this software has improved.  Kanfer and Ackerman (1989) 
2. I thought about how well I was doing.  Kanfer and Ackerman (1989) 
3. I thought about how enjoyable the task was.  Kanfer and Ackerman (1989) 
4. I thought about outdoing my previous performance.  Kanfer and Ackerman (1989) 
Self‐regulation (self‐monitoring)   
1. I monitored how well I was learning the software requirements.  Grafill and Compeau (2008) 
2. I asked myself questions to make sure I understood.  Grafill and Compeau (2008) 
3. I made sure I sorted it out as soon as I could before moving on.  Grafill and Compeau (2008) 
4. I tried to monitor where I needed the most improvement.  Grafill and Compeau (2008) 
5. I made up questions to help focus on my learning. (dropped)  Grafill and Compeau (2008) 
6. I monitored my time spent learning skills to determine if it’s worth it.  Grafill and Compeau (2008) 
(dropped) 
General computer self‐efficacy   
1. I was able to use MS. Access if there was no one around me to tell me what  Thatcher and Perrewe (2002) 
to do. 
2. I was able to use MS. Access if I had never used a package like it before.  Thatcher and Perrewe (2002) 
3. I was able to use MS. Access if I had only the software manuals for  Thatcher and Perrewe (2002) 
references. 
Focus of attention (Used in Chapter 4)   
When I was using the website to complete the task,   Nadkarni and Gupta (2007) 
1. …, I had a distinct identifiable purpose. 
2. …, I was able to look up specific information.  Nadkarni and Gupta (2007) 
3. …, I was able to focus on finding specific information.   Nadkarni and Gupta (2007) 
4. …, I was absorbed in finding specific information.  Nadkarni and Gupta (2007) 
Mind wandering (Used in Chapter 4)   
When I was using the website to complete the task,   Mrazek et al. (2013) 
1. …, sometimes I cannot easily pay attention when I should. 
2. …, I found myself always distracted by other things in mind.  McVay and Kane (2009) 
3. …, I had so many things in mind.  Mrazek et al. (2013) 
4. …, my mind wandered.  Mrazek et al. (2013) 
5. …, I was easily distracted by unnecessary information in mind.  Mrazek et al. (2013) 
6. …, sometimes I had trouble concentrating on the tasks.  McVay and Kane (2009) 

162 
 
 
 
 
 
 
 
 
 
 
 
 
 
APPENDIX C 

LEARNING PERFORMANCE MEASURES 

   

163 
 
1. When you export data from MS. Access to MS. Excel, you need to make sure each of the 
following, except:  
(a) Verify my access database is selected in the Navigation Pane 
(b) Specify the destination of my file  
(c) Specify the export option 
(d) Open the destination file after the export is complete  
 
2. What would you do if you only want to export only a subset or a portion of the table?  
(a) Create a query, and then copy it manually to excel 
(b) Filter a table and then export the data 
(c) Remove the unnecessary record and then export the data 
(d) a and b are correct 
 
3. What is the major different between exporting data to Excel and Word format?  
(a) In Excel, I have to make sure the file type is .xlsx or .xls, whereas in Word, I have to 
make sure the file type is .rtf or .doc. 
(b) They are similar, except that I will not be able to analyze my data in Word.   
(c) There is no major different, all the steps are all the same. 
(d) a and b are correct 
 
4. What is an RTF file?  
(a) A word file format introduced by the new version of Microsoft. 
(b) A word format that supports a tabular report.  
(c) A word format that enables documents to be flexibly opened using different 
application.  
(d) A word format that supports special text formatting, such as bold, italic, and 
underline.  
 
5. What types of Access format can be exported to PDF?  
(a) Only tables and queries 
(b) Only tables and forms 
(c) Tables, queries, and forms 
(d) Only forms, reports, and limited tables.   
 
6. What is the initial step you need to do when you want to export your access data to 
another access database? 
(a) Make sure my query is opened in Design view 
(b) Make sure I create a new Table in the Query Type group 
(c) Make sure I combine two databases in Design view 
(d) Make sure the categories in my databases belong to the same tables 
 
7. Why would you decide to export the definition of a table to a different database, but 
not the data? 
(a) Because you generally only need the table format for a new database.  

164 
 
(b) Because you don’t want to have an overlapping data. 
(c) Because data from the original table may not less useful for the new purpose.  
(d) Because if you export the data, you will not be able to edit the table.  
 
