Engagement versus Motivation:

A Confirmatory Factor Analysis of the Motivation and Engagement Wheel

A DISSERTATION
SUBMITTED TO THE FACULTY OF THE
UNIVERSITY OF MINNESOTA
BY

Alyssa A. Schardt

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

Faith G. Miller, Adviser

May 2019
©Alyssa A. Schardt, 2019
 i

Acknowledgements

Graduate school, and the dissertation process in particular, has been arduous and

only possible because of the support and guidance I have received from so many

individuals. First, I would like to whole-heartedly thank Dr. Faith G. Miller, my academic

advisor, whose guidance, encouragement, and patience throughout my graduate school

experience has helped shape who I am professionally. I am forever grateful she endorsed

the vision I had for my research line, and encouraged my curiosity. I also wish to

acknowledge the continued support of my committee chair, Dr. Amanda Sullivan, who

has pushed me to deepen my understanding and appreciation for my research. I would

also like to thank my committee members, Drs. Timothy Piehler and Annie Hansen-

Burke. Dr. Piehler’s input on my methodology and analyses has been invaluable, and Dr.

Hansen-Burke has acted as a mentor and provided unyielding support throughout

graduate school. Dr. Hansen-Burke’s words of wisdom and dedication to her students’

completion continues to inspire me. In addition, I would like to thank Dr. Scott

McConnell, who modeled compassion as a supervisor and encouraged me to follow my

own path.

I would also like to acknowledge my cohort-mates who have become my close

friends over the past five years. Our mutual support for one another created an

environment that allowed us to flourish and find joy in stressful times (and set a high

expectation for applause). I would also like to formally thank Jordan Thayer who has

been key to my accomplishments throughout graduate school. Whether recommending

me for graduate assistantships or allowing me to take over his personal computer so I

could run my analyses, I am genuinely lucky to have him in my corner. I am also

 ii

fortunate for my family and their continued support. Specifically, I want to express my

appreciation for my sister, Lauren Mann, for always taking the time to put a smile on my

face (even from across the country), and my parents, Paul Schardt and Donna Brenner

Schardt, who could not have been more encouraging throughout every step of my

journey.

Lastly, I gratefully recognize the time and involvement from the students and

teachers who participated in this study. I can only hope I continue to work with educators

as dedicated to understanding and developing best practices throughout the rest of my

career.
 iii

Abstract


Both engagement and motivation have been identified as constructs that are critical to

student success and are linked with later academic achievement. However, the multitude

of conceptualizations around these constructs and how they relate to one another has

become a point of contention within the field. The primary aim of the current study is to

examine a model of academic engagement and motivation, the motivation and

engagement wheel (MEW), using the Motivation and Engagement Scale (MES; Martin,

2009; Martin, Ginns, & Papworth, 2017) and replicating the model structure with a more

diverse, elementary-aged American population. Secondarily, the current study aimed to

expand the MEW by examining an adapted model structure that included alternate

components of motivation and engagement as measured by the Engagement Versus

Disaffection with Learning-teacher and student reports and Patterns of Adaptive Learning

Scale. Participants included 270 students in 3rd - 6th grade (predominately African

American, 67.4%), from an urban area in the Midwestern United States. Confirmatory

factor analysis (CFA) techniques were used to compare the higher order structure across

models. As hypothesized, the four-factor higher order models, comprised of adaptive

engagement, adaptive motivation, maladaptive engagement, and maladaptive motivation,

which best align with the MEW, demonstrated best fit across both the replication and

adaptability models. Thus, this study provided additional support for the structure of the

MEW, and preliminary evidence for its adaptability as a theoretical model.

Keywords: adaptive engagement, adaptive motivation, maladaptive engagement,

maladaptive motivation
 iv

Table of Contents

Acknowledgements .................................................................................................. i

Abstract
 ................................................................................................................ iii

List of Tables .......................................................................................................... v

List of Figures ....................................................................................................... vii

List of Appendices ............................................................................................... viii

Chapter 1: Introduction ........................................................................................... 1

Chapter 2: Literature Review .................................................................................. 8

Chapter 3: Methods ............................................................................................... 30

Chapter 4: Results ................................................................................................. 41

Chapter 5: Discussion ........................................................................................... 46

References ............................................................................................................. 56
 v

List of Tables

Table 1: Demographic Information for the Sample .............................................. 63

Table 2: Comparing Means and Cronbach’s alphas for Subscales of MES Across

Samples ..................................................................................................... 64

Table 3: Descriptive Statistics for Subscales of MES, EvsD, and PALS .............. 65

Table 4: Descriptive Statistics for MES Items ...................................................... 66

Table 5: Descriptive Statistics for Student EvsD Items ........................................ 68

Table 6: Descriptive Statistics for PALS Items ..................................................... 69

Table 7: Descriptive Statistics for Teacher EvsD Items ....................................... 71

Table 8: Pearson’s Correlation Coefficients for Subfactors of Adaptive

Motivation ................................................................................................. 72

Table 9: Pearson Correlation Coefficients for Subfactors of Adaptive Engagement

................................................................................................................... 73

Table 10: Pearson Correlation Coefficients for Subfactors of Maladaptive

Motivation ................................................................................................. 74

Table 11: Pearson’s Correlation Coefficients for Subfactors of Maladaptive

Engagement............................................................................................... 75

Table 12: Standardized Item Factor Loadings and Errors for Model 4 ............... 76

Table 13: Model 4 First-Order Standardized Factor Loadings and Errors on

Higher Order Factors ............................................................................... 78

Table 14: Standardized Item Factor Loadings and Errors for Model 8 ............... 79
 vi

Table 15: Model 8 First-Order Standardized Factor Loadings and Errors on

Higher Order Factors ............................................................................... 83

 vii

List of Figures

Figure 1. Motivation and Engagement Wheel (reproduced from Martin et al.,

2007). ........................................................................................................ 84

Figure 2. Higher-order single-factor replication model of motivation/engagement.

................................................................................................................... 85

Figure 3. Higher-order two-factor replication model of adaptive and maladaptive

traits........................................................................................................... 86

Figure 4. Higher-order two-factor replication model of motivation and

engagement. .............................................................................................. 87

Figure 5. Higher-order four-factor model of adaptive and maladaptive motivation

and engagement. ....................................................................................... 88

Figure 6. Higher-order single-factor model of motivation/engagement. ............. 89

Figure 7. Higher-order two-factor model of adaptive and maladaptive traits. ..... 90

Figure 8. Higher-order two-factor model of motivation and engagement. .......... 91

Figure 9. Higher-order four-factor model of adaptive and maladaptive motivation

and engagement. ....................................................................................... 92

Figure 10. Four-factor replication model (Model 4) with standardized factor

loadings ..................................................................................................... 93

Figure 11. Four-factor adaptability model (Model 8) with standardized factor

loadings ..................................................................................................... 94
 viii

List of Appendices

Appendix A: Theoretical alignment for subscale categorization ......................... 97

Appendix B: School-based professional demographic sheet ................................ 99

Appendix C: Motivation and Engagement Scale – Junior School Items………...98

Appendix D: Engagement Versus Disaffection with Learning: Student Report

Items ........................................................................................................ 103

Appendix E: Engagement Versus Disaffection with Learning: Teacher Report

Items ........................................................................................................ 105

Appendix F: Patterns of Adaptive Learning Scales: Student Survey Items ....... 107
 1

CHAPTER 1

Introduction

Engagement and motivation in school make it possible for students to learn in the

classroom, and therefore are foundational to academic success (e.g., Greenwood, Terry,

Marquis, & Walker, 1994). To this end, both motivation and engagement have been

identified as academic enablers (DiPerna & Elliott, 2002). While variation exists in how

these constructs are defined, generally motivation is conceptualized as an internal drive

or desire to make an attempt or complete a task or behavior, while engagement is

conceptualized as attention and actions aligned to expected behaviors. More specifically,

engagement aligns more to actions while motivation aligns more to thoughts. Throughout

educational literature both constructs have been identified as critical components of

academic achievement (e.g., Finn & Rock, 1997). Academic engagement is positively

correlated with standardized test scores and grades (e.g., Marks, 2000; Singh, Granville,

& Dika, 2002), and has been established as a protective factor against school risk factors

(Finn & Rock, 1997). Alternatively, a lack of engagement has been associated with

school dropout (Archambault, Janosz, Morizot, & Pagani, 2009). This is true for students

with and without exceptionality; if engagement is low, students are less likely to

complete high school (Reschly & Christenson, 2006). Additionally, early difficulties with

academic engagement have been linked to a long-term negative impact on grades and test

scores (Alexander, Entwisle, & Horsey, 1997).

From teachers’ perspectives, motivating students has long been perceived as a

major problem across all grade levels, regardless of experience levels of the teachers

(Veenman, 1984). In other words, a lack of motivation is a pervasive academic concern.

 2

Students’ engagement is also a concern for teachers; when a child is not academically

engaged it not only contributes to their own lack of learning, but also has the potential to

negatively impact the teacher and other students in the classroom. According to one

survey, 36% of teachers reported losing more than two hours of instruction time per week

due to behavioral concerns (Walker, Ramsey, & Gresham, 2003). Findings were even

more drastic for teachers in urban areas—Walker and colleagues reported 21% of

teachers in urban settings lost more than four hours per week because of behavioral

redirection. According to Archambault et al. (2009), children who present low

engagement by middle school should be considered a high-priority for targeted

intervention. Consequently, preventative interventions that focus on increasing

motivation and engagement could begin as early as elementary school.

Over the past several decades, the impact engagement and motivation have on

academic success has become well-established in the research literature, with the benefits

universally recognized across varied theoretical perspectives (e.g., Christenson, Reschly,

& Wylie, 2012; DiPerna & Elliott, 2002; Finn & Rock, 1997). Indeed, to date over eight

theories have been developed to help explain the phenomena of motivation and

engagement. While diverse perspectives do exist, a primary theme throughout each

theory is that engagement and motivation are generally considered to be closely related

constructs with a spectrum of inhibitors and facilitators. However, close examination of

this body of research also reveals inconsistencies with how motivation and engagement

are conceptualized, measured, and defined. Thus, while there is agreement on the concept

of motivation and engagement broadly and their academic benefits, there is still

disagreement in the field regarding operationalization of these constructs and how they
 3

relate to one another. From some theoretical perspectives, engagement and motivation

have been conceptualized as interchangeable and measured as a single construct rather

than two separate factors. This was the case for the original motivation and engagement

wheel, a model of engagement and motivation established by Martin (2007), which has

since been revised. However, many researchers posit engagement and motivation are

closely related, but separate constructs—a more prominent and current perspective in the

field (e.g., DiPerna, Volpe, & Elliott, 2002; Martin, Ginns, & Papworth, 2017; Reschly &

Christenson, 2012). Given these differing perspectives, the conceptual and theoretical

relationship between engagement and motivation has become a point of contention in the

field (Reschly & Christenson, 2012). Consequently, the current state of the science lacks

coherence and consensus. The primary aim of this research study was to add clarity to

conceptualization of adaptive and maladaptive engagement and motivation, and better

understand the relationship between these broad constructs. These perspectives were

examined in the current study through comparing model fit for models that represent a

single combined engagement and motivation latent construct and other models that

separated engagement and motivation into multiple latent constructs, replicating models

associated with the updated version of the motivation engagement wheel (MEW; Martin

et al., 2017).

Recently, Martin and colleagues (2017) proposed an updated model for the MEW

in which motivation and engagement are conceptualized as distinct constructs that

interact in a cyclical fashion. From their perspective, motivation is defined as “the

inclination, energy, emotion, and drive relevant to learning, working effectively, and

achieving,” while engagement is defined as “the behaviors that reflect this inclination,
 4

energy, emotion, and drive” (Martin et al., 2017, p. 150). In this most recent study,

Martin and colleagues (2017) used confirmatory factor analysis to compare four higher-

order models of engagement and motivation: a single-factor higher-order factor in which

engagement and motivation were a single construct, a two-factor higher-order model that

separated engagement and motivation, a two-factor higher-order model that separated

adaptive and maladaptive components, and a four-factor higher-order model made up of

adaptive engagement, adaptive motivation, maladaptive engagement, and maladaptive

motivation. Fit indices indicated that the four-factor model had the best fit. Thus, within

this updated model, engagement is broken into adaptive and maladaptive behaviors, and

motivation is also separated into adaptive and maladaptive cognitions (see Figure 1).

While this model holds promise, several gaps and limitations exist that require further

study.

Rationale

The MEW model was chosen for replication over other models of engagement

and motivation for a variety of reasons. First, the MEW holds promise because it posits a

bi-directional relationship between engagement and motivation, accepting different

aspects from a variety of conceptualizations. While some models propose that motivation

drives engagement (i.e., DiPerna & Elliott, 2002) and others suggest that engagement

influences motivation (i.e., Reeve & Lee, 2014), the MEW posits a cyclical relationship

demonstrating a broader conceptualization of engagement and motivation that

encompasses multiple theoretical perspectives (a more detailed review of different

theoretical perspectives and models of engagement and motivation can be found in

Chapter 2). Another reason the MEW was chosen in the current study is the explicit
 5

inclusion of both positive and negative components of engagement and motivation within

the model. While inhibitors and facilitators of these constructs are universally accepted

and measured, the MEW provides a parsimonious conceptualization and is the only

model that parses apart adaptive and maladaptive factors into their own constructs. In

other words, most models will measure both positive and negative components on the

same scale, but Martin’s model provides additional information by separating adaptive

and maladaptive components into different constructs. For instance, measuring

maladaptive components of engagement (e.g., self-handicapping) does not necessarily

provide information about adaptive engagement (e.g., task management). Thus, more

information is gained in separating into constructs. This conceptualization also holds

particular promise for intervention; by differentiating between adaptive and maladaptive

cognitions and behaviors, interventions could be better targeted to address the underlying

problem. For example, maladaptive cognitions could be targeted through cognitive

behavioral therapy, while maladaptive behaviors may be better served through a self-

monitoring intervention. Therefore, the current study aimed to replicate Martin and

colleagues’ (2017) MEW model and assess the adaptability of their model using

additional measures of engagement and motivation.

While Martin and colleagues’ (2017) MEW is a well-known model of academic

engagement and motivation, there are number of limitations associated with their

research. First, neither the original nor the updated model has been replicated outside of

the original research group despite being widely accepted and used over the past 10

years. Replication is particularly essential with confirmatory factor analysis (CFA) due to

the influence of sample on factor analytic techniques (Tabachnick & Fidell, 2013).
 6

Additionally, the model relies solely on the Motivation and Engagement Scale (MES)

measure, which is potentially problematic because MES was developed by Martin’s

research team. Models can be strengthened by utilizing multiple measures to represent

latent variables. Maruyama (1998) suggested using multiple measures for each construct

is the only defensible way in which to create viable models. Additionally, utilizing

measures that are not directly derived from the MEW will permit further validation of the

model, allowing for the possibility to better understand how the MEW bridges multiple

conceptualizations of engagement and motivation. Lastly, Martin and colleagues’ (2017)

study was conducted with a relatively homogenous Australian population, within a

secondary educational setting, which impacts the generalizability of their findings to

students in the United States. In addition, although the MES was developed and has been

validated for students nine years of age and older, the higher order structure of the MEW

has only been tested with a secondary sample. Thus, there is a need to further examine

this relationship between engagement and motivation utilizing a more diverse,

elementary aged population.

Purpose. The primary purpose of this study was to examine the relationship

between adaptive and maladaptive engagement and motivation by replicating the models

presented by Martin et al. (2017) with a more diverse, and elementary aged American

sample to confirm their four-factor higher-order model. Secondarily, this study aimed to

capture additional components that make up these constructs beyond those included in

the MES. More specifically, by including additional measures of motivation and

engagement that include components of these constructs beyond those currently assessed

in the MEW, it may strengthen our conceptualization of positive and negative

 7

engagement and motivation beyond what is captured by the measure created by Martin

and colleagues.

Research questions. Based on prior research and theory and using the MEW as

the conceptual base and the MES, Engagement Versus Disaffection with Learning-

teacher and student reports (EvsD), and Patterns of Adaptive Learning Scale (PALS) as

the measurement base, this study aimed to answer the following research questions:

1. Which higher-order structure, as replicated from models proposed by

Martin and colleagues (2017), best fits an American, elementary aged

population as compared to findings of a four-factor structure for the MEW

from an Australian high school population (Martin et al., 2017) as

assessed through CFA of students’ MES scores?

2. Do alternate measures of adaptive and maladaptive motivation and

engagement uphold the higher-order four-factor structure of the MEW

(e.g., the adaptive and maladaptive engagement and motivation quadrants)

as determined by CFA?

Based on findings from Martin and colleagues, it was expected that the four-

factor structure of the MEW would also fit a more diverse, American, elementary aged

population. Additionally, we anticipated that alternative measures of engagement and

motivation, that capture additional components of these constructs (e.g., interest and on-

task behavior; see Appendix A), would fit within the adaptive motivation, adaptive

engagement, maladaptive engagement, and maladaptive motivation quadrants represented

in Martin’s (2017) updated MEW, thus upholding the four-factor higher order structure

of the model.
 8

CHAPTER 2

Literature Review

Central to understanding the research literature on engagement and motivation

involves understanding the theoretical perspectives and definitions underlying these

constructs. Prior to reviewing these prominent theories and corresponding definitions, it

is important to first contextualize the intricacy and complexity of doing so. For example,

engagement can be conceptualized as unidimensional (e.g., on-task behaviors; Fisher &

Berliner, 1985) or, more commonly, multidimensional (e.g., including internal and

external components of engagement; Appleton, Christenson, Kim, & Reschly, 2006;

Fredricks, Blumenfeld, & Paris, 2004). Within this multidimensional conceptualization of

engagement, differences even exist with regard to the number and nature of proposed

components. While the majority of engagement scholars align with a three-part

conceptualization (i.e., behavioral, cognitive, and emotional engagement; Christenson et

al., 2012; Fredricks et al., 2004), the operationalization of the individual subtypes of

engagement varies across researchers.