8. Which document formats can be edited once you export your data from Access? 
(a) Word and XPS 
(b) Word and PDF 
(c) Word and Excel 
(d) Excel and PDF 
 
9. What advantage of exporting your Access data to Excel?  
(a) There is no advantage at all.  
(b) For easy reporting.  
(c) For easy editing. 
(d) For easy reporting and editing 
 
10. Which type of objects can be exported from one database to another?  
(a) Table only 
(b) Tables and queries 
(c) Reports only 
(d) All objects 
   

165 
 
 
 
 
 
 
 
 
 
 
 
 
APPENDIX D 

EXECUTIVE FUNCTIONS TASKS 

 
   

   

166 
 
  Executive Functions were measured using three cognitive tasks (i.e., color shape task, 

Stroop task, and N‐Back task). All three tasks were administered in Inquisit Web Version 4 

(Millisecond Software LLC, Seattle). The experiment materials are available on the Millisecond 

Test Library (http://www.millisecond.com/download/library/) and customized to fit the needs 

of this dissertation. 

Color Shape Task 

The version of color shape task in this study followed a similar procedure used in Miyake 

et al. (2004). Before starting the target trials, participants completed a total of 80 training trials, 

administered in five blocks of 16 trials each, to familiarize themselves with the response‐

mapping rules. The responses for both color and shape tasks were mapped onto two keyboard‐

selection rules in a bivalent manner (e.g., the left button for “red” and “circle” and the right 

button for “green” and “triangle”). The response‐mapping rules were kept constant for all 

participants in order to maintain consistency. In each training trial, a cue letter (C for color and 

S for shape) appeared above a colored rectangle with a shape in it. Participants were instructed 

to indicate whether the color was red or green when the cue was C, and whether the shape was 

a circle or triangle when the cue was S. If the response was correct, the next trial appeared 

after a 600 ms intertribal interval (ITI). If it was incorrect, the computer beeped and error 

message (i.e., “incorrect”) was displayed for 1000 ms followed by an ITI of 600 ms (Miyake et 

al., 2004).  

In the first three blocks of the training trials, the stimulus was univalent and was 

applicable to only one task—either a color or shape task. In the first block, the shape task (i.e., a 

167 
 
black line‐drawing of either a circle or square shape) was presented on each trial, whereas in 

the second block, the color task (i.e., a square‐shaped color patch of either red or green color) 

was presented. In the third block, a shape or a color patch was presented in a random order 

(e.g., a black line drawing of a square on one trial, followed by a red color patch on the next). In 

the last two blocks of the training session, the stimulus was bivalent and consisted of a shape 

(either circle or triangle) superimposed on a square color patch (either red or green). 

Participants performed the shape task throughout the fourth block and the color task for the 

last, fifth block.  

After completing the training trials, participants performed two blocks of target pure‐

task trials with 64 trials in each block. The trial set‐up was identical to the training trials.  

However, participants no longer received an error message for incorrect responses but were 

informed of their accuracy (percentage correct) at the end of each block. After completing the 

pure‐task trials, participants performed blocks of mixed‐task trials. In these trials, the cues were 

the combination of both color and shape tasks (C for color and S for shape). To familiarize 

participants with the procedure, they were first asked to practice mixed‐task trials for two 

blocks consisting of 16 trials in each block. The stimulus was the same as in the target pure‐task 

trials and consisted of a shape (circle or triangle) superimposed on a square‐shaped color patch 

(red or green) presented in the middle of the screen. The cue‐to‐stimulus interval (CSI) was 200 

ms. The cue and stimulus remained on‐screen until responses from participants were received. 

During this practice phase, participants were given immediate feedback for incorrect responses 

using the same procedure as in the pure‐task trials. After practicing the mixed‐task trials, 

participants performed two blocks consisting of 50 target mixed‐task trials in each block. The 

168 
 
trial setup for these mixed‐task trials was identical to the practice trials except that feedback 

was no longer provided for incorrect responses (Miyake et al., 2004).  

Stroop Task 

  On each trial of the Stroop task, participants saw words presented in one of four colors: 

red, blue, green, or black. Participants are instructed to name the color of each stimulus while 

ignoring what the words actually say as quickly and as accurately as possible. The response 

rules were kept constant for all participants (i.e., d for red words, f for green words, j for blue 

words, and k for black words). If an incorrect response was made, a red x was flashed on the 

screen. In this experiment, participants were given three types of trials: (a) incongruent colors 

condition—30 trials with a color word printed in a different color (e.g., BLUE printed in red 

color); (b) congruent colors condition—30 trials with a color word printed in the same color 

(BLUE in blue color); and (c) control condition—30 trials with rectangles printed in one of four 

colors (red, green, blue, or black). Participants also received three short blocks of approximately 

6 trials apiece for practice.  

N‐Back Task 

  This task followed the same procedure used in Jaeggi et al. (2010). In this task, 

participants were shown a sequence of visual stimuli and they had to respond each time the 

current stimulus was identical to the one presented n positions back in the sequence. The 

stimulus material consisted of 8 random shapes. The shapes were all shown in yellow and 

presented centrally on a black background for 500 ms each, followed by a 2500 ms inter‐

169 
 
stimulus interval. Participants were required to press a pre‐defined key for targets, and their 

response window lasted from the onset of the stimulus until the presentation of the next 

stimulus (3000 ms). No response was required for non‐targets. Participants completed a total of 

9 blocks (2‐, 3‐, and 4‐back levels, with each level presented for 3 consecutive blocks). 

Participants completed 10 trials of practice session per level n tested and then three blocks of 

each n‐back task—with each block consisted of 20 trials, including 6 target and 14 non‐target 

trials. Omissions were counted as errors.  

   

170 
 
 
 
 
 
 
 
 
 
 
 
 
APPENDIX E 

SAMPLE EXPERIMENT MATERIALS 

 
   

171 
 
Instructions for Experiment7 

Please read this following information carefully.  
1. Your primary task is to do an online shopping search, and then make a decision based on 
the criteria provided by the researchers.  
2. The excel file on your computer provides the criteria you need to make your final 
decision.  
3. Please save your response regularly.  
4. Please let the researcher know if you have any questions.   
5. Now, you are about to begin your task. Your completion time will be automatically 
monitored. Click ‘START’ once you are ready to begin and click ‘STOP’ once you finished 
making your decision.  
 
Scenario8 

  Your birthday is just around the corner. Your parents want to buy you a new laptop. So, 

you think of doing some research about laptops/computers. Because you are so excited about 

it, you go online to search for information about laptop, including the features, specifications, 

and the price. So, you turn on your computer, open internet explorer, and go to 

www.amazon.com. You particularly are interested in five different laptops (see the list below). 

But, you don’t know which one is the best. In order to help you make the best decision, you 

asked for help from your introduction for information systems’ instructor and he provided you 

with a scoring sheet in Excel where you can give a point score based on 15 criteria for each 

laptop.  

Note: Please copy the product name and paste it in the website’s search box.    
 
A.  MacBook Pro MD101LL/A (750 Storage) 
                                                       
7
 The instruction was printed and was also conveyed verbally before the experiment.  
8
 The scenario shown here was for the high complexity condition. For the low complexity condition, only the first 
five products were presented and for the medium complexity condition, only the first ten products were 
presented. 

172 
 
B. Asus X202E‐DB21T  
C. HP Envy 14‐3010NR Spectre  
D. Samsung Series 5 550 (Wi‐Fi)  
E. MacBook Air MD711LL/A  
 
  You talk to your friend, Amanda, about it and she suggests you to take a look at another 

website (www.bestbuy.com). She believes you can find a good laptop on that website. She also 

gives you a list of laptops you may be interested in.  

F. Acer E14314404 
G. Toshiba C75DA7310 
H. Asus Q550LF‐BBI7T07 
I. Dell I7437T‐2509SLV  
J. HP Envy 15‐J175NR 
 
  You tell your parents that you get confused because you have so many options. Because 

your dad works as a Manager at the Staples, he thinks you should check out some laptop on his 

company’s website (www.staples.com). He gives you some recommendations to look up.  

K. Lenovo G710 ‐ 59395518 
L. Gateway NV570P07u 
M. HP Pavilion 15‐N020US 
N. Samsung NP600B4C‐A01US 
O. Acer NX.V8WAA.001 
 
  After you calculated the total point of each laptop, you decided to buy the product with 

the highest total points. Please enter your product selection below: 

___________________________________________  
Total Completion Time: _______ (Minutes, Seconds) 

You have just finished your tasks. Do not forget to save your Excel file. Thank you for your 

participation.  

173 
 
Table E.1 

Excel Scoring Sheet 

Product 
Criteria  Guidelines  A  B  …  O 
Operating  Guidelines:          
System (OS)  1 point if Chrome OS 
2 points if Windows 7                         
3 points if Windows 8 
4 points if Mac OS 
Laptop Size  Guidelines (rounded to the closest decimal):         
No point if >16" 
3 points if 14‐16"  
5 points if 11‐13" 
Memory and  Guidelines:         
Battery Life  No point if battery life < 8 hours; memory < 4 GB  
2 point if battery life < 8 hours; memory > 4 GB  
4 points if  > 8 hours; memory > 4 GB    
Price (New Item)  Guidelines:         
No point if > US$1,000 
1 points if US$1,000‐750 
2 points if US$750‐500 
3 points if < US$500 
Harddrive  Guidelines:         
No point if < 128 GB 
2 points if 128‐256 GB 
4 points if > 256 GB 
Processor  Guidelines:         
No point if Intel Celeron/Atom/Core 2 Duo/AMD 
1 point if Intel Pentium 
2 points if Intel core i3 
3 points if Intel core i5 
4 points if intel core i7 
Item Weight  Guidelines:         
No point if > 7 pounds 
2 point if 6.99‐5 pounds  
4 points if 4.99‐3 pounds 
5 points if < 3 pounds 
Wi‐Fi and  Guidelines:         
Bluetooth  No point if neither are available 
1 point if Wi‐Fi OR Bluetooth is available 
2 points if both Wi‐Fi and Bluetooth are available 
(table continues) 