To further complicate matters, the term motivation is often not defined (e.g.,

Christenson & Anderson, 2002) or defined subjectively (e.g., “to be moved to do

something;” Ryan & Deci, 2000, p.54). It is generally agreed amongst researchers that

motivation is an internal or private construct (Reeve, 2012). However, different

theoretical orientations are associated with distinct conceptualizations surrounding the

underlying components of motivation (e.g., self-efficacy, goal setting, mastery

orientation). To add to the confusion, similar terms are used to describe vastly different

concepts and vastly different terms are used to describe similar concepts for both
 9

engagement and motivation. For instance, both affective engagement and emotional

engagement are used interchangeably throughout the literature. Conversely, behavioral

engagement is often defined in terms of on-task behaviors, but also used to describe

attendance, suspensions, and being on time depending on the researcher (e.g.,

Christenson et al., 2012).

While engagement and motivation are conceptualized in uni- and multi-

dimensional formats across educational literature (Reschly & Christenson, 2012),

assumptions regarding the existence of a relationship between these constructs remains

consistent. However, the conceptualization of this relationship differs across theoretical

perspectives and how engagement and motivation are defined. Depending on the

theoretical perspective, engagement and motivation have been conceptualized as an

interchangeable single construct (e.g., Martin, 2007; Skinner & Belmont, 1993) or closely

related, but separate constructs (e.g., DiPerna et al., 2002; Martin et al., 2017; Reschly &

Christenson, 2012). Given the influence engagement and motivation have on academic

achievement, it is of particular interest to better understand these constructs and how they

relate to one another from both theoretical and applied perspectives. This chapter outlines

prominent theories of engagement and motivation, discusses the impact these theories

have on the conceptualization of engagement and motivation, and describes how they

relate to one another in order to synthesize findings from theoretical literature and

increase conceptual clarity of these constructs to better inform research and practice. For

the purpose of the current study, theories were operationalized as underlying ideas used

to explain a phenomenon. Models, on the other hand, were conceptualized as

mathematical representations based on theory (Grüne-Yanoff, 2013). There are a number

 10

of key differences between theories and models: a) theory is a generalized explanation,

while models are specific to populations and measures utilized; b) theories are abstract

conceptualizations, while models are statistically driven; and c) theories are based on

observed phenomena, while models are based on experimentation and mathematical

manipulation.

Prominent Theories of Engagement and Motivation

Achievement Goal Theory

According to achievement goal theory (AGT) there are multiple contrasting

motivational processes, which underlie behaviors in achievement-type activities. Thus,

the basis of AGT focuses mainly on the construct of motivation without explicitly

defining engagement. Specifically, one’s approach and engagement in an achievement

activity, concept of self, and the task outcome determine if success is of intrinsic or

extrinsic value (Ames, 1992). Across researchers, the name for the opposing processes

differ; however, the theory remains aligned to orienting towards internal gains, such as

acquiring knowledge, versus orienting towards external gains, such as grades or money.

Maehr and Nicholls (1980) refer to these processes as task-involvement versus ego-

involvement goals, while Ames and Archer (1988) utilized the terminology of mastery

versus performance goals. Similarly, Elliott and Dweck (1988) preferred the terms

learning versus performance goals. In other words, task-involvement, learning, and

mastery goals all refer to success as determined by internal gains. Ego-involved and

performance goals, on the other hand, refer to success as determined through external

markers.
 11

While AGT does not explicitly discuss engagement, there is a focus on motivation

in which “active engagement” is an essential component. Within this theory, active

engagement refers to the application of learning and problem-solving strategies, with an

underlying belief that hard work leads to success, and failure can be remedied by effort

(Ames, 1992). Motivation is not overtly defined within this theory. Instead, AGT

provides the framework that informs motivational patterns (Ames, 1992). Specifically,

mastery goals are aligned to a motivation to learn, which is associated with a higher

quality of participation that will likely sustain across settings (Ames, 1992; Brophy,

1983). Consequently, mastery orientation is linked with risk-taking behaviors and

persistence throughout achievement activities (Elliott & Dweck, 1988; Meece,

Blumenfeld, & Hoyle, 1988). Conversely, performance orientation is associated with

ability as determined through surpassing peers or norm-based standards. Through a

performance goal lens, achievement with little effort is perceived as a greater success

than achievement through hard work (Ames, 1992). Additionally, performance

orientation leads to a motivational pattern that is associated with avoiding failure. Thus,

performance oriented individuals are more likely to take on less challenging tasks in

order to maintain success through preventing failure (Ames, 1992).

While sometimes categorized as its own theory (e.g., Martin et al., 2017),

subsequent researchers have expanded on AGT’s perspective of goal orientation to

include perceptions of success and failure within mastery and performance goal

orientations. More specifically, goal orientations can be categorized as success-oriented

compared to failure-avoidant (Martin & Marsh, 2003). Success-oriented students are

often characterized as proactively oriented towards their studies. Failure-avoidant

 12

orientation, on the other hand, is associated with self-handicapping or procrastination in

case of poor performance. It is important to note that neither success-oriented nor failure-

avoidant orientations is associated with a specific goal-type, but rather both mastery and

performance goals can be split into success- or failure avoidant-orientations.

When AGT was originally conceptualized, goal orientations were often thought of

as more stable traits rather than a changeable perspective within an individual. However,

as the theory developed, researchers began proposing the malleability of these

orientations. According to Dweck (2007), some individuals view intelligence as fixed

while others perceive it to be malleable (i.e., fixed- versus growth-mindset). Similar to

performance goal orientations, individuals with fixed mindsets believe that, if you have

the ability, then effort is unnecessary. In contrast, individuals with growth mindset value

effort and are more likely to take risks in achievement activities; thus, growth mindset

parallels mastery goal orientation. Dweck also emphasized that fixed versus growth

mindsets are not traits, but perspectives that can be taught. In other words, according to

the growth mindset perspective, educators could help facilitate learners’ understanding

that goals can be achieved through effort and persistence through challenges.

To summarize, AGT is a theory that focuses on the impact an individual’s goal

orientation has on motivational patterns. Students who are more aligned to the

performance orientation care less about the process of gaining knowledge, but are

motivated by norm-based standards. Mastery orientation, on the other hand, is associated

with a motivation to learn. Thus mastery orientated individuals believe effort and success

are linked, and are therefore more likely to persist through challenges. AGT has been

expanded throughout educational literature to include the impact of failure-avoidant

 13

versus success oriented perspectives, and has even driven much of the research on growth

mindset, which is currently receiving a lot of attention in schools. Thus, AGT is a

primary theory of motivation within the field.

Social Cognitive Theory

Social cognitive theory (SCT) is another prominent theory of motivation and

engagement. It posits human behavior as a function of the interaction between personal

dynamics (e.g., concepts of self), actions, and environmental factors (Bandura, 1986).

Based in this theory, goal setting, causal attributions, self-regulation, and self-efficacy are

considered key constructs that underlie both engagement and motivation (Eccles &

Wigfield, 2002; Schunk & Mullen, 2012). Similar to AGT, neither motivation nor

engagement are explicitly defined within this theory. Instead, it is the reciprocal

relationship between the personal factors, actions, and environment that influence both

engagement and motivation.

According to Schunk and Mullen (2012), self-efficacy—an individual’s perceived

ability to learn or complete tasks at different levels of difficulty—can be used to

demonstrate a cyclical relationship. For instance, self-efficacy has been shown to

influences task choice, effort, persistence, and use of learning strategies (i.e., behaviors;

Schunk & Mullen, 2012). Conversely, these actions lead to increased learning, which

affects self-efficacy and an individual’s motivation to continue engaging in behaviors.

This process is often discussed using the term self-regulation within engagement and

motivation literature (Eccles & Wigfield, 2002). Self-regulated students are viewed as

motivationally and behaviorally active in their own learning process and in achieving

their personal goals. Thus, goal setting also plays an important role within the SCT
 14

perspective of engagement and motivation. In a study conducted by Bandura and Schunk

(1981), the intervening process of goal setting and self-evaluation affects self-motivation,

with self-efficacy being associated with greater achievement gains. Thus, unlike AGT

which traditionally views underlying constructs of motivational patterns as traits (e.g.,

mastery or performance oriented), SCT views motivational patterns and engagement as

continuously changing through a reciprocal relationship between concept of self,

environmental factors, and actions.

Self-Determination Theory

Self-determination theory (SDT) also conceptualizes motivation and engagement

in cyclical pattern. SDT is a theory of human motivation, which is defined as being

“energized or activated towards” a task (Ryan & Deci, 2000, p. 54). Thus the primary

focus of SDT is motivation, with an underlying focus on engagement. According to SDT,

humans are driven to maintain an optimal level of stimulation while balancing a need for

competence (Eccles & Wigfield, 2002). In other words, individuals seek tasks that they

are successful in completing, but are not overly simplistic. According to Deci and Ryan

(1985), the developers of SDT, motivation is expressed through different causal

orientations of intrinsic, and extrinsic motivation. Intrinsic motivation refers to doing

something because it is interesting or enjoyable, while extrinsic motivation leads to a

separate outcome (e.g., getting a good grade; Ryan & Deci, 2000). Intrinsic and extrinsic

motivation are maintained when individuals feel both competent and self-determined

(Eccles & Wigfield, 2002).

However, within this theory, motivation is not dichotomously intrinsic or extrinsic

(Ryan & Deci, 2000). Instead, individuals who are truly self-determined transfer external
 15

regulation inside through a process called internalization. The progression of

internalization can be explained through four different levels: external, introjected

(internal based on a feeling of requirement), identified (internal based on the usability of

the behavior), and integrated regulation (internal based on one’s own values; Ryan &

Deci, 2000). For example, consider the possible progression of a student in school –

initially a student may complete homework for an extrinsic reason (e.g., a grade), but

later may recognize that good grades are expected from his or her parents (introjected

regulation). Later that student may realize that good grades will lead to a better college

(identified regulation), and eventually the student may begin to value the act of learning

(integrated regulation). Thus, increased internalization is associated with greater

persistence, a more positive sense of self, and higher quality engagement in achievement

activities (Ryan & Deci, 2000).

Additional Theories

Attribution theory. Attribution theory posits that an individual’s causal

attributions (i.e., explanation) for achievement outcomes determine future achievement

efforts (Weiner, 1985). In other words, their explanation for success or failure will not

only drive their engagement in an activity, but also their future motivation for

achievement. Ability, effort, task difficulty, and luck are some examples of achievement

attributions, which can be categorized under locus of control, stability, and controllability

(Eccles & Wigfield, 2002). Locus of control is represented by internal versus external

believed causality. Stability refers to whether causes can change with time (e.g., effort

can change over time). Lastly, the controllability dimension refers to causes that can and

cannot be controlled (e.g., skill versus luck; Weiner, 1985). Similar to SCT and AGT,
 16

attribution theory also suggests goal setting as an essential, but not the solitary

component of motivation. According to the attribution theory, goal setting and

expectancy to attain that goal are central to motivational beliefs (Eccles & Wigfield,

2002).

Flow theory. Another theory that emphasizes intrinsically motivated behavior is

Csikszentmihalyi’s flow theory (Csikszentmihalyi, 1997). Flow theory defines intrinsic

motivation as the immediate subjective experience that occurs when people are engaged

in an activity. While SDT interprets intrinsic motivation in terms of ultimate as opposed

to immediate reason for behavior, Eccles and Wigfield (2002) suggest that these theories

reflect a different perspective of the same point. Flow theory is also closely aligned to

engagement in that “flow”—a holistic feeling of being immersed in an activity, a merging

of action and awareness, lack of self-consciousness, and a feeling of control over both

actions and the environment—is experienced only when an individual is fully engaged

emotionally (Nakamura & Csikszentmihalyi, 2009). Also, flow can only be achieved

when skill and challenge level are matched, which aligns with SDT’s optimal

performance perspective.

One limitation of flow theory is the fact that it primarily describes optimal levels

of engagement and intrinsic motivation; there is little to no discussion surrounding any

type of engagement or motivation outside of flow. This is quite limiting as flow is

typically experienced by experts (Eccles & Wigfield, 2002). Nakamura and

Csikszentmihalyi (2009) discussed different behaviors in terms of intrinsically rewarding

and aversive situations; relaxation and flow are both intrinsically rewarding, while

overwhelming demands and apathy are aversive situations. Thus, flow theory describes
 17

optimal functioning through complete emotional engagement through intrinsic motivation

and a complete lack of engagement and motivation, but nothing in between.

Interest theories. While there is not a single theory for interests’ role in

motivation and engagement, there was an upsurge in this theoretical framework during

the 1990’s (Eccles & Wigfield, 2002). Across these theories, individual versus situational

interests are differentiated; individual interests refer to stable orientations toward domains

while situational interest is described more as an emotional state associated with an

activity. Individual interests can be broken down even further into feeling- (feelings

associated with a task) and value-related (personal significance) valences (Eccles &

Wigfield, 2002). Specific to learning, both individual and situational interest are

associated with deeper levels of learning (Schiefele, 1991). Thus, interest, which is

merely implied from other theoretical perspectives, may be an important factor missing

from prominent theories previously discussed.

Self-system theory. The self-system theory of motivational development suggests

that autonomy, competence, and relatedness are fundamental motivational needs that lead

to engagement (Fredricks & McColskey, 2012). This model is closely aligned to an

expansion of SDT, but also incorporates some components of interest theories. SDT was

developed beyond intrinsic and extrinsic motivation to include additional causal

orientations: autonomy orientation (decisions based on own choice), control orientation

(decisions based on expectations or environment), and interpersonal orientation

(decisions are beyond own control; Deci & Ryan, 1985), which parallel those represented

in the self-system model of motivation. Regarding the relationship between engagement

and motivation, self-system model assumes motivation is a necessary, but not sufficient,
 18

precursor to engagement. Specific to school settings, Fredricks and McColskey (2012)

suggest that students who have opportunities aligned with these motivational factors (e.g.,

task choice, scaffolded learning opportunities) will be more engaged in the classroom.

Expectancy X value theory. According to the expectancy X value theory, the

amount of effort an individual puts forth to reach a goal is a function of their expectancy

to reach the goal with effort and the value they place on reaching said goal (Brophy,

1983). Brophy argued that motivation to learn must not be maximized, but rather

optimized. The relationship between motivation and performance can be described as an

inverted U, in that an individual reaches their peak performance when motivation is at an

optimal level. Despite attribution theory being founded from expectancy X Value theory,

it also closely aligns to SDT’s perspective that humans are motivated to maintain an

optimal level of stimulation and competence. Additionally, flow theory also emphasizes

the concept of optimizing intrinsic motivation to gain complete engagement.

Models and Measurement

Through previously discussed theories that focus on the constructs of engagement

and motivation it is clear that the relationship between these constructs can be described

in a number of ways. While some (e.g., SCT & SDT) view engagement and motivation as

mutually reinforcing of one another, others (e.g., AGT) view motivation as an underlying

contributor to engagement. Additionally, despite the abundance of theories underlying

engagement and motivation, theories are more general conceptualizations of phenomena.

Models, on the other hand, are statistical representations based on theory. Therefore, a

limited number of models that describe how these constructs relate to one another, using

specific measures and mathematical representations, are included below.

 19

Model of academic enablers. DiPerna and Elliott (2002) define academic

enablers as a set of attitudes and behaviors that allow students to benefit from school.

Within their model, engagement and motivation, along with study and interpersonal

skills, were identified as the constructs that enable students to learn. In this model,

motivation was defined as a person's internal desire to complete a task based on interest,

persistence, and approach regarding academic subjects (DiPerna & Elliott, 1999).

Engagement, on the other hand, was defined as attention and active participation in

classroom activities (DiPerna & Elliott, 2002). Similar to the self-system model of

motivation, DiPerna and colleagues (2002) found that motivation leads to engagement. In

other words, motivation is foundational to engaged behaviors. However, unlike the self-

system model of motivation, the model of academic enablers does not clearly align to any

theoretical perspectives previously discussed.

For their models, DiPerna et al. (2002) relied on the Academic Competence

Evaluation Scales (ACES), which is a measure they created (DiPerna & Elliott, 1999).

While an ACES – Student Rating exists for grades 6th through 12th, the model relied on a

single-informant using the ACES – Teacher Rating for kindergarten through 12th grade.

The ACES provide subscale ratings for both motivation and engagement. With regard to

measurement, the ACES have adequate psychometric properties; test-retest reliability

was between .88 and .97 for teachers. Internal consistency fell between .97 - .98 for

teachers. Regarding validity, concurrent validity between academic performance and the

teacher ACES rating fell between .31 and .87. The ACES demonstrates adequate

psychometric properties for measuring academic enablers and skills, but narrowly aligned

to DiPerna’s model of academic enablers rather than broader theoretical perspectives.

 20

Reeve’s longitudinal model of engagement and motivation. Reeve and Lee

(2014) conducted a longitudinal study to explore the relationship between engagement

and motivation. Unlike the self-system model or the model of academic enablers, which

suggest motivation is necessary for engagement, Reeve and colleagues hypothesized that

engagement could influence motivation. Their hypothesis was founded in SDT, which

posits that individuals who are self-determined transition from external to internal

regulation (Ryan & Deci, 2000). The progression of internalization can be explained

through four different levels: external, introjected (internal based on a feeling of

requirement), identified (internal based on the usability of the behavior), and integrated

regulation (internal based on one’s own values; Ryan & Deci, 2000). Thus, increasing

engagement could help facilitate internalization. In other words, while a student may start

off engaged because of external expectations within the classroom, eventually that

engagement could internalize based on an individual’s values.

Similar to DiPerna and Elliott, Reeve and Lee (2014) relied on a single-informant.

However, unlike in the model of academic enablers which relied on teachers as

informants, Reeve’s longitudinal model relied solely on students as informants.

Additionally, Reeve and Lee relied on multiple measures of motivation and engagement

within their model. In order to measure engagement, they used the behavioral

engagement subscale of the EvsD. While the EvsD reports adequate psychometric

properties for the full measure, it is important to note that the whole measure was not

utilized. Regarding the measurement of motivation, PALS and the mastery goals scale

from the Achievement Goal Questionnaire (AGQ) were used. PALS has an internal

consistency that fell between .60 and .84, and a test-retest reliability between .34 and .61.
 21

Additionally, convergent validity has been reported with goal questionnaires as falling

between.63 and .67 (Fredricks et al., 2011; Skinner, Kindermann, & Furrer, 2009).

Lastly, the mastery goal subscales of the AGQ has reported internal consistency between

.74 and .76, and discriminant validity between .3 and .51 (Finney, Pieper, & Barron,

2004). Reeve and Lee (2014) utilized multiple measures and theoretical perspectives to

create their model; however, only particular subscales were used rather than full

measures, undermining the psychometric properties.

MEW. Martin’s (2007) original MEW was a model made up of 11 factors that

fall within four quadrants: adaptive cognitions, adaptive behaviors, maladaptive

behaviors and impeding cognitions. The adaptive cognitive dimension was made up of

self-efficacy, mastery orientation, and valuing of school, while the adaptive behavioral

dimension was made up of persistence, planning, and study management. Anxiety, failure

avoidance, and uncertain control were factors within the impeding cognitive dimension.

Lastly, maladaptive behavioral dimension was made up of disengagement and self-

handicapping (Martin, 2007). According to Martin, motivation is defined as a set of

beliefs and emotions that influence and guide behavior, while engagement is described

through behavioral, emotional, and cognitive lenses (Way, Reece, Bobis, Anderson, &

Martin, 2015). Despite defining these constructs, motivation and engagement are used

interchangeably within this original model, treating them as the same construct.

Over the past decade Martin’s MEW has undergone a series of modifications. In

the most recent model, Martin and colleagues (2017) parsed apart engagement and

motivation within the wheel. They replaced terminology of the four quadrants with

adaptive motivation, adaptive engagement, maladaptive motivation, and maladaptive

 22

engagement rather than considering positive and negative behavior and cognitions. Thus

engagement is broken into adaptive (i.e., persistence, planning, and task management)

and maladaptive (i.e., self-sabotage and disengagement) behaviors. Similarly, motivation

is also separated into adaptive (i.e., self-belief, learning focus, and valuing) and

maladaptive (i.e., anxiety, failure avoidance, and uncertain control) cognitions (see

Figure 1).

In order to explore the distinctions between motivation and engagement, Martin

et al. (2017) used confirmatory factor analysis (CFA) to examine four possible higher

order factors. The main sample consisted of 5432 secondary students (grades 7th through

12th) from 12 different independent high schools. Just under half of the participants were

female (43%) and 57% were male. The mean socioeconomic status of the participants

was higher than the national average, with 8% of the sample from a non-English speaking

background. For this population, it was determined that the four-factor structure, most

aligned with the MEW, had superior fit (χ2(885) = 8357, CFI = 0.92, RMSEA = 0.039,

SRMR = 0.060) in comparison to the positive and negative two-factor, the engagement

and motivation two-factor, or the motivation/engagement single-factor higher order

models. According to fit indices, both the four-factor and the positive and negative two-

factor models yielded acceptable fit as demonstrated by CF > 0.90, RMSEA < 0.08, and

SRMR < 0.08. However, the four factor higher order model had the best fit,

demonstrating a statistically significantly different chi square (χ2 = 1749, df = 5, p <

0.001). While this study demonstrated that engagement and motivation are unique

constructs, further examination and replication is necessary to confirm Martin’s findings.

 23

Similar to DiPerna’s model of academic enablers, Martin and colleagues also

relied on a single-informant through a single measure that was created by their own

research team. The MES was used to measure students’ perception of their engagement

and motivation in all models of the MEW. The MES has demonstrated test-retest

reliability between .61 and .81, and internal consistency between .70 and .87, as well as

criterion related validity between .40 and .63 for achievement and academic outcomes

(Fredricks et al., 2011; Liem & Martin, 2012). Thus, the MES appears to demonstrate

adequate psychometric properties for measuring engagement and motivation, but similar

to the ACES, the MES is narrowly aligned to MEW model.

The MEW encompasses multiple theoretical perspectives that most closely align

to AGT and SCT. AGT, which focuses on goal orientation, aligns to the learning focus

and failure avoidance factors found within the motivation quadrants. SCT on the other

hand, focuses on the relationship between personal dynamics, environmental factors, and

causal attributions. Within the MEW self-belief, persistence, anxiety, and uncertain

control all fall within the SCT perspective. However, according to Martin et al. (2017)

the MEW was also influenced by attribution theory, expectancy X value theory, and

SDT. Thus, unlike other models of engagement and motivation that outline the direction

of influence (i.e., motivation effects engagement, or engagement effects motivation), the

MEW posits a reciprocal relationship between the two constructs. Therefore, the MEW

was utilized in the current study because of the broad conceptualization of engagement

and motivation, and incorporation of multiple theoretical perspectives.

Limitations of Current Models

 24

Many of the models presented offer promising explanations for how motivation

and engagement relate to one another in an educational context. However, there are a

number of limitations to these models. These limitations provide an opportunity for

further research to explore our understanding of motivation and engagement and how

they relate to one another.

One limitation to existing models is the lack of replication. To date, the models

presented above have not been replicated outside of the original research teams.

Specifically for the MEW, which is the theoretical basis of the present study, Martin and

colleagues have researched the factor structure of their model over the past decade.

Despite the fact that they have run extensive studies across multiple countries, replication

outside of their research team has not occurred. According to Tabachnick & Fidell

(2013), replication is particularly important within CFA, the analyses used to compare

models and their underlying factor structure. Beyond CFA, replication is an area of

critical concern within the field and should be viewed as an essential aspect of science

(Simons, Holcombe, & Spellman, 2014). Therefore, replication of theoretical models

should be viewed as not only critical, but also foundational.

Another limitation is the fact that many of these models rely on a single-informant

or single measure of engagement or motivation within their model. According to

Maruyama (1998), using multiple measures for each construct is the only defensible way

in which to create viable models. However, it is common for researchers to rely on a

single measure to represent a latent variable as long as it is made up of at least the

minimum number of recommended items. Martin and colleagues, for example, relied

solely on their MES to build their model of motivation and engagement, while DiPerna
 25

and Elliott relied on their ACES. Thus, violating Maruyama’s recommendation for

creating a defensible model. Relying on the MES is also problematic because it was

created by Martin’s team. Thus, the MES was created based on the MEW model rather

than relying on measures of motivation and engagement to inform the construction of the

model. For example, the MES includes items that are aligned to SDT (e.g., valuing), but

the MES does not include any questions aligned to interest (another aspect of SDT). It is

possible that interest still fits within their model under adaptive motivation, but has not

been assessed due to relying solely on the MES when examining the structure of the

MEW. Therefore, there is a need to not only replicate the MEW but also examine the

adaptability of the model through including additional measures of adaptive and

maladaptive motivation and engagement within the model.

Summary of Literature Review

Specific to engagement, AGT and flow theory described a multidimensional

perspective of engagement (i.e., including both internal and external factors). The model

of academic enablers, on the other hand, utilized the unidimensional perspective of

engagement as on-task behaviors. While there is some overlap across theories of

motivation and engagement, each theory provides a unique perspective of motivational

patterns and engagement. Similarly, key components and terminologies associated with

motivation and engagement varied across theories. Based on the theories and models

presented, motivation can be described through goal setting, concept of self, desire,

reward, interest, and expectancy of success. These components align to the terminologies

outlined in a review conducted by Murphy and Alexander (2000), which included goals

(e.g., AGT), intrinsic and extrinsic factors (e.g., SDT), interest (e.g., interest theories),
 26

and concept of self (e.g., SCT) which were described as central terminologies associated

with motivation. The current study aimed to capture these central components of

engagement and motivation that are not assessed in the MES (e.g., desire, on-task

behavior) in order to determine if they align with the four-factor higher-order structure of

the MEW.

While each theory of motivation and engagement has distinct characteristics,

there are a number of overlapping constructs across theories and models. For instance,

individuals who are more aligned to mastery goal orientation tend to have higher self-

efficacy, relating AGT and SCT. This makes sense given the reciprocal relationship

described between self-efficacy and behavior; according to AGT, mastery oriented

individuals are more likely to take on and persevere through challenging achievement

tasks. Thus, the behaviors of mastery oriented individuals fuel the reciprocal relationship

described in SCT increasing self-efficacy. Similarly, SDT, flow theory, and the

expectancy X value model all emphasize the optimal level of performance based on

motivation. Specific to the various interest theories, Eccles and Wigfield (2002) suggest

that flow falls within the feeling-valence dimension of individual interest. Additionally,

intrinsic and extrinsic motivation, which are primarily discussed in SDT, align to the

concept of individual versus situational interests, respectively.

Regarding the relationship, in some instances engagement and motivation are

seen as interchangeable (e.g., the original MEW; Martin, 2007). Other theories include

engagement as a subcomponent of motivation (e.g., Self-System Model of Motivation;

AGT). However, some theories conceptualize engagement and motivation as completely

separate from one another (e.g., SCT; SDT). SDT is of particular interest because it
 27

allows for an alternate directionality of the relationship between motivation and

engagement. While other theories explain how motivation impacts engagement (e.g.,

AGT; attribution theory; Reschly & Christenson, 2012), SDT posits the concept that

engaging tasks actually drive motivation (Reeve, 2012). Similarly, SCT discusses the

reciprocal nature of self-efficacy and behaviors. Through this lens, motivational patterns

and engagement continuously change through a cyclical relationship between concept of

self, environmental factors, and actions, which influences both engagement and

motivation, which aligns to the updated MEW (Martin et al., 2017). The current study

examined the relationship between engagement and motivation through comparison of

different higher-order models.

Gaps in the literature. There are a considerable number of questions remaining

within educational literature that cannot fully be answered at the present time. First, there

is no model that encompasses all underlying constructs of motivation and engagement

across theoretical perspectives. However, Martin’s updated MEW holds promise in that it

incorporates multiple theoretical perspectives and 11 factors that make up adaptive and

maladaptive motivation and engagement. There are a number of reviews of different

theoretical perspectives, but there is little consensus across researchers as to the essential

components of these constructs. The lack of agreement surrounding operational

definitions has led to an abundance of misconceptions within the field. This is

problematic because engagement and motivation as broad constructs are known

predictors of academic achievement. In other words, engagement and motivation are

essential to academic success (DiPerna & Elliott, 2002). However, how can educators
 28

and researchers intervene to help foster growth in these areas if we continue to lack clear

and consistent operational definitions or theoretical understanding?

Additionally, how do motivation and engagement truly relate to one another? The

two constructs are often discussed in tandem, but are researched separately; there are

separate interventions that focus on engagement and motivation, and separate measures

for engagement and motivation. While multiple researchers have aimed to answer how

motivation and engagement are related, there is little to no replication, with each research

team reaching a different conclusion. For instance, DiPerna and Elliott (2002) used

structural equation modeling in their model of academic enablers and concluded that

motivation leads to engagement. Conversely, Reeve and Lee (2014) found that

engagement can actually alter motivation. Most recently, Martin and colleagues (2017)

concluded that engagement and motivation actually influence one another. In addition to

the lack of replication, many of these studies relied on a single measure of engagement

and motivation that was created by the same researchers; DiPerna and Elliott relied on the

ACES, while Martin and colleagues utilized the MES. If motivation and engagement are

as linked as theoretically believed, interventions could be optimized to target these

constructs to help prevent the loss that occurs throughout adolescence (e.g., Archambault

et al., 2009; Harter, 1981).

Summary

The literature surrounding motivation and engagement is extensive, but

also chaotic. Despite clear associations between engagement, motivation, and academic

achievement, there is confusion surrounding the conceptualization of both engagement

and motivation. Operational definitions of these constructs not only differ across theories,
 29

but often are not explicit. However, in synthesizing various theories and models of

motivation and engagement, patterns emerge; both engagement and motivation are

typically considered from a multidimensional perspective. Thus, engagement is

conceptualized to include both internal and external components, while motivation is best

described through a synthesis of various theoretical perspectives. In other words, there is

no single underlying component that fully captures motivation or engagement, but instead

a number of subcomponents that can be used to help conceptualize these constructs.

Thus, there is a need for additional research that not only replicate previous findings, but

also utilizes multiple measures to better understand the relationship between motivation

and engagement. The primary purpose of the current study was to examine the higher

order structure of the MEW, a model that encompasses multiple theoretical perspectives,

by replicating the models presented by Martin et al. (2017) with a more diverse, and

elementary-aged American sample. Secondarily, this study aimed to capture additional

components of adaptive and maladaptive engagement and motivation by including

additional measures of these constructs, beyond what was used by Martin and colleagues.
 30

CHAPTER 3

Method

Setting

Data collection took place at two elementary schools (grades pre-kindergarten

through sixth) in an urban area within the Midwestern United States. According to 2017-

2018 school data for the first school, there were 338 students, in which 85% were African

American, and 15% were Hispanic. For the second school, which had 381 students, 37%

of students were African American, 31% were White, 13% were Asian, and 10% were

Hispanic. Additionally, 96% of the students at the first school were identified as coming

from a low-income background as determined by the percentage of students who received

free or reduced price lunch. For the second school, 66% of students were identified as

coming from a low-income background and 12% had limited English proficiency.

Participants

Participants were recruited at the classroom level. Thus, teachers of 3 rd through 6th

grade students were first recruited for participation. After teachers were recruited,

students in their class were then recruited for participation. To this end, parents first

received a passive consent form, which determined which students were eligible to

participate. Parents who signed and returned the opt-out form if they did not want their

child to participate. Students whose parents opted-out of participating were provided with

an alternative activity (e.g., reading time, work-sheet) while the class completed the

questionnaire. Finally, prior to the start of the study, student assent was obtained from

those students for whom parental passive consent was obtained. Students were not given
 31

any incentive for participating in this study. However, participating teachers received a

$10 gift card as compensation for their time completing study measures.

Thus, there were 360 potential participating students and 16 teachers initially

recruited for study participation. However, one teacher did not agree to have her class

participate (potential student n = 340), 14 parents waived consent, seven students did not

assent, 16 discontinued after starting, 15 were absent, and there were 18 technological

errors. Thus, 270 students and 15 teachers were included in the final sample. See Table 1

for demographic information of the final sample.

Procedures

Data were collected using a multi-informant and multi-method design, allowing

for in-depth analysis of both children and teachers’ perspectives of academic engagement

and motivation. Data were collected at a single time point during the spring of 2018. The

primary method of data collection involved teacher and student questionnaires

administered using Qualtrics on iPads. Teacher questionnaires were made up of a single

measure (Engagement Versus Disaffection with Learning-teacher report [EvsD]) with 25

items per student, while student questionnaires included three measures (Motivation and

Engagement Scale [MES], EvsD-student report, and Patterns of Adaptive Learning Scale

[PALS]) consisting of a total of 129 items. The order of the measures within the student

questionnaires were randomized across classrooms, creating multiple versions of the

student questionnaires in an effort to guard against potential order effects.

The questionnaire was administered class-wide so all participating students

completed the measures at the same time on iPads. The student questionnaire took

approximately 40 minutes to complete. Additionally, trained graduate students

 32

administered the student questionnaires and read the items out loud in order to help

students comprehend each item and respond to the best of their abilities. Teachers were

asked to complete questionnaires for each individual student, in a randomly assigned

order. Teacher questionnaires were completed while the graduate students administered

the student questionnaire.

Measures

Student demographic information was obtained through the district online portal,

and teachers completed a demographic questionnaire to gather information such as their

level of education and number of years teaching (see Appendix B). Four measures of

motivation and engagement were used to capture students’ engagement and motivation:

MES, EvsD – teacher report and student report, and PALS. Each of these measures has

been validated with upper elementary aged students and captures both adaptive and

maladaptive components of motivation and engagement, therefore aligning to Martin’s

model theoretically.

MES. Martin’s (2016) MES – Junior School is a student self-report measure for

elementary aged students that assesses adaptive motivation, adaptive engagement,

maladaptive motivation, and maladaptive engagement. The MES measures 11 subscales

that are grouped under four scales—self-belief (e.g., “If I try hard, I believe I can do my

schoolwork well”), learning focus (e.g., “I feel very pleased with myself when I really

understand what I am taught at school”), and valuing (e.g., “Learning at school is

important to me;” adaptive motivation); persistence (e.g. “If I can’t understand my

schoolwork at first, I keep going over it until I understand it”), planning (e.g., “Before I

start an assignment I plan out how I am going to do it”), and task management (e.g.,
 33

“When I study, I usually study in places where I can concentrate;” adaptive engagement);

anxiety (e.g., “When exams and assignments are coming up, I worry”), failure avoidance

(e.g., “Often the main reason I work at school is because I don’t want to disappoint my

parents”), and uncertain control (e.g., “I’m often unsure how I can avoid doing poorly at

school;” maladaptive motivation); and self-sabotage (e.g., “I sometimes don’t study very

hard before exams so I have an excuse if I don’t do as well as I hoped”) and

disengagement (e.g., “I often feel like giving up at school;” maladaptive engagement; see

Appendix C for the full measure). Students provided a rating for each item on a Likert

sale (1 = strongly disagree and 5 = strongly agree). The MES provides scores for the

eleven subscales and four scales, where higher scores are better for the adaptive factors

and lower scores are better for the maladaptive factors. The MES – Junior School takes

approximately 10 minutes to administer and has been validated for students between the

ages nine and 13 (Martin, 2016). Research has demonstrated adequate test-retest

reliability (.61-.81) and internal consistency (.70-.87), as well as criterion related validity

(.40-.63) for achievement and academic outcomes (Fredricks et al., 2011; Liem & Martin,

2012).