174 
 
Table E.1 (continued) 

Product 
Criteria  Guidelines  A  B  …  O 
Screen  Guidelines:         
Resolution  No point if less than 1024 x 768 
1 point if between 1152 x 768 to 1280 x 854 
2 points if greater than 1280 x 854 or 1366 x 768  

Number of USB 3  Guidelines:         
Port  No point if none 
1 point if 1  
2 point if 2 or more 
TOTAL  Guidelines: The total point of each product is         
automatically computed using the SUM Function in 
Excel.  
 
 

   

175 
 
 

APPENDIX F 

INFORMED CONSENT NOTICE 

   

176 
 
177 
 
 

178 
 
 

179 
 
 

180 
 
REFERENCES FOR APPENDICES 

Agarwal, R., Sambamurthy, V., & Stair, R. M. (2000). The evolving relationship between general 
and specific computer self‐efficacy‐an empirical assessment. Information Systems 
Research, 11(4), 418‐430. 
Gravill, J., & Compeau, D. (2008). Self‐regulated learning strategies and software training. 
Information & Management, 45(5), 288‐296. 
Inquisit (Version 4.0.3) [Computer software]. (2014). Seattle, WA: Millisecond Software LLC. 
Jaeggi, S. M., Studer‐Luethi, B., Buschkuehl, M., Su, Y. F., Jonides, J., & Perrig, W. J. (2010). The 
relationship between N‐back performance and matrix reasoning‐implications for training 
and transfer. Intelligence, 38(6), 625‐635. 
Jiang, Y., & Punj, G. N. (2010). The effects of attribute concreteness and prominence on 
selective processing, choice, and search experience. Journal of the Academy of Marketing 
Science, 38(4), 471‐489. 
Kanfer, R., & Ackerman, P. L. (1989). Motivation and cognitive abilities: An integrative/aptitude‐
treatment interaction approach to skill acquisition. Journal of Applied Psychology, 74(4), 
657‐690. 
Kanfer, R., Ackerman, P. L., Murtha, T. C., Dugdale, B., & Nelson, L. (1994). Goal setting, 
conditions of practice, and task performance: A resource allocation perspective. Journal of 
Applied Psychology, 79(6), 826‐835. 
McVay, J. C., & Kane, M. J. (2009). Conducting the train of thought: Working memory capacity, 
goal neglect, and mind wandering in an executive‐control task. Journal of Experimental 
Psychology: Learning, Memory, and Cognition, 35(1), 196‐204. 
Miyake, A., Emerson, M. J., Padilla, F., & Ahn, J. C. (2004). Inner speech as a retrieval aid for task 
goals: The effects of cue type and articulatory suppression in the random task cuing 
paradigm. Acta Psychologica, 115(2), 123‐142. 
Mrazek, M. D., Phillips, D. T., Franklin, M. S., Broadway, J. M., & Schooler, J. W. (2013). Young 
and restless: Validation of the mind wandering questionnaire (MWQ) reveals disruptive 
impact of mind wandering for youth. Frontiers in Psychology, 4, 1‐7. 
Nadkarni, S., & Gupta, R. (2007). A task‐based model of perceived website complexity. MIS 
Quarterly, 31(3), 501‐524. 
Ophir, E., Nass, C., & Wagner, A. D. (2009). Cognitive control in media multitaskers. Proceedings 
of the National Academy of Sciences of the United States of America, 106(37), 15583‐
15587. 
Sarason, I. G., Sarason, B. R., Keefe, D. E., Hayes, B. E., & Shearin, E. N. (1986). Cognitive 
interference: Situational determinants and traitlike characteristics. Journal of Personality 
and Social Psychology, 51(1), 215‐226. 

181 
 
Seijts, G. H., Latham, G. P., Tasa, K., & Latham, B. W. (2004). Goal setting and goal orientation: 
An integration of two different yet related literatures. Academy of Management Journal, 
47(2), 227‐239. 
Thatcher, J. B., & Perrewe, P. L. (2002). An empirical examination of individual traits as 
antecedents to computer anxiety and computer self‐efficacy. MIS Quarterly, 26(4), 381‐
396. 
Winters, D., & Latham, G. P. (1996). The effect of learning versus outcome goals on a simple 
versus a complex task. Group & Organization, 21(2), 236‐250. 
 

182 
 

You might also like