EvsD. In addition to the MES, teachers and students completed the EvsD. The

EvsD includes four subscales for both the teacher and student reports: behavioral

engagement (e.g., “I try hard to do well in school”), behavioral disaffection (e.g., “When

I’m in class, my mind wanders”), emotional engagement (e.g., “When we work on

something in class, I get involved”), and emotional disaffection (e.g., “When we start

something new in class, I feel nervous;” see Appendices D and E for the full measures).

For each item, respondents used a 4-point Likert-type scale: ranging from 1 (not at all
 34

true) to 4 (very true). The EvsD, which was designed for students between the ages of

eight and 13, has demonstrated strong internal consistency for the teacher report (.81-

.87), but more variable internal consistency for the student report (.61-.85). Test-retest

reliability was also higher for the teacher report (.65-.82) compared to the student report

(.53-.68). Additionally, adequate evidence of criterion related validity has been

demonstrated for both the teacher report (.50-.81) and the student report (.34-.61;

Fredricks et al., 2011; Skinner et al., 2009).

PALS. Students also completed the PALS student self-report. The PALS Personal

Achievement Goal Orientations includes three scales: Mastery Goal Orientation (e.g.,

“It’s important to me that I improve my skills this year”), Performance-Approach Goal

Orientation (e.g., “It’s important to me that I look smart compared to others in my

class”), Performance-Avoid Goal Orientation (e.g., “One of my goals in class is to avoid

looking like I have trouble doing the work”). The PALS Academic–Related Perceptions,

Beliefs, and Strategies includes 8 scales: Academic Efficacy (e.g., “I am certain I can

master the skills taught in class this year”), Academic Pressure (e.g., “When I’ve figured

out how to do a problem, my teacher gives me more challenging problems to think

about”), Academic Self-Handicapping Strategies (e.g., “Some students purposely don’t

try hard in class. Then if they don’t do well, they can say it is because they didn’t try.

How true is this of you?”), Avoiding Novelty (e.g., “I would prefer to do class work that

is familiar to me, rather than work I would have to learn how to do”), Cheating Behavior

(e.g., “I sometimes cheat on my class work”), Disruptive Behavior (e.g., “I sometimes get

into trouble with my teacher during class”), Self-Presentation of Low Achievement (e.g.,

“One of my goals in class is to avoid looking smarter than other kids”), and Skepticism
 35

About the Relevance of School for Future Success (e.g., “My chances of succeeding later

in life don’t depend on doing well in school;” see Appendix F for the full scales). PALS

uses five point Likert-type scales: items are anchored at 1 = "Not at all true,” 3 =

"Somewhat true,” and 5 = "Very true." The PALS has been validated with students

between the ages of six and 18. It has demonstrated adequate internal consistency (.60-

.84), and test-retest reliability (.34-.61). Additionally, convergent validity has been

demonstrated with alternate goal questionnaires (.63-.67; Midgley et al., 2000).

Data Analysis Plan

Descriptive analyses. Before testing the CFA, preliminary descriptive analyses

were conducted using R version 1.0.136 and Jamovi version 0.9.5.15 to examine

descriptive statistics. Distribution of variables were examined using graphs to address

potential problems with outliers and account for skewed distribution. Additional analyses

were conducted to ensure no statistically significant differences based on the order of

questionnaire completion or other demographic features were present. Lastly, correlation

matrices for all subscales were calculated in order to create measurement models for the

additional measures of engagement and motivation (i.e., EvsD teacher and student

reports, and PALS).

Structural Equation Modeling. To assess the measurement models, CFA were

performed through MPlus 8.1 on the four measures presented above. CFA were

conducted following Rindskopf and Rose’s (1998) method of comparing nested models.

Models were tested by comparing the fit of least to most restrictive competing models.

While maximum likelihood estimation is most commonly used, it is only appropriate for

normally distributed continuous data. Therefore, robust weighted least squares

 36

(WLSMV) estimation was used to estimate the model parameters because it is robust

against smaller sample sizes and variables with floor or ceiling effects and is the best

estimator for categorical data (Brown, 2014).

Primary analysis: Proposed replication models. Four higher-order models were

tested based upon models used by Martin et al. (2017). The hypothesized path diagrams

are presented in Figures 2 through 5 in order of their restrictiveness, where circles

represent higher-order latent variables and ovals represent first-order factors. Absence of

a line connecting variables implies no hypothesized direct effect, while a curved line

represents a correlation.

1) Higher-Order Single-Factor Model of Motivation/Engagement (Figure 2): This

model hypothesized the presence of a combined motivation/engagement higher-

order factor, which all 11 first-order factors load on (i.e., self-belief, learning

focus, valuing, persistence, planning, task management, anxiety, failure

avoidance, uncertain control, self-sabotage, and disengagement).

2) Higher-Order two-factor Model of Adaptive and Maladaptive Qualities

(Figure 3): This model hypothesized two higher order factors, Adaptive and

Maladaptive behaviors and cognitions. Six first-order factors load on the Adaptive

factor (i.e., self-belief, learning focus, valuing, persistence, planning, and task

management), while five first-order factors load on the Maladaptive factor (i.e.,

anxiety, failure avoidance, uncertain control, self-sabotage, and disengagement).

3) Higher-Order two-factor Model of Motivation and Engagement (Figure 4):

This model hypothesized two factors, Motivation and Engagement. Six first-order

factors load on the Motivation factor (i.e., self-belief, learning focus, valuing,
 37

anxiety, failure avoidance, and uncertain control), and five first-order factors load

on the Engagement factor (i.e., persistence, planning, task management, self-

sabotage, and disengagement).

4) Higher-Order four-factor Model of Adaptive Motivation, Adaptive

Engagement, Maladaptive Motivation, and Maladaptive Engagement (Figure 5):

This model hypothesized four factors, Adaptive Motivation, Adaptive

Engagement, Maladaptive Motivation, and Maladaptive Engagement. Three first-

order factors load on the Adaptive Motivation (i.e., self-belief, learning focus, and

valuing), Adaptive Engagement (i.e., persistence, planning, and task

management), and Maladaptive Motivation (i.e., anxiety, failure avoidance, and

uncertain control), while two factors load on Maladaptive Engagement (i.e., self-

sabotage, and disengagement).


Secondary analysis: Proposed adaptability models. Four higher-order models

were tested based upon models used by Martin et al. (2017), but including alternate

subfactors of engagement and motivation as measured by PALS and EvsD. The

hypothesized path diagrams are presented in Figures 6 through 9 in order of their

restrictiveness.

5) Higher-Order Single-Factor Model of Motivation/Engagement (Figure 6): This

model hypothesized the presence of a combined motivation/engagement higher-

order factor, which 19 first-order factors load on (i.e., teacher and student rated

emotional engagement, mastery goal orientation, performance-approach goal

orientation, academic efficacy, teacher and student rated behavioral engagement,

academic press, teacher and student rated emotional disaffection, performance-

 38

avoid goal orientation, avoiding novelty, self-presentation of low achievement,

skepticism about relevance of school for future success, teacher and student rated

behavioral disaffection, academic self-handicapping, cheating behavior, and

disruptive behavior).

2) Higher-Order two-factor Model of Adaptive and Maladaptive Qualities

(Figure 7): This model hypothesized two higher order factors, Adaptive and

Maladaptive behaviors and cognitions. Seven first-order factors load on the

Adaptive factor (i.e., teacher and student rated emotional engagement, mastery

goal orientation, academic efficacy, teacher and student rated behavioral

engagement, academic press), while twelve first-order factors load on the

Maladaptive factor (i.e., teacher and student rated emotional disaffection,

performance-approach goal orientation, performance-avoid goal orientation,

avoiding novelty, self-presentation of low achievement, skepticism about

relevance of school for future success, teacher and student rated behavioral

disaffection, academic self-handicapping, cheating behavior, and disruptive

behavior).

3) Higher-Order two-factor Model of Motivation and Engagement (Figure 8):

This model hypothesized two factors, Motivation and Engagement. Eleven first-

order factors load on the Motivation factor (i.e., teacher and student rated

emotional engagement, mastery goal orientation, performance-approach goal

orientation, academic efficacy, teacher and student rated emotional disaffection,

performance-avoid goal orientation, avoiding novelty, self-presentation of low

achievement, skepticism about relevance of school for future success), and eight
 39

first-order factors load on the Engagement factor (i.e., teacher and student rated

behavioral engagement, academic press, teacher and student rated behavioral

disaffection, academic self-handicapping, cheating behavior, and disruptive

behavior).

4) Higher-Order four-factor Model of Adaptive Motivation, Adaptive

Engagement, Maladaptive Motivation, and Maladaptive Engagement (Figure 9):

This model hypothesized four factors, Adaptive Motivation, Adaptive

Engagement, Maladaptive Motivation, and Maladaptive Engagement. Four first-

order factors load on Adaptive Motivation (i.e., teacher and student rated

emotional engagement, mastery goal orientation, academic efficacy). Three first-

order factors load on Adaptive Engagement (i.e., teacher and student rated

behavioral engagement, academic press). Seven first-order factors load on

Maladaptive Motivation (i.e., teacher and student rated emotional disaffection,

performance-approach goal orientation, performance-avoid goal orientation,

avoiding novelty, self-presentation of low achievement, skepticism about

relevance of school for future success), and five factors load on Maladaptive

Engagement (i.e., teacher and student rated behavioral disaffection, academic

self-handicapping, cheating behavior, and disruptive behavior).


Difference tests and χ2 probability were used to assess model fit. A χ2 that is not

statistically significant suggests that the hypothesized model does not differ from the

population model meaning that the model fits the population of students. Model fit was

assessed with the following goodness-of-fit criteria:

1) Comparative fit index (CFI): A ratio of the fit of the estimated model over the
 40

null hypothesis model where the closer to 1.0 the better the fit (Brown, 2014).

Values greater than .90 indicate an adequate fit and values above .95 indicate a

good fit (Hu & Bentler, 1999; Tabachnick & Fidell, 2013). 


2) Root mean square error of approximation (RMSEA): A calculation of the lack

of a good fit in comparison to the ideal model, where the closer to 0 the better the

fit (Brown, 2014). RMSEA values equal to or below .08 indicate an adequate fit,

values equal to or below .06 indicate a good fit, and values equal to or below .05

indicate excellent fit (Browne & Cudeck, 1993, Hu & Bentler, 1999; Schumacker

& Lomax, 2010). 


3) Standardized root mean square residual (SRMR): A calculation similar to

RMSEA, where the closer to 0 the better the fit. SRMR values equal to or below

.08 indicate close fit, while a value equal to or below .05 indicates excellent fit

(Schumacker & Lomax, 2010).

Chi-squared difference tests were used to determine if the nested models were a

statistically significant improvement over the baseline model. In the case of χ 2 difference

tests, a statistically significant finding suggests an improved model fit. Models that

resulted in statistically significant improvement over the baseline model were compared

and the final model was determined by the best overall fit indices.
 41

CHAPTER 4

Results

Analytic Assumptions

Based on the number of parameters within the models, the initial objective was a

sample size of a minimum of 300 student participants. Despite fewer participants than

anticipated (i.e., n = 270), two separate analyses confirmed the sample size was adequate

for the replication models: Bartlett’s test of Sphericity [X2(946) = 4951, p < .001] and

Kaiser (KMO) measure of sampling adequacy (all items < .6 indicating adequate).

Sample size was smaller for the adaptability models due to the fact that it included

teacher surveys. Based on teacher response, only 185 students were included for the

adaptability models. Thus, KMO indicated some items > .6. However, Bartlett’s

Sphericity still indicted adequate sample size, X2(5995) = 13,529, p < .001. Based on a

Little’s MCAR test that was not statistically significant, data appeared to be missing

completely at random. Additionally, logistic regression was implemented with dummy

coded items (i.e., missing data were coded as 1 while non-missing were coded as 0),

which indicated neither school, grade, race, nor order of measures predicted missingness

(p > .05 across all items). In other words, data were missing with no clear pattern of

clustering based on demographics. Because less than 1% of data were missing, missing

data were left empty and pairwise deletion was utilized, which is recommended with

WLSMV estimation.

Normality of variables was assessed by visually examining histograms using R

Studio and by calculating standardized skewness and kurtosis statistics (see Tables 2-6).

Skewness and kurtosis were present across measures. However, this was expected given
 42

the nature of the scales (i.e., it was expected that students would rate their positive

motivation and engagement higher and their negative motivation and engagement lower).

Consequently, WLSMV estimation was used, which is recommended for both smaller

sample sizes and non-normal data. Linearity was assessed using scatterplots and best-fit

lines – variables appeared to be linearly related.

Descriptive Statistics and Correlations

Multiple linear regression analyses were used to assess any statistically significant

differences in motivation and engagement, as measured by the MES, EvsD, and PALS,

based on race, grade level, gender, school, and the order measures were presented. Race

( = -0.02, p > .05), grade level ( = -0.02, p > .05), biological sex ( = 0.01, p > .05),

school ( = 0.04, p > .05), and order the measures were presented ( = -0.01, p > .05)

were not statistically significant predictors of engagement and motivation. The overall

model fit was R2 = 0.01.

Across measures, students rated their Adaptive Engagement and Motivation higher

and their Maladaptive Engagement and Motivation lower. On the MES, which provided

raw scores out of 5 and converted scores out of 100, the average subscale scores and

Cronbach’s alphas from the current sample aligned to Martin’s (2016) sample. On

average, subscale scores differed by only 1.5 for the converted scores out of 100 across

the two samples; the largest average score difference was found in the Valuing subscale

in which the current sample had a Mean of 84 (raw score = 4.20) compared to an average

of 89 (raw score = 4.45) for Martin’s sample, while some subscales demonstrated no

difference across samples (i.e., Self-Belief and Task Management). Differences between

Cronbach’s alpha coefficients were also minimal across samples with alphas ranging
 43

between questionable (.6 <  < .7) and good (.8 <  < .9), with the majority of alpha

coefficients falling in the adequate range (.7 <  < .8); see Table 2). Similar results (i.e.,

higher positive and lower negative components of motivation and engagement and

adequate alpha coefficients) were found on the student and teacher EvsD and PALS

measures. For information regarding average subscale and scale scores and Cronbach’s

alpha coefficients across measures see Table 3. For more detailed descriptive statistics by

item, see Tables 4 through 7.

Tables 8 through 11 present the Pearson’s correlation matrices for the subscales of the

four measures by adaptive and maladaptive motivation and engagement. Most of the

subscales demonstrated statistically significant correlations with one another, with the

exception of some scales from the teacher EvsD. The MES and student EvsD

demonstrated statistically significant (p < .001) and moderate correlations (r = 0.39 –

0.61) across subscales. Additionally, both were weakly to moderately correlated with

subscales of the PALS (r = 0.04 – 0.65).

Confirmatory Factor Analysis

Based on the research questions that asked which higher-order structure, as

replicated from models proposed by Martin and colleagues (2017), best fits an American,

elementary aged population across various measures, we examined eight possible higher

order factor structures using CFA. Evidence of motivation and engagement as distinct

constructs would be indicated by superior fit for the multi-factor higher order models.

Replication models. Four replication models were run to assess the higher order

structure behind motivation and engagement. Model 1, which was a single-factor higher

order model that combined engagement and motivation, was run first in order to allow for
 44

model comparisons. Fit indices for Model 1 indicated poor fit, χ2(891) = 2372, CFI =

0.86, RMSEA = 0.08, SRMR = 0.10. Model 2, a two-factor higher order model that

separated adaptive and maladaptive constructs, yielded χ2(890) = 1981, CFI = 0.90,

RMSEA = 0.07, SRMR = 0.09. Model 3, a two-factor higher order model that separated

engagement and motivation yielded χ2(890) = 2354, CFI = 0.86, RMSEA = 0.08, SRMR

= 0.10. Lastly, Model 4, a four-factor higher order model that separates engagement and

motivation into adaptive and maladaptive constructs yielded χ2(885) = 1577, CFI = 0.94,

RMSEA = 0.05, SRMR = 0.08. Despite statistically significant χ2, indicating that the

models do not fit the population, the four-factor and two-factor adaptive and maladaptive

model yielded acceptable fit (CFI > 0.90, RMSEA < 0.08, and SRMR < 0.08) on

additional fit indices. Of these two models, the hypothesized four-factor higher order

model (Model 4) fit best, as determined by a statistically significantly different chi square

(Δχ2 = 175, df = 5, p < 0.001). Standardized factor loadings and errors for Model 4 are

presented in Tables 12 and 13, and Figure 10 presents the final model with standardized

coefficients.

Adaptability models. In order to answer the second research question, four

additional adaptability models, with the same higher order structure as the replication

models, were run. Model 5, a single-factor higher order model that combined engagement

and motivation could not converge. Model 6, a two-factor higher order model that

separated adaptive and maladaptive constructs yielded χ2(5865) = 9769, CFI = 0.89,

RMSEA = 0.05, SRMR = 0.14. Model 7, a two-factor higher order model that separated

engagement and motivation could not converge. Lastly, Model 8, a four-factor higher
 45

order model that separates engagement and motivation into adaptive and maladaptive

constructs yielded χ2(5860) = 9696, CFI = 0.90, RMSEA = 0.05, SRMR = 0.14. While

two of the models were unable to converge, the four-factor and two-factor adaptive and

maladaptive model yielded estimates and met criteria for adequate fit in at least a single

fit index. Of these two models, the hypothesized four-factor higher order model fit best,

as determined by a statistically significantly different chi square (Δχ2 = 93, df = 5, p <

0.001). Standardized factor loadings and errors are presented in Tables 14 and 15, and

Figure 11 presents the final model with standardized coefficients.

 46

CHAPTER 5

Discussion

The primary purpose of this study was to examine the relationship between

adaptive and maladaptive engagement and motivation by replicating the models

presented by Martin et al. (2017) with a more diverse, and elementary aged American

sample to assess the four-factor higher-order model that represents the MEW.

Specifically, it investigated if a model with adaptive engagement, adaptive motivation,

maladaptive engagement, and maladaptive motivation as four separate constructs had

better fit in comparison to models that combined adaptive and maladaptive engagement

and motivation. In order to strengthen the MEW as a theoretical model and our

conceptualization of positive and negative engagement and motivation, this study

secondarily aimed to examine how additional components of these constructs, beyond

those included in the current MEW, fit within the four-factor structure proposed by

Martin and colleagues.

Engagement Versus Motivation: How Many Constructs?

Of the four models replicated from Martin and colleagues (2017), the two-factor

adaptive and maladaptive model (Model 2) and four-factor model (Model 4) indicated

adequate fit for at least two fit indices. Although χ2 values were statistically significant

across models, indicating models did not fit, χ2 are impacted by several factors (e.g.,

sample size, correlations) and are often criticized for having compromised statistical

significance (Brown, 2014). Thus, additional fit indices for absolute fit (i.e., SRMR),

parsimony correction (i.e., RMSEA), and comparative fit (i.e., CFI) were included. Based

on these additional fit indices, Model 2 demonstrated adequate fit based on RMSEA and
 47

CFI, while Model 4 demonstrated adequate fit based on CFI, close fit based on SRMR,

and excellent fit based on RMSEA. In other words, the four-factor model (Model 4)

indicated better fit across the different fit indices. Additionally, χ2 indicated Model 4

was a statistically significantly better fit compared to Model 2. Thus, findings from the

current study replicated Martin and colleagues’ (2017) outcomes, indicating adaptive

engagement, adaptive motivation, maladaptive engagement and maladaptive motivation

are four separate constructs.

Despite demonstrating better fit compared to the single-factor model (Model 1),

the two-factor model that conceptualized engagement and motivation as separate

constructs (Model 3) did not yield adequate fit. While this corresponded with findings

from Martin and colleagues (2017), it is still thought-provoking. The overlap between

engagement and motivation, and the components of these constructs, is widely

recognized (e.g., Fredricks et al., 2004; Martin, 2012; Martin et al., 2017). However,

conceptualizing these constructs as highly related, but separate factors has become more

prevalent (Reschly & Christenson, 2012). While current findings show support for the

separation of engagement from motivation, as indicated by Model 4’s superior fit, results

also suggest the interconnectedness of these constructs (as indicated by the lack of fit in

Model 3). In other words, it was not until the positive and negative constructs were

separated that the differentiation between engagement and motivation was evident. This

is particularly interesting because theories of motivation and engagement typically do not

explicitly discuss maladaptive components of these constructs (e.g., SCT and SDT).

Thus, one possible explanation is that the inclusion of the maladaptive components in the

current model interfered with the precision of engagement or motivation as latent factors.
 48

In other words, the inclusion of the negative components of engagement and motivation

created constructs that were too expansive, which therefore did not fit as independent

latent factors. It is possible that engagement and motivation would still be best explained

as separate constructs if they were defined and measured in a way that was more

theoretically aligned (i.e., only including positive aspects of these constructs), but the

current study did not explore this further. However, the current study did highlight the

information gained, and importance of the inclusion of maladaptive aspects of these

constructs, as demonstrated by the superior fit of Model 4.

Findings from the replication models, aligned to the first research question, were

important for a number of reasons. First, neither the original nor the updated MEW has

been replicated outside of the original research group. Additionally, replication is

particularly critical with CFA due to the influence of the sample on factor analytic

techniques (Tabachnick & Fidell, 2013). Lastly, replication is an essential component to

research. Recently, psychology has been faced with a “replication crisis.” A lack of

replication undermines findings and limits our understanding for whom and under what

conditions conclusions can be drawn. This is the first study that has examined the higher

order relationship between adaptive and maladaptive engagement and motivation for a

diverse sample. Martin et al. (2017) noted that their sample, which was made up of

mostly White upper-class Australians, was a limitation to their findings and replicating

their models with a more diverse population was suggested as a next step for research.

Thus, findings from the current study extend Martin’s findings, allowing us to expand the

representation of the MEW.

The findings from the current study also indicate that the relationship between
 49

these constructs may be stable across age. While Martin and colleagues focused on a

secondary population, the current study utilized an elementary aged sample. The fact that

the results replicated is particularly interesting given our current understanding of the

development of both engagement and motivation; engagement and motivation have been

shown to decrease from childhood into adolescence (e.g., Eccles et al., 1993; Greenwood,

2002; Harter, 1981). However, there have not been studies to date that have explored how

or if the relationship between these constructs differs throughout development. Findings

from the current study support the notion that engagement and motivation could endure a

similar relationship over time, despite decreases in engagement and motivation across

development.

Beyond the MES: Additional Measures for the MEW

While the primary research question focused on replication supporting the

structure of the MEW, the secondary research question aimed to provide additional

support for the MEW as a theoretical model through supplementing alternative measures

of adaptive and maladaptive motivation and engagement. Of the four adaptability models,

the two-factor adaptive and maladaptive model (Model 6) and four-factor model (Model

8) were the only models that produced estimates; the single-factor (Model 5) and two-

factor engagement and motivation (Model 7) models did not converge. A lack of

convergence could occur for a number of reasons (e.g., complexity or grossly

misspecified models; Brown, 2014). Thus, it is likely that Models 5 and 7 would have

poor-fit. A number of adaptations were attempted in order to encourage convergence (i.e.,

increasing the maximum number of iterations, co-varying variables that were

theoretically and statistically related). However, Models 5 and 7 failed to converge across
 50

attempts. Both the two-factor adaptive and maladaptive (Model 6) and four-factor higher

order model (Model 8) had adequate fit based on RMSEA and CFI. Of the two fitted

models, Model 8 yielded a statistically significant improvement over Model 6. Thus, in

alignment to the replication models, the four-factor higher order structure best

represented the sample.

Theoretically speaking, the components captured by the EvsD and PALS were

well aligned to the underlying theories that are foundational to the MEW. For instance,

goal orientation—captured by the Mastery Goal Orientation and Performance-Approach

Goal Orientation subscales in the PALS—aligns to AGT, SCT, and SDT, which were

foundational to the conceptualization of the MEW. Additionally, the subscales of the

adapted measures were statistically significantly correlated with subscales from the

MEW. Thus, it was expected that alternative measures would fit the four-factor structure

of Martin’s MEW model; a hypothesis that was supported by current findings (i.e., Model

8 demonstrated superior fit).

The findings from the adaptability models provided insight on the malleability of

the MEW, but should be interpreted with caution. Based on these preliminary findings, it

appears that the four-factor structure of the MEW is upheld even without the inclusion of

the MES measure. This is of interest for a number of reasons. First, the MEW has

previously relied solely on the MES measure, which is potentially problematic because

MES was developed by Martin’s research team. According to Maruyama (1998),

utilizing multiple measures for each construct is the most defensible way in which to

create viable theoretical models. Thus, while the MEW is referred to as a theoretical

model, it has not previously met the criteria; instead, it is more aligned to a model of the
 51

MES measure. Including measures that are not directly derived from the MEW provided

further validation of the MEW as a theoretical model by demonstrating its adaptability

across measures of engagement and motivation. Additionally, the adaptability models

included multiple sources of information; both student and teacher’s responses were

captured in the model. Models can be strengthened by utilizing multiple measures and

multiple sources to represent latent variables. Thus, these findings could enhance our

understanding and application of the MEW, allowing a better understanding of how this

theoretical model bridges multiple conceptualizations of engagement and motivation.

Future Research Directions and Limitations

While this study provided additional support for the four-factor higher order

MEW model across various measures of motivation and engagement, there are several

limitations that must be considered and recommendations for future research. First, the

current study ended up with a lower than expected sample size (n = 270 for replication

and n = 185 for adaptability). While this met the threshold of acceptability from some

perspectives, it would be considered lower than advisable based on other rules of thumb

(Brown, 2014). Hence, findings should not be interpreted outside the context of the

sample and methodology used in the current study, especially with the adaptability

models. Additional research is necessary to replicate these preliminary findings, ideally

with a larger sample size. The current study only included an elementary aged sample.

Therefore, there is a need for further replication with secondary students from researchers

not associated with Martin. Additionally, despite using a racially diverse sample, the

majority of participants were from a low-income background so there is still a need to

examine the impact of SES on engagement and motivation.

 52

The current study did not examine the directionality of the relationship between

positive and negative engagement and motivation. Future research should utilize

longitudinal methods to explore the potential cyclical relationship between these

constructs. While other models have explored the impact engagement and motivation

have on each other (e.g., DiPerna or Reeve’s models), the directionality within the MEW

has not yet been explored.

Another limitation was the use of iPads to gather responses across the PALS,

MES, and EvsD measures. While these measures are psychometrically sound, they are

typically completed in paper-pencil format and have not been validated in an online

format. Thus, while measurement equivalence was assumed, it is possible that the use of

an online survey completed on an iPad could have impacted the psychometric properties

of these measures. The international test commission recommends any new version of a

measure, even including a change in format, should be evaluated for psychometrics

separately. However, a number of studies have compared online to paper-pencil self-

reports and found no statistically significant differences between the formats and that

online versions can be used with confidence (e.g., Carlbring et al., 2007;

Holländare, Andersson, & Engström, 2010).

Another possible limitation of the current study is the reliance upon self-report to

measure student engagement and motivation. The current study attempted to balance this

by including teacher-ratings of students’ motivation and engagement, however most of

the measures were self-report. While self-report forms are the most common measures of

both engagement and motivation, there can be developmental issues depending on the

age range or cognitive ability of respondents, which can negatively impact validity
 53

(Fulmer & Frijters, 2009). Although, issues like social desirability may impact responses

to self-report measures (Devellis, 2003), research has indicated that social desirability

bias is not as problematic with low-risk items such as those found in engagement or

motivation questionnaires (e.g., Miller, 2012). However, there are still issues of

interpretation for both self-report and ratings; it is possible that the individual completing

a measure interpreted an item differently from how the researcher or author intended.

While this study replicated and examined the higher order structure of Martin’s

(2017) analysis, it did not examine the first order factors. Many of the observed variables

were strongly correlated. Thus, it is possible that the observed variables could load into

different first order factors. This is especially true for the adaptability models. While

Martin’s research team has extensively researched the first order factor structure, the

current study relied on theoretical alignment to categorize the subscales of the different

measures into higher-order factors. Measurement models were used to support theoretical

alignment, but alternate first-order structures were not explored. Lastly, this study was

largely theoretical in nature and lacked explicit alignment to school based interventions.

Future research should explore the impacts of these different elements of adaptive and

maladaptive engagement and motivation on students’ education.

Practical implications

Given the influence engagement and motivation have on academic achievement,

it is of particular interest to better understand these constructs and how they relate to one

another from both theoretical and applied perspectives. Findings from the current study

are of particular interest in practice because it will allow educators to better understand

students’ specific areas of need. In other words, altering the conceptualization of

 54

engagement and motivation to include adaptive and maladaptive components will allow

educators to create more precise and well-matched interventions and strategies for

working with students. For instance, it is possible for a student to have high adaptive

motivation, but also have high maladaptive motivation. Thus, while many educators

might automatically choose an intervention that focuses on increasing motivation in

general, a more targeted intervention for decreasing maladaptive motivation could be

more efficient and effective for that individual student. Future research should explore

this possibility further.

Conclusion

This study was the first replication, outside of the original research team, of

Martin and colleagues’ (2017) examination of the higher-order structure of the MEW.

Overall, results indicated that the four-factor higher order structure, which separated

adaptive engagement, adaptive motivation, maladaptive engagement, and maladaptive

motivation into individual constructs, demonstrated the best fit for a diverse, American,

elementary-aged sample, thus replicating findings from Martin. Secondarily, this study

aimed to examine the adaptability of the MEW by including additional measures of

motivation and engagement. Based on the hypothesized adaptability models, the four-

factor higher order structure demonstrated best fit despite not including the MES. This

supports the hypothesis that the MEW can be adapted to include alternate components of

positive and negative engagement and motivation, thus strengthening the notion that the

MEW is a theoretical model. However, there was no way to compare the fit of the

replication and adaptability models because of the method of estimation used for the

current analysis. It is recommended that future research continues to validate the structure
 55

and adaptability of the MEW and explore the impacts of various interventions on these

distinct components of motivation and engagement in order to better understand how this

theoretical model can be applied in schools.

 56

References

Alexander, K. L., Entwisle, D. R., & Horsey, C. S. (1997). From first grade forward:

Early foundations of high school dropout. Sociology of education, 87-107.

Ames, C. (1992). Classrooms: Goals, structure, and student motivation. Journal of

Educational Psychology, 84, 261–271. doi:10.1037/0022-0663.84.3.261

Ames, C., & Archer, J. (1988). Achievement goals in the classroom: Students' learning

strategies and motivation processes. Journal of Educational Psychology, 80, 260–

267. doi:10.1037/0022-0663.80.3.260

Appleton, J. J., Christenson, S. L., Kim, D., & Reschly, A. L. (2006). Measuring

cognitive and psychological engagement: Validation of the Student Engagement

Instrument. Journal of School Psychology, 44, 427–445.

doi:10.1016/j.jsp.2006.04.002

Archambault, I., Janosz, M., Morizot, J., & Pagani, L. (2009). Adolescent behavioral,

affective, and cognitive engagement in school: Relationship to dropout. Journal of

School Health, 79, 408–415. doi:10.1111/j.1746-1561.2009.00428.x

Bandura, A. (1986). The explanatory and predictive scope of self-efficacy

theory. Journal of Social and Clinical Psychology, 4, 359–373.

Bandura, A., & Schunk, D. H. (1981). Cultivating competence, self-efficacy, and intrinsic

interest through proximal self-motivation. Journal of Personality and Social

Psychology, 41, 586–598. doi:10.1037/0022-3514.41.3.586

Brophy, J. (1983). Conceptualizing student motivation. Educational psychologist, 18,

200–215. doi:10.1080/00461528309529274
 57

Brown, T. A. (2014). Confirmatory factor analysis for applied research. Guilford

Publications.

Browne, M.W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K.A.

Bollen, & J.S. Long (Eds.), Testing structural equation models (pp. 136-162).

Newbury Park, CA: Sage.

Carlbring, P., Brunt, S., Bohman, S., Austin, D., Richards, J., Öst, L. G., & Andersson,

G. (2007). Internet vs. paper and pencil administration of questionnaires

commonly used in panic/agoraphobia research. Computers in Human

Behavior, 23, 1421–1434.

Christenson, S. L., & Anderson, A. R. (2002). Commentary: The centrality of the

learning context for student's academic enabler skills. School Psychology

Review, 31, 378–393.

Christenson, S. L., Reschly, A. L., & Wylie, C. (Eds.). (2012). Handbook of research on

student engagement. New York, NY: Springer.

Csikszentmihalyi, M. (1997). Flow and the psychology of discovery and

invention. Harper Perennial, New York, 39.

Deci, E. L., & Ryan, R. M. (1985). The general causality orientations scale: Self-

determination in personality. Journal of Research in Personality, 19, 109–134.

doi:10.1016/0092-6566(85)90023-6

DeVellis, R.F. (2003). Scale development: Theory and applications (2nd ed). Thousand

Oaks, CA: Sage Publications.

 58

DiPerna, J. C., & Elliott, S. N. (1999). Development and validation of the academic

competence evaluation scales. Journal of Psychoeducational Assessment, 17,

207–225. doi:10.1177/073428299901700302

DiPerna, J. C., & Elliott, S. N. (2002). Promoting academic enablers to improve student

achievement: An introduction to the mini-series. School Psychology Review, 31,

293–297.

DiPerna, J. C., Volpe, R. J., & Elliott, S. N. (2002). A model of academic enablers and

elementary reading/language arts achievement. School Psychology Review, 31,

298–312.

Dweck, C. S. (2007). The perils and promises of praise. Educational Leadership, 65, 34–

39.

Eccles, J. S., Midgley, C., Wigfield, A., Buchanan, C. M., Reuman, D., Flanagan, C., &

Mac Iver, D. (1993). Development during adolescence: The impact of stage-

environment fit on young adolescents' experiences in schools and in

families. American Psychologist, 48, 90 –101. doi:10.1037/003-066X.48.2.90

Eccles, J. S., & Wigfield, A. (2002). Motivational beliefs, values, and goals. Annual

Review of Psychology, 53, 109–132.

doi:10.1146/annurev.psych.53.100901.135153

Elliott E. S., & Dweck C. S. ( 1988). Goals: An approach to motivation and achievement.

Journal of Personality and Social Psychology, 54, 5–12. doi:10.1037/0022-

3514.54.1.5
 59

Finn, J. D., & Rock, D. A. (1997). Academic success among students at risk for school

failure. Journal of Applied Psychology, 82, 221–234. doi:10.1037/0021-

9010.82.2.221

Finney, S. J., Pieper, S. L., & Barron, K. E. (2004). Examining the psychometric

properties of the Achievement Goal Questionnaire in a general academic

context. Educational and Psychological Measurement, 64, 365–382.

doi:10.1177/0013164403258465

Fisher, C. W., & Berliner, D. C. (1985). Perspectives on instructional time. New York,

NY: Longman.

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, 59–109.

doi:10.3102/00346543074001059

Fredricks, J. A., & McColskey, W. (2012). The measurement of student engagement: A

comparative analysis of various methods and student self-report instruments.

In Christenson, S. L., Reschly, A. L., & Wylie, C. (Eds.), Handbook of research

on student engagement (pp. 763-782). New York, NY: Springer.

Fredricks, J., McColskey, W., Meli, J., Mordica, J., Montrosse, B., & Mooney, K. (2011).

Measuring Student Engagement in Upper Elementary through High School: A

Description of 21 Instruments. Issues & Answers. REL 2011-No. 098. Regional

Educational Laboratory Southeast.

Fulmer, S. M., & Frijters, J. C. (2009). A review of self-report and alternative approaches

in the measurement of student motivation. Educational Psychology Review, 21,

219-246.
 60

Greenwood, C. (2002). Academic engagement: Current perspectives on research and

practice. School Psychology Review., 31, 328–349.

Greenwood, C. R., Terry, B., Marquis, J., & Walker, D. (1994). Confirming a

performance-based instructional model. School Psychology Review.

Grüne-Yanoff, T. (2013). Relations between theory and model in psychology and

economics. Perspectives on Science 21(2), 196-201. The MIT Press. Retrieved

March 12, 2018, from Project MUSE database.

Harter, S. (1981). A new self-report scale of intrinsic versus extrinsic orientation in the

classroom: Motivational and informational components. Developmental

Psychology, 17, 300–312. doi:10.1037/0012-1649.17.3.300

Holländare, F., Andersson, G., & Engström, I. (2010). A comparison of psychometric

properties between internet and paper versions of two depression instruments

(BDI-II and MADRS-S) administered to clinic patients. Journal of Medical

Internet Research, 12.

Hu, L., & 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, 1-55.

Liem, G. A. D., & Martin, A. J. (2012). The Motivation and Engagement Scale:

Theoretical framework, psychometric properties, and applied yields. Australian

Psychologist, 47, 3–13.

Maehr, M. L., & Nicholls, J. G. (1980). Culture and achievement motivation: A second

look. In N. Warren (Ed.), Studies in cross cultural psychology. San Diego, CA:

Academic Press.
 61

Marks, H. M. (2000). Student engagement in instructional activity: Patterns in the

elementary, middle, and high school years. American Educational Research

Journal, 37, 153-184.

Martin, A. J. (2007). Examining a multidimensional model of student motivation and

engagement using a construct validation approach. British Journal of Educational

Psychology, 77, 413–440. doi:10.1348/000709906X118036

Martin, A. J. (2009). Motivation and engagement across the academic life span: A

developmental construct validity study of elementary school, high school, and

university/college students. Educational and Psychological Measurement, 69,

794–824. doi:10.1177/0013164409332214

Martin, A.J. (2016). The Motivation and Engagement Scale (16th Edition). Sydney,

Australia: Lifelong Achievement Group (www.lifelongachievement.com).

Martin, A. J., Ginns, P., & Papworth, B. (2017). Motivation and engagement: Same or

different? Does it matter?. Learning and Individual Differences, 55, 150–162.

doi:10.1016/j.lindif.2017.03.013

Martin, A. J., & Marsh, H. W. (2003). Fear of failure: Friend or foe?. Australian

Psychologist, 38, 31–38. doi:10.1080/00050060310001706997

Maruyama, G. (1998). Basics of structural equation modeling. Sage.

Meece, J. L., Blumenfeld, P. C., Hoyle, R. H. (1988). Students’ goal orientations and

cognitive engagement in classroom activities. Journal of Educational

Psychology, 80, 514–523. doi:10.1037/0022-0663.80.4.514

 62

Midgley, C., Maehr, M. L., Hruda, L. Z., Anderman, E., Anderman, L., Freeman, K. E.,

& Urdan, T. (2000). Manual for the patterns of adaptive learning scales. Ann

Arbor, 1001, 48109-1259.

Miller, A. L. (2012). Investigating Social Desirability Bias in Student Self-Report

Surveys. Educational Research Quarterly, 36, 30–47. Retrieved from

http://login.ezproxy.lib.umn.edu/login?url=http://search.ebscohost.com/login.aspx

?direct=true&AuthType=ip,uid&db=aph&AN=87515376&site=ehost-live

Murphy, P. K., & Alexander, P. A. (2000). A motivated exploration of motivation

terminology. Contemporary Educational Psychology, 25, 3–53.

Nakamura, J., & Csikszentmihalyi, M. (2009). Flow theory and research. Handbook of

positive psychology, 195-206.

Reeve, J. (2012). A self-determination theory perspective on student engagement.

In Christenson, S. L., Reschly, A. L., & Wylie, C. (Eds.), Handbook of research

on student engagement (pp. 149-172). New York, NY: Springer.

Reeve, J., & Lee, W. (2014). Students’ classroom engagement produces longitudinal

changes in classroom motivation. Journal of Educational Psychology, 106, 527–

540. doi:10.1037/a0034934

Reschly, A. L., & Christenson, S. L. (2006). Prediction of dropout among students with

mild disabilities a case for the inclusion of student engagement

variables. Remedial and Special Education, 27, 276–292.

doi:10.1177/07419325060270050301

Reschly, A. L., & Christenson, S. L. (2012). Jingle, jangle, and conceptual haziness:

Evolution and future directions of the engagement construct. In Christenson, S.

 63

L., Reschly, A. L., & Wylie, C. (Eds.), Handbook of research on student

engagement (pp. 3-19). New York, NY: Springer.

Rindskopf, D., & Rose, T. (1988). Some theory and applications of confirmatory second-

order factor analysis. Multivariate Behavioral Research, 23, 51–67.

Ryan, R. M., & Deci, E. L. (2000). Intrinsic and extrinsic motivations: Classic

definitions and new directions. Contemporary Educational Psychology, 25, 54–

67. doi:10.1006/ceps.1999.1020

Schiefele, U. (1991). Interest, learning, and motivation. Educational Psychologist, 26,

299–323.

Schumacker, R. E., & Lomax, R. G. (2010). A beginner's guide to structural equation

modeling. New York: Routledge.

Schunk, D. H., & Mullen, C. A. (2012). Self-efficacy as an engaged learner.

In Christenson, S. L., Reschly, A. L., & Wylie, C. (Eds.), Handbook of research

on student engagement (pp. 219-235). New York, NY: Springer.

Simons, D. J., Holcombe, A. O., & Spellman, B. A. (2014). An introduction to registered

replication reports at perspectives on psychological science. Perspectives on

Psychological Science, 9, 552-555.

Singh, K., Granville, M., & Dika, S. (2002). Mathematics and Science Achievement:

Effects of Motivation, Interest, and Academic Engagement. The Journal of

Educational Research, 95, 323–332. Retrieved from

http://www.jstor.org.ezp2.lib.umn.edu/stable/27542398
 64

Skinner, E. A., & Belmont, M. J. (1993). Motivation in the classroom: Reciprocal effects

of teacher behavior and student engagement across the school year. Journal of

Educational Psychology, 85, 571–581. doi:10.1037/0022-0663.85.4.571

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.

doi:10.1177/0013164408323233

Tabachnick, B. G., & Fidell, L.S. (2013). Using Multivariate Statistics (6th ed). New

Jersey: Pearson.

Veenman, S. (1984). Perceived problems of beginning teachers. Review of Educational

Research, 54, 143–178. doi:10.3102/00346543054002143

Walker, H. M., Ramsey, E., & Gresham, F. M. (2003). Heading off disruptive behavior:

How early intervention can reduce defiant behavior—and win back teaching

time. American Educator, 26, 6-45.

Way, J., Reece, A., Bobis, J., Anderson, J., & Martin, A. (2015). Improving student

motivation and engagement in mathematics through one-to-one interactions.

Mathematics Education Research Group of Australasia.

Weiner, B. (1985). An attributional theory of achievement motivation and emotion.

Psychological Review, 92, 548.

 65

Table 1

Demographic Information for the Sample

Characteristics Frequency Percentages

Biological Sex
Male 158 56%
Female 124 44%
Year in School
Third 80 28.4%
Fourth 97 34.4%
Fifth 84 29.8%
Sixth 21 7.4%
School
School 1 135 47.9%
School 2 147 52.1%
Race/Ethnicity
Asian 16 5.7%
Black/African American 190 67.4%
Caucasian 52 18.4%
Hispanic/Latino 30 10.6%
Other 4 1.4%
 66

Table 2

Comparing Means and Cronbach’s alphas for Subscales of MES Across Samples

M (SD) Cronbach’s 
Current Martin’s Current Martin’s
Factor Sub-Factor Sample Sample Sample Sample
Adaptive Self-Belief 86 (16.1) 86 (14.0) .83 .77
Motivation Learning Focus 84 (16.1) 86 (14.1) .82 .82
Valuing 84 (17.4) 89 (12.4) .79 .75
Adaptive Persistence 79( 16.0) 78 (15.2) .74 .77
Engagement Planning 73 (19.5) 74 (18.7) .79 .85
Task Management 79 (19.6) 79 (18.5) .82 .86
Maladaptive Anxiety 60 (20.8) 61 (18.5) .73 .66
Motivation Failure Avoidance 54 (22.7) 56 (23.9) .77 .85
Uncertain Control 48 (18.9) 49 (19.3) .68 .78
Maladaptive Self-Sabotage 40 (18.1) 41 (19.3) .74 .79
Engagement Disengagement 39 (16.6) 36 (15.8) .64 .70

Note. Average scores from the current sample were transformed from scores out of 5 to
converted scores out of 100 (as presented in the user manual).
 67

Table 3

Descriptive Statistics for Subscales of MES, EvsD, and PALS

Cronbach’s
Factor Measure Sub-Factor M SD Skew Kurtosis 
Adaptive MES Self-Belief 4.28 0.80 -1.29 1.45 .83
Motivation Learning Focus 4.21 0.80 -1.28 1.62 .82
Valuing 4.20 0.83 -1.31 1.47 .79
Scale Average 4.23 0.73 -1.27 1.44 .91
S. EvsD Emotional Engagement 3.05 0.74 -0.89 0.33 .80
T. EvsD Emotional Engagement 3.18 0.78 -0.76 -0.15 .95
PALS Academic Efficacy 4.05 0.87 -1.09 0.89 .79
Mastery Goal Oriented 4.30 0.82 -1.47 2.02 .81
Scale Average 3.67 0.68 -0.55 0.88 .86
Adaptive MES Persistence 3.97 0.80 -0.68 0.08 .74
Engagement Planning 3.64 0.97 -0.45 -0.51 .79
Task Management 3.95 0.98 -0.87 0.03 .82
Scale Average 3.85 0.78 -0.57 -0.20 .89
S. EvsD Behavioral Engagement 3.41 0.56 -0.94 0.48 .80
T. EvsD Behavioral Engagement 2.90 0.93 -0.45 -0.89 .95
PALS Academic Pressure 4.01 0.77 -0.86 0.75 .71
Maladaptive MES Anxiety 2.98 1.04 -0.08 -0.69 .73
Motivation Failure Avoidance 2.69 1.13 0.39 -0.71 .77
Uncertain Control 2.40 0.94 0.38 -0.32 .68
Scale Average 2.69 0.81 0.13 -0.14 .82
S. EvsD Emotional Disaffection 2.17 0.61 0.19 -0.33 .83
T. EvsD Emotional Disaffection 1.73 0.70 0.79 -0.03 .93
PALS Performance Avoid 2.56 0.98 0.25 -0.55 .49
Avoid Novelty 2.47 0.94 0.33 -0.33 .70
Low Achievement 2.11 0.77 0.51 -0.40 .63
Skepticism 1.86 0.89 1.17 1.05 .76
Performance Approach 2.66 1.07 0.42 -0.61 .77
Scale Average 2.25 0.62 0.61 0.46 .83
Maladaptive MES Self-Sabotage 2.02 1.17 0.84 -0.38 .74
Engagement Disengagement 1.76 1.07 1.45 1.37 .64
Scale Average 1.98 0.75 0.74 0.54 .78
S. EvsD Behavioral Disaffection 2.10 0.66 0.11 -0.63 .67
T. EvsD Behavioral Disaffection 2.12 0.90 0.27 -1.04 .93
PALS Self-Handicapping 2.18 0.95 0.58 -0.60 .77
Cheating Behavior 1.57 0.84 1.77 3.16 .79
Disruptive Behavior 2.42 1.08 0.46 -0.66 .83
Scale Average 2.06 0.73 0.72 0.12 .84

Note. (S. EvsD) refers to student’s self-ratings; (T. EvsD) refers to teacher’s ratings of
students.
 68

Table 4
Descriptive Statistics for MES Items

Frequencies
Strongly Strongly
Factor Sub-Factor Item Missing M SD Skew Kurtosis Disagree Disagree Neither Agree Agree
Adaptive Self-Belief 13 2 4.43 0.91 -1.81 3.09 1.9% 3.0% 8.7% 22.7% 63.6%
Motivation 23 5 4.23 1.04 -1.40 1.44 3.4% 3.4% 14.2% 24.1% 54.8%
33 4 4.19 1.01 -1.37 1.49 3.1% 4.6% 11.5% 32.1% 48.9%
40 4 4.29 0.98 -1.39 1.40 1.9% 4.2% 13.0% 24.8% 56.1%
Learning Focus 2 1 4.23 0.94 -1.30 1.51 1.9% 3.8% 12.5% 33.6% 48.3%
7 1 4.32 1.01 -1.71 2.63 3.8% 2.6% 9.4% 26.4% 57.7%
25 3 4.18 0.97 -1.31 1.72 3.0% 1.9% 15.2% 33.8% 46.0%
26 4 4.10 1.07 -1.15 0.72 3.4% 5.0% 16.4% 28.3% 46.9%
Valuing 4 2 3.88 1.10 -0.82 -0.03 3.8% 8.3% 19.7% 33.0% 35.2%
14 2 4.48 0.94 -2.00 3.49 1.9% 4.5% 6.1% 18.6% 68.9%
34 2 4.18 1.14 -1.45 1.37 6.1% 2.7% 13.3% 23.5% 54.5%
41 3 4.31 1.02 -1.60 2.07 3.4% 3.0% 11.8% 22.8% 58.9%
Adaptive Persistence 1 1 4.04 0.99 -1.17 1.31 3.4% 3.8% 15.1% 41.1% 36.6%
Engagement 9 1 3.99 1.14 -1.08 0.44 5.3% 5.7% 16.6% 29.8% 42.6%
28 4 3.93 1.04 -0.82 0.07 2.7% 7.3% 20.2% 34.0% 35.9%
36 6 3.93 1.09 -0.73 -0.38 2.3% 9.6% 20.8% 27.7% 39.6%
Planning 21 0 3.56 1.28 -0.54 -0.78 9.0% 13.2% 20.3% 28.2% 29.3%
27 2 3.95 1.13 -1.10 0.60 5.7% 5.7% 14.8% 36.0% 37.9%
30 3 3.69 1.28 -0.62 -0.71 7.6% 11.4% 21.7% 22.8% 36.5%
39 4 3.35 1.29 -0.29 -0.95 10.7% 15.3% 26.7% 23.3% 24.0%
Task 3 2 4.00 1.20 -1.08 0.23 6.1% 6.4% 15.9% 25.0% 46.6%
Management 17 2 3.86 1.19 -0.85 -0.17 5.7% 8.3% 19.3% 28.0% 38.6%
32 3 3.91 1.21 -0.98 0.03 6.5% 7.2% 16.7% 27.8% 41.8%
44 4 4.05 1.21 -1.19 0.48 6.9% 4.6% 15.6% 22.9% 50.0%
Maladaptive Anxiety 10 1 2.99 1.35 0.01 -1.18 17.4% 21.5% 23% 20.8% 17.4%
Motivation 19 1 3.2 1.45 -0.23 -1.29 18.9% 14.3% 19.6% 21.9% 25.3%
Continued
 69

Frequencies
Strongly Strongly
Factor Sub-Factor Item Missing M SD Skew Kurtosis Disagree Disagree Neither Agree Agree
Maladaptive Anxiety 37 7 2.91 1.36 0.00 -1.17 22.0% 15.8% 26.6% 20.1% 15.4%
Motivation 43 4 2.78 1.44 0.20 -1.27 27.1% 17.9% 22.5% 14.5% 17.9%
Failure 11 4 2.48 1.52 0.55 -1.17 39.3% 18.3% 15.3% 9.2% 17.9%
Avoidance 20 1 2.55 1.47 0.43 -1.22 35.5% 18.5% 17.4% 13.2% 15.5%
31 1 3.21 1.47 -0.21 -1.32 19.6% 13.2% 21.9% 17.4% 27.9%
38 2 2.51 1.40 0.50 -1.01 33.3% 20.8% 21.6% 10.2% 14.0%
Uncertain 6 2 2.63 1.33 0.31 -1.03 26.9% 21.6% 25.0% 15.2% 11.4%
Control 12 1 2.66 1.36 0.29 -1.11 27.2% 20.8% 23.8% 15.5% 12.8%
16 1 2.40 1.34 0.49 -0.96 37.0% 17.4% 23.8% 12.8% 9.1%
18 6 1.91 1.19 1.13 0.10 52.7% 22.3% 10.0% 11.2% 3.8%
Maladaptive Self-Sabotage 5 1 2.02 1.17 0.84 -0.38 47.5% 19.2% 20.4% 9.4% 3.4%
Engagement 24 2 1.94 1.20 1.09 0.14 52.3% 19.3% 15.9% 7.6% 4.9%
35 3 1.97 1.19 1.11 0.24 47.9% 25.9% 12.5% 8.4% 5.3%
42 2 2.06 1.21 0.89 -0.30 45.5% 22.7% 16.7% 10.2% 4.9%
Disengagement 8 0 1.76 1.07 1.45 1.37 56.0% 24.4% 10.5% 5.6% 3.4%
15 3 1.92 1.19 1.13 0.31 52.5% 19.0% 17.5% 5.7% 5.3%
22 0 2.11 1.24 0.95 -0.06 42.5% 24.8% 19.2% 6.0% 7.5%
29 2 1.98 1.22 1.11 0.25 49.6% 22.0% 15.9% 6.1% 6.4%
 70

Table 5
Descriptive Statistics for Student EvsD Items
Frequencies
Not at all Not Very Sort of
Factor Item Missing M SD Skew Kurtosis True True True Very True
Emotional 2 0 3.00 0.92 -0.74 -0.20 9.5% 13.7% 44.1% 32.7%
Engagement 9 4 3.02 0.89 -0.67 -0.26 7.3% 16.6% 42.9% 33.2%
13 1 2.99 0.98 -0.79 -0.34 12.6% 10.7% 41.6% 35.1%
22 2 3.21 0.95 -1.05 0.12 8.8% 10.3% 32.2% 48.7%
24 2 3.34 0.81 -1.09 0.58 3.4% 10.7% 34.5% 51.3%
Behavioral 1 0 3.66 0.621 -2.13 5.16 1.9% 2.3% 23.6% 72.2%
Engagement 8 1 3.57 0.64 -1.37 1.50 0.8% 5.7% 29.4% 64.1%
15 2 3.25 0.86 -1.09 0.59 6.1% 9.2% 38.3% 46.4%
18 2 3.36 0.80 -1.27 1.27 4.6% 6.9% 36.8% 51.7%
23 1 3.19 0.84 -0.81 -0.01 4.2% 14.5% 38.9% 42.4%
Emotional 4 1 1.66 0.88 1.02 -0.12 57.3% 22.5% 16.8% 3.4%
Disaffection 5 2 2.46 1.04 -0.10 -1.18 24.5% 21.8% 36.8% 16.9%
6 1 2.79 1.06 -0.41 -1.06 16.4% 19.5% 32.4% 31.7%
7 0 2.41 1.07 0.03 -1.27 26.6% 24.0% 30.8% 18.6%
11 1 2.06 1.04 0.48 -1.04 40.1% 25.2% 23.7% 11.1%
12 1 1.87 1.01 0.78 -0.69 49.6% 22.5% 19.1% 8.8%
16 1 2.18 1.08 0.34 -1.20 35.9% 24.8% 24.8% 14.5%
17 1 1.87 1.02 0.86 -0.52 49.2% 25.2% 15.3% 10.3%
19 2 2.62 1.11 -0.16 -1.30 21.5% 23.0% 28.0% 27.6%
20 1 2.25 1.11 0.30 -1.28 34.0% 25.2% 22.5% 18.3%
26 1 2.23 1.06 0.24 -1.22 33.2% 24.4% 28.6% 13.7%
27 0 1.68 0.97 1.22 0.26 60.1% 20.2% 11.8% 8.0%
Behavioral 3 2 1.63 0.94 1.26 0.32 63.2% 16.9% 13.8% 6.1%
Disaffection 10 0 1.47 0.81 1.68 1.95 69.2% 18.3% 8.7% 3.8%
14 1 2.32 1.14 0.15 -1.41 34.0% 19.5% 26.7% 19.8%
21 0 2.60 1.01 -0.30 -1.01 20.2% 19.0% 41.8% 19.0%
25 1 2.47 1.10 -0.06 -1.32 26.7% 20.2% 32.1% 21.0%
 71

Table 6
Descriptive Statistics for PALS Items
Frequencies
Not at all Somewhat Very
Factor Sub-Factor Item Missing M SD Skew Kurtosis True True True
Adaptive Academic 1 0 3.98 1.06 -0.96 0.45 3.9% 4.2% 21.2% 30.9% 39.8%
Motivation Efficacy 11 4 3.57 1.31 -0.50 -0.88 9.4% 12.2% 23.9% 21.2% 33.3%
52 6 4.10 1.25 -1.21 0.31 6.7% 5.5% 16.2% 13.8% 57.7%
56 4 4.42 1.00 -1.68 1.86 1.6% 6.7% 8.6% 14.9% 68.2%
58 6 4.27 1.17 -1.57 1.41 5.5% 4.7% 10.7% 15.0% 64.0%
Mastery Goal 9 2 4.33 1.04 -1.39 0.94 1.9% 5.4% 15.2% 12.8% 64.6%
Orientation 25 2 4.39 1.11 -1.72 1.89 3.9% 4.7% 11.7% 8.6% 71.2%
29 4 4.30 1.11 -1.48 1.19 3.5% 5.1% 14.1% 12.2% 65.1%
38 3 4.00 1.24 -1.00 -0.07 6.3% 6.6% 19.1% 17.2% 50.8%
49 4 4.47 1.01 -2.02 3.42 3.5% 2.0% 11.0% 10.6% 72.9%
Adaptive Academic 6 0 3.83 1.35 -0.87 -0.47 10.0% 6.9% 18.9% 17.8% 46.3%
Engagement Pressure 10 3 4.02 1.19 -0.99 0.02 5.1% 5.9% 21.1% 18.0% 50.0%
15 1 4.21 1.16 -1.34 0.80 4.7% 4.7% 16.3% 13.6% 60.9%
17 2 3.98 1.32 -1.05 -0.13 8.6% 7.0% 15.2% 16.7% 52.5%
19 2 3.98 1.22 -0.97 -0.06 5.8% 6.2% 20.6% 18.3% 49.0%
53 5 4.04 1.19 -1.03 0.08 5.1% 5.9% 20.1% 18.1% 50.8%
57 5 4.07 1.24 -1.14 0.11 5.9% 7.9% 14.6% 16.1% 55.5%
Maladaptive Performance 3 1 3.04 1.64 0.01 -1.60 28.7% 12.8% 18.2% 6.6% 33.7%
Motivation Avoid Goal 33 3 2.08 1.41 1.02 -0.35 53.5% 14.8% 13.7% 6.3% 11.7%
51 3 2.73 1.59 0.28 -1.45 35.5% 12.1% 20.3% 7.8% 24.2%
55 4 2.38 1.50 0.70 -0.95 42.4% 18.0% 16.9% 5.1% 17.6%
Performance 8 1 2.97 1.41 0.06 -1.20 21.3% 15.9% 29.1% 12.4% 21.3%
Approach 26 3 3.19 1.48 -0.17 -1.35 19.5% 14.5% 21.9% 15.6% 28.5%
41 2 2.38 1.46 0.66 -0.93 41.2% 17.1% 19.5% 7.0% 15.2%
45 2 2.4 1.49 0.56 -1.12 40.9% 15.6% 18.3% 8.9% 16.3%
48 5 2.37 1.54 0.66 -1.10 45.7% 14.6% 14.2% 8.3% 17.3%
Avoid 7 0 2.54 1.40 0.46 -1.01 32.4% 19.3% 24.3% 9.7% 14.3%
Novelty 20 4 2.12 1.33 1.00 -0.18 45.9% 22.7% 15.3% 5.9% 10.2%
Continued
 72

Frequencies
Not at Somewhat Very
Factor Sub-Factor Item Missing M SD Skew Kurtosis all True True True
Maladaptive Avoid 23 2 2.62 1.42 0.38 -1.07 31.1% 16.7% 27.6% 8.2% 16.3%
Motivation Novelty 35 4 2.63 1.35 0.35 -0.98 28.7% 18.4% 29.4% 10.2% 13.7%
40 3 2.50 1.42 0.48 -1.05 35.2% 17.6% 23.0% 10.2% 14.1%
Self 2 0 1.84 1.27 1.33 0.50 61.5% 13.5% 11.2% 6.9% 6.9%
Presentation 5 2 2.15 1.46 0.87 -0.74 54.5% 9.7% 14.8% 8.6% 12.5%
Low 21 4 1.90 1.37 1.33 0.37 61.6% 12.5% 11.4% 3.1% 11.4%
Achievement 24 2 2.76 1.57 0.23 -1.46 33.9% 13.2% 18.7% 11.3% 23.0%
27 2 2.13 1.40 0.93 -0.48 51.0% 14.0% 17.5% 5.8% 11.7%
37 4 1.97 1.27 1.12 0.11 53.7% 16.5% 16.5% 5.9% 7.5%
46 4 2.04 1.33 1.01 -0.24 52.9% 14.5% 16.9% 6.7% 9%
Skepticism 4 1 1.81 1.26 1.44 0.83 62.4% 14.3% 10.5% 5.0% 7.8%
Relevance 13 1 1.78 1.22 1.48 1.09 63.2% 13.2% 13.2% 3.5% 7.0%
28 2 1.86 1.29 1.33 0.55 61.1% 12.5% 14.0% 3.9% 8.6%
32 3 1.94 1.32 1.23 0.27 57.8% 13.3% 16.0% 3.1% 9.8%
36 2 1.91 1.34 1.32 0.41 59.1% 16.0% 10.1% 4.7% 10.1%
43 2 1.74 1.20 1.56 1.35 65.0% 12.8% 12.1% 3.5% 6.6%
Maladaptive Academic 12 1 2.22 1.44 0.76 -0.83 50.0% 10.5% 19.0% 8.5% 12.0%
Engagement Self 16 1 2.34 1.44 0.67 -0.91 42.6% 15.9% 19.8% 7.8% 14.0%
Handicap 18 1 2.34 1.51 0.66 -1.05 47.3% 10.9% 18.2% 7.8% 15.9%
42 2 2.21 1.48 0.85 -0.74 50.2% 13.6% 15.6% 5.8% 14.8%
44 3 1.95 1.29 1.19 0.26 55.9% 14.8% 16.8% 3.9% 8.6%
47 3 2.04 1.33 1.09 0.01 51.2% 18.0% 16.8% 3.5% 10.5%
Cheating 22 2 1.39 0.85 2.50 6.11 78.2% 10.9% 7.0% 1.9% 1.9%
Behavior 31 4 1.61 1.02 1.79 2.67 65.5% 17.3% 11.4% 2.4% 3.5%
39 2 1.70 1.12 1.63 1.80 63.0% 17.1% 11.3% 3.5% 5.1%
Disruptive 14 1 2.38 1.40 0.63 -0.84 38.4% 18.6% 22.5% 7.4% 13.2%
Behavior 30 4 2.77 1.48 0.25 -1.28 28.6% 16.9% 23.9% 10.2% 20.4%
34 2 2.44 1.42 0.57 -0.92 37.4% 16.7% 24.9% 6.2% 14.8%
50 4 2.41 1.30 0.54 -0.73 34.1% 19.2% 28.6% 7.8% 10.2%
54 5 2.14 1.37 0.92 -0.43 48.4% 17.3% 16.9% 6.3% 11.0%
 73

Table 7

Descriptive Statistics for Teacher EvsD Items

Frequencies
Not at all Not Very
Factor Item Missing M SD Skew Kurtosis True True Sort of True Very True
Emotional 6 0 3.12 0.86 -0.65 -0.41 4.3 18.7 37.4 39.6
Engagement 7 0 3.29 0.78 -0.90 0.30 2.7 11.8 39.6 46.0
8 0 3.19 0.83 -0.83 0.07 4.3 13.9 40.1 41.7
9 0 3.12 0.91 -0.81 -0.19 7.0 15.0 36.9 41.2
10 0 3.15 0.89 -0.85 -0.04 6.4 13.9 38.0 41.7
Behavioral 1 0 3.00 1.01 -0.66 -0.70 11.2 17.1 32.1 39.6
Engagement 2 0 3.15 0.94 -0.86 -0.24 7.5 15.0 32.6 44.9
3 0 3.00 1.00 -0.59 -0.82 9.6 20.9 29.4 40.1
4 0 2.52 1.08 0.08 -1.27 19.8 33.7 20.9 25.7
5 0 2.82 1.05 -0.36 -1.11 13.9 24.1 28.3 33.7
Emotional 16 1 2.01 0.98 0.51 -0.88 38.7 29.6 23.7 8.1
Disaffection 17 0 1.96 0.97 0.62 -0.73 41.2 29.9 20.9 8.0
18 0 1.60 0.81 1.22 0.76 56.7 29.4 10.7 3.2
19 0 1.63 0.86 1.21 0.53 58.3 25.1 12.3 4.3
20 1 1.63 0.89 1.31 0.81 58.1 26.3 9.7 5.9
21 0 1.51 0.75 1.48 1.73 62.0 27.8 7.5 2.7
22 0 1.61 0.91 1.37 0.81 62.0 21.4 10.2 6.4
23 0 1.79 0.95 0.97 -0.16 50.8 27.3 14.4 7.5
24 0 1.76 0.95 1.02 -0.03 51.9 27.3 13.4 7.5
25 0 1.82 1.02 0.95 -0.37 51.9 24.1 13.9 10.2
Behavioral 11 0 2.37 1.03 0.01 -1.19 26.7 24.1 34.8 14.4
Disaffection 12 0 1.93 0.98 0.66 -0.71 43.4 28.3 20.3 8.0
13 0 1.94 1.01 0.66 -0.78 44.4 26.2 20.3 9.1
14 5 2.25 1.09 0.23 -1.30 34.1 22.5 27.5 15.9
15 0 2.12 1.03 0.42 -1.04 35.8 28.3 24.1 11.8
 74

Table 8

Pearson’s Correlation Coefficients for Subfactors of Adaptive Motivation

1 2 3 4 5 6 7
1. Self-Belief – 0.70* 0.73* 0.56* 0.61* 0.52* 0.16*
2. Valuing – 0.69* 0.40* 0.65* 0.57* 0.15*
3. Learning Focus – 0.46* 0.60* 0.59* 0.17*
4. Academic Efficacy – 0.57* 0.38* 0.10
5. Mastery Goal Orientation – 0.54* 0.11
6. Student Emotional Engagement – 0.12
7. Teacher Emotional Engagement –

Note. (*) indicate p < 0.05

 75

Table 9

Pearson Correlation Coefficients for Subfactors of Adaptive Engagement

1 2 3 4 5 6
1. Planning – 0.63* 0.55* 0.31* 0.51* 0.11
2. Task Management – 0.63* 0.27* 0.56* 0.12
3. Persistence – 0.30* 0.61* 0.15*
4. Academic Pressure – 0.35* 0.06
5. Student Behavioral Engagement – 0.27*
6. Teacher Behavioral Engagement –

Note. (*) indicate p < 0.05

 76

Table 10

Pearson Correlation Coefficients for Subfactors of Maladaptive Motivation

1 2 3 4 5 6 7 8 9 10
1. Anxiety – 0.36* 0.40* 0.12 0.22* 0.18* 0.26* 0.04 0.48* 0.10
2. Failure Avoidance – 0.46* 0.32* 0.42* 0.20* 0.40* 0.17* 0.39* 0.10
3. Uncertain Control – 0.17* 0.25* 0.36* 0.48* 0.30* 0.43* 0.27*
4. Performance Approach – 0.47* 0.10 0.21* 0.12 0.17* 0.06
5. Performance Avoid – 0.22* 0.34* 0.18* 0.32* 0.08
6. Avoid Novelty – 0.43* 0.35* 0.31* 0.16*
7. Self Presentation – 0.41* 0.36* 0.12
8. Relevance Skeptic – 0.29* 0.33*
9. Student Emotional Disaffection – 0.25*
10. Teacher Emotional Disaffection –

Note. (*) indicate p < 0.05

 77

Table 11

Pearson’s Correlation Coefficients for Subfactors of Maladaptive Engagement

1 2 3 4 5 6 7
1. Self-Sabotage – 0.52* 0.40* 0.31* 0.32* 0.45* 0.25*
2. Disengagement – 0.35* 0.29* 0.36* 0.45* 0.19*
3. Self-Handicapping – 0.29* 0.39* 0.28* 0.24*
4. Cheating Behavior – 0.42* 0.36* 0.10
5. Disruptive Behavior – 0.48* 0.50*
6. Student Behavioral Disaffection – 0.31*
7. Teacher Behavioral Disaffection –

Note. (*) indicate p < 0.05

 78

Table 12
Standardized Item Factor Loadings and Errors for Model 4
Factors
Learning Task Failure Uncertain
Item Self-Belief Focus Valuing Persistence Planning Management Anxiety Avoid Control
13 0.80 (0.03)
23 0.84 (0.03)
33 0.81 (0.03)
40 0.86 (0.02)
2 0.75 (0.03)
7 0.81 (0.03)
25 0.84 (0.03)
26 0.79 (0.03)
4 0.69 (0.04)
14 0.82 (0.04)
34 0.73 (0.04)
41 0.88 (0.03)
1 0.59 (0.05)
9 0.77 (0.03)
28 0.69 (0.04)
36 0.82 (0.03)
21 0.70 (0.04)
27 0.70 (0.04)
30 0.84 (0.03)
39 0.81 (0.03)
3 0.65 (0.04)
17 0.81 (0.03)
32 0.88 (0.03)
44 0.87 (0.03)
10 0.70 (0.05)
19 0.60 (0.05)
Continued
 79

Factors
Learning Task Failure Uncertain Self-
Item Focus Valuing Persistence Planning Manage Anxiety Avoid Control Sabotage Disengagement
37 0.77 (0.05)
43 0.64 (0.05)
11 0.77 (0.04)
20 0.89 (0.04)
31 0.58 (0.06)
38 0.75 (0.05)
6 0.54 (0.05)
12 0.69 (0.04)
16 0.70 (0.04)
18 0.76 (0.05)
5 0.69 (0.05)
24 0.76 (0.04)
35 0.78 (0.04)
42 0.70 (0.05)
8 0.75 (0.04)
15 0.69 (0.05)
22 0.51 (0.06)
29 0.76 (0.04)
 80

Table 13
Model 4 First-Order Standardized Factor Loadings and Errors on Higher Order Factors

Higher-Order Factors
Adaptive Adaptive Maladaptive Maladaptive
First-Order Factors Motivation Engagement Motivation Engagement
Self-Belief 0.97 (0.02)
Learning Focus 0.92 (0.02)
Valuing 0.93 (0.02)
Persistence 0.96 (0.02)
Planning 0.81 (0.03)
Task Management 0.92 (0.02)
Anxiety 0.47 (0.06)
Failure Avoidance 0.57 (0.06)
Uncertain Control 1.24 (0.08)
Self-Sabotage 0.79 (0.04)
Disengagement 0.95 (0.04)
 81

Table 14
Standardized Item Factor Loadings and Errors for Model 8

Factors
Student Teacher Student Teacher Student
Emotional Emotional Academic Mastery Goal Behavioral Behavioral Academic Emotional
Item Engage Engage Efficacy Orientation Engage Engage Pressure Disaffection
Student EvsD2 0.67 (0.05)
Student EvsD9 0.65 (0.05)
Student EvsD13 0.74 (0.04)
Student EvsD22 0.91 (0.04)
Student EvsD24 0.65 (0.07)
Teacher EvsD6 0.89 (0.02)
Teacher EvsD7 0.83 (0.03)
Teacher EvsD8 0.98 (0.01)
Teacher EvsD9 0.98 (0.01)
Teacher EvsD10 0.99 (0.01)
PALS1 0.69 (0.06)
PALS11 0.50 (0.07)
PALS52 0.77 (0.05)
PALS56 0.85 (0.06)
PALS58 0.82 (0.05)
PALS9 0.68 (0.07)
PALS25 0.87 (0.06)
PALS29 0.67 (0.06)
PALS38 0.65 (0.07)
PALS49 0.90 (0.06)
Student EvsD1 0.76 (0.05)
Student EvsD8 0.75 (0.04)
Student EvsD15 0.60 (0.06)
Student EvsD18 0.92 (0.03)
Student EvsD23 0.81 (0.04)
Continued
 82

Factors
Student Teacher Student Teacher
Behavioral Behavioral Academic Emotional Emotional Performance Avoid Self Relevance
Item Engage Engage Pressure Disaffection Disaffection Avoid Novelty Presentation Skeptic
Teacher EvsD1 0.92 (0.02)
Teacher EvsD2 0.95 (0.01)
Teacher EvsD3 0.99 (0.01)
Teacher EvsD4 0.92 (0.01)
Teacher EvsD5 0.96 (0.01)
PALS6 0.37 (0.09)
PALS10 0.51 (0.09)
PALS15 0.51 (0.10)
PALS17 0.87 (0.09)
PALS19 0.58 (0.10)
PALS53 0.74 (0.08)
PALS57 0.38 (0.10)
Student EvsD4 0.58 (0.07)
Student EvsD5 0.85 (0.04)
Student EvsD6 0.33 (0.07)
Student EvsD7 0.82 (0.04)
Student EvsD11 0.79 (0.04)
Student EvsD12 0.44 (0.06)
Student EvsD16 0.27 (0.07)
Student EvsD17 0.53 (0.07)
Student EvsD19 0.53 (0.06)
Student EvsD20 0.78 (0.05)
Student EvsD26 0.40 (0.07)
Student EvsD27 0.67 (0.07)
Teacher EvsD16 0.96 (0.01)
Teacher EvsD17 0.95 (0.01)
Teacher EvsD18 0.69 (0.04)
Continued
 83

Factors
Student Teacher
Academic Emotional Emotional Performance Avoid Self Relevance Performance
Item Pressure Disaffection Disaffection Avoid Novelty Presentation Skeptic Approach
Teacher EvsD19 0.74 (0.04)
Teacher EvsD20 0.91 (0.02)
Teacher EvsD21 0.90 (0.02)
Teacher EvsD22 0.83 (0.03)
Teacher EvsD23 0.79 (0.03)
Teacher EvsD24 1.00 (0.01)
Teacher EvsD25 1.00 (0.01)
PALS3 0.39 (0.13)
PALS33 0.47 (0.13)
PALS51 0.46 (0.11)
PALS55 0.73 (0.11)
PALS7 0.50 (0.08)
PALS20 0.88 (0.08)
PALS23 0.58 (0.07)
PALS35 0.56 (0.08)
PALS40 0.63 (0.07)
PALS2 0.61 (0.10)
PALS5 0.58 (0.09)
PALS21 0.52 (0.10)
PALS24 0.29 (0.10)
PALS27 0.34 (0.10)
PALS37 0.85 (0.09)
PALS46 0.35 (0.09)
PALS4 0.60 (0.08)
PALS13 0.78 (0.08)
PALS28 0.81 (0.06)
PALS32 0.60 (0.07)
PALS36 0.69 (0.06)
PALS43 0.73 (0.07)
Continued
 84

Factors
Student Teacher
Self Relevance Performance Behavioral Behavioral Disruptive
Item Presentation Skeptic Approach Disaffection Disaffection Self-Handicap Cheating Behaviors
PALS8 0.49 (0.09)
PALS26 0.43 (0.08)
PALS41 0.73 (0.06)
PALS45 0.80 (0.06)
PALS48 0.93 (0.06)
Student EvsD3 0.66 (0.07)
Student EvsD10 0.78 (0.07)
Student EvsD14 0.31 (0.07)
Student EvsD21 0.75 (0.05)
Student EvsD25 0.72 (0.05)
TeacherEvsD11 0.89 (0.02)
Teacher EvsD12 0.80 (0.03)
Teacher EvsD13 0.96 (0.01)
Teacher EvsD14 0.87 (0.02)
Teacher EvsD15 0.98 (0.01)
PALS12 0.61 (0.08)
PALS16 0.83 (0.06)
PALS18 0.65 (0.06)
PALS42 0.54 (0.09)
PALS44 0.79 (0.06)
PALS47 0.61 (0.08)
PALS22 0.80 (0.07)
PALS31 0.93 (0.05)
PALS39 0.86 (0.06)
PALS14 0.74 (0.05)
PALS30 0.83 (0.04)
PALS34 0.78 (0.04)
PALS50 0.77 (0.04)
PALS54 0.75 (0.05)
 85

Table 15

Model 8 First-Order Standardized Factor Loadings and Errors on Higher Order Factors

Higher-Order Factors
Adaptive Adaptive Maladaptive Maladaptive
First-Order Factors Motivation Engagement Motivation Engagement
Student Emotional Engagement 0.58 (0.05)
Teacher Emotional Engagement 0.88 (0.04)
Academic Efficacy 0.50 (0.05)
Mastery Goal Orientation 0.51 (0.05)
Student Behavioral Engagement 0.57 (0.06)
Teacher Behavioral Engagement 0.66 (0.06)
Academic Pressure 0.24 (0.05)
Student Emotional Disaffection 0.58 (0.04)
Teacher Emotional Disaffection 0.77 (0.04)
Performance Avoid 0.39 (0.07)
Avoid Novelty 0.46 (0.05)
Self Presentation 0.46 (0.06)
Relevance Skeptic 0.54 (0.05)
Performance Approach 0.21 (0.06)
Student Behavioral Disaffection 0.73 (0.04)
Teacher Behavioral Disaffection 0.92 (0.02)
Self-Handicapping Behavior 0.46 (0.05)
Cheating Behavior 0.43 (0.06)
Disruptive Behaviors 0.68 (0.03)
 86

Figures

Adaptive Motivation Adaptive Engagement

Valuing Persistence

Learning Planning
Focus

Self- Task
Belief Management

Disengagement Anxiety

Failure
Avoidance

Self- Uncertain
Sabotage Control

Maladaptive Engagement Maladaptive Motivation

Figure 1. Motivation and Engagement Wheel (reproduced from Martin et al., 2007).
 87

Figure 2. Higher-order single-factor replication model of motivation/engagement.

 88

Figure 3. Higher-order two-factor replication model of adaptive and maladaptive traits.

 89

Figure 4. Higher-order two-factor replication model of motivation and engagement.

 90

Figure 5. Higher-order four-factor model of adaptive and maladaptive motivation and

engagement.
 91

Figure 6. Higher-order single-factor model of motivation/engagement.

 92

Figure 7. Higher-order two-factor model of adaptive and maladaptive traits.

 93

Figure 8. Higher-order two-factor model of motivation and engagement.

 94

Figure 9. Higher-order four-factor model of adaptive and maladaptive motivation and engagement.
 95

Figure 10. Four-factor replication model (Model 4) with standardized factor loadings
 96

Figure 11. Four-factor adaptability model (Model 8) with standardized factor loadings
 97

Appendices

Appendix A

Theoretical alignment for subscale categorization

Higher Sub Factor Example Item Theoretical

Order Factor Orientation
Adaptive Self-Belief If I try hard, I believe I can do my SCT
Motivation schoolwork wella
Learning Focus I feel very pleased with myself AGT; SDT
when I really understand what I am
taught at schoola
Valuing Learning at school is important to SDT
mea
Emotional When we work on something in SDT
Engagement class, I feel interestedb
Mastery Goal One of my goals in class is to learn AGT; SCT;
Orientation as much as I canc SDT
Academic Even if the work is hard, I can SCT
Efficacy learn itc
Adaptive Persistence If I can't understand my AGT; SCT
Engagement schoolwork at first, I keep going
over it until I understand ita
Planning Before I start an assignment I plan
out how I am going to do ita
Task When I study, I usually study in
Management places where I can concentratea
Behavioral I try hard to do well in schoolb Attribution
Engagement Theory
Academic When I’ve figured out how to do a AGT
Press problem, my teacher gives me
more challenging problems to
think aboutc
Maladaptive Anxiety When exams and assignments are SCT
Motivation coming up, I worry a lota
Failure Often the main reason I work at AGT; SDT
Avoidance school is because I don't want to
disappoint my parentsa
Uncertain I'm often unsure how I can avoid SCT;
Control doing poorly at schoola Attribution
Theory
Emotional When we work on something in SDT
Disaffection class, I feel boredb
 98

Performance- One of my goals in class is to AGT; SCT;

Avoid Goal avoid looking like I have trouble
Orientation doing the workc
Avoiding I would prefer to do class work that AGT;
Novelty is familiar to me, rather than work Attribution
I would have to learn how to doc Theory
Performance- One of my goals is to show others AGT; SCT;
Approach Goal that I’m good at my class workc SDT
Orientation
Self- I would avoid participating in class AGT; SCT
Presentation of if it meant that other students
Low would think I know a lotc
Achievement
Skepticism Doing well in school doesn’t Attribution
About improve my chances of having a Theory
Relevance of good life when I grow upc
School for
Future Success
Maladaptive Self-Sabotage I sometimes don't study very hard Attribution
Engagement before exams so I have an excuse if Theory
I don't do as well as I hopeda
Disengagement I often feel like giving up at SCT
schoola
Behavioral When I’m in class, I just act like
Disaffection I’m workingb
Academic Self- Some students fool around the Attribution
Handicapping night before a test. Then if they Theory
Strategies don’t do well, they can say that is
the reason. How true is this of
you?c
Cheating I sometimes copy answers from
Behavior other students during testsc
Disruptive I sometimes annoy my teacher
Behavior during classc

Note. (a) indicates MES; (b) indicates EvsD; (c) PALS.

 99

Appendix B

School-Based Professional Demographic Sheet

Please provide the following information about you:

Name: ______________________________ School: _______________________

Grade: ____________ # of Students in Class: ______________

Gender: Male
Female

Race: White
Black or African-American
American Indian or Alaska Native
Asian
Pacific Islander
Other: _______________

Ethnicity: Non-Hispanic Hispanic

Total years teaching: 1-5

6-10
11-15
16-20
20+

Years teaching this grade: 1-5

6-10
11-15
16-20
20+

Highest degree attained: Bachelor’s

Master’s
Master’s Plus
Doctorate
Other: _________________

Area(s) of certification: General Education

Special Education
Other: _________________
 100

Appendix C

Motivation and Engagement Scale – Junior School Items

1. If I can’t understand my schoolwork, I keep trying until I do

2. I feel very happy with myself when I really understand what I’m taught at school

3. I usually do my homework in places where I can concentrate

4. I’m able to use some of the things I learn at school in other parts of my life

5. Sometimes I don’t try hard at school so I can have a reason if I don’t do well

6. When I don’t do well at school I don’t know how to avoid that happening next time

7. I feel very happy with myself when I do well at school by working hard

8. Each week I’m trying less and less at school

9. If my homework is difficult, I keep working at it trying to figure it out

10. When I have a project to do, I worry about it a lot

11. The main reason I try at school is because I don’t want people to think that I’m dumb

12. When I get a good mark I often don’t know how I’m going to get that mark again

13. If I try hard, I believe I can do my schoolwork well

14. Learning at school is important

15. I don’t really care about school anymore

16. When I get a bad mark I don’t know how to avoid that happening next time

17. When I do homework, I get organized so I can do it well

18. I don’t know how to get good marks at school

19. I worry about getting bad marks in tests and projects

20. The main reason I try at school is because I don’t want people to think bad things

about me
 101

21. I usually have a plan for how to do my homework when I start it

22. I’m not involved in things like class activities and class discussion at school

23. If I don’t give up, I believe I can do schoolwork that is hard

24. I sometimes don’t work very hard at school so I can have a reason if I don’t do well

25. I feel very happy with myself when what I learn at school shows me how something

works

26. I feel very happy with myself when I learn new things at school

27. Before I start a project, I plan out how I am going to do it

28. When I’m taught something that doesn’t make sense, I spend time to try to

understand it

29. I’ve given up being interested in school

30. I have a plan for how to do my homework or projects when I start them

31. The main reason I try at school is because I don’t want to disappoint my parents

32. When I do homework, I try to find a place where I can do it well

33. If I have enough time, I believe I can do well in my schoolwork

34. What I learn at school will be useful in the future

35. I sometimes waste time the night before a test so I can have a reason if I don’t do well

36. I’ll keep working at difficult schoolwork until I’ve figured it out

37. When I do tests I don’t feel very good

38. The main reason I try at school is because I don’t want my teacher to think bad things

about me

39. I usually stick to a homework plan

40. If I try hard enough, I believe I can do all my schoolwork

 102

41. It’s important to understand what I’m taught at school

42. I sometimes leave homework until the last moment so I can have a reason if I don’t

do so well

43. I worry about school and schoolwork

44. When I do homework, I usually do it where I can concentrate best

 103

Appendix D

Engagement Versus Disaffection with Learning: Student Report Items

Behavioral Engagement

1. I try hard to do well in school.

2. In class, I work as hard as I can.

3. When I’m in class, I participate in class discussions.

4. I pay attention in class.

5. When I’m in class, I listen very carefully.

Emotional Engagement

1. When I’m in class, I feel good.

2. When we work on something in class, I feel interested.

3. Class is fun.

4. I enjoy learning new things in class.

5. When we work on something in class, I get involved.

Behavioral Disaffection

1. When I’m in class, I just act like I’m working. (–)

2. I don’t try very hard at school. (–)

3. In class, I do just enough to get by. (–)

4. When I’m in class, I think about other things. (–)

5. When I’m in class, my mind wanders. (–)

Emotional Disaffection

1. a. When we work on something in class, I feel bored. (–)

b. When I’m doing work in class, I feel bored. (–)

 104

c. When my teacher first explains new material, I feel bored. (–)

2. a. When I’m in class, I feel worried. (–)

b. When we start something new in class, I feel nervous. (–)

c. When I get stuck on a problem, I feel worried. (–)

3. When we work on something in class, I feel discouraged. (–)

4. Class is not all that fun for me. (–)

5. a. When I’m in class, I feel bad. (–)

b. When I’m working on my classwork, I feel mad. (–)

c. When I get stuck on a problem, it really bothers me. (–)

d. When I can’t answer a question, I feel frustrated. (–)

Note. Adapted from Wellborn (1991). The items added to the Emotional Disaffection

subscale can be used to tap the more differentiated disaffected emotions.

 105

Appendix E

Engagement Versus Disaffection with Learning: Teacher Report Items

Behavioral Engagement

1. In my class, this student works as hard as he/she can.

2. When working on classwork in my class, this student appears involved.

3. When I explain new material, this student listens carefully.

4. In my class, this student does more than required.

5. When this student doesn’t do well, he/she works harder.

Emotional Engagement

1. In my class, this student is enthusiastic.

2. In class, this student appears happy.

3. When we start something new in class, this student is interested.

4. When working on classwork, this student seems to enjoy it.

5. For this student, learning seems to be fun.

Behavioral Disaffection

1. When we start something new in class, this student thinks about other things. (–)

2. In my class, this student comes unprepared. (–)

3. When faced with a difficult assignment, this student doesn’t even try. (–)

4. In my class, this student does just enough to get by. (–)

5. When we start something new in class, this student doesn’t pay attention. (–)

Emotional Disaffection

1. a. When we work on something in class, this student appears to be bored. (–)

b. When doing work in class, this student looks bored. (–)

 106

2. a. When working on classwork, this student seems worried. (–)

b. In my class, this student is anxious. (–)

3. a. In class, this student seems unhappy. (–)

b. In my class, this student appears to be depressed. (–)

4. a. In my class, this student is angry. (–)

b. When working on classwork, this student appears frustrated. (–)

5. a. When I explain new material, this student doesn’t seem to care. (–)

b. When working on classwork in my class, this student seems uninterested. (–)

Note. Adapted from Wellborn (1991). The items added to the Emotional Disaffection

subscale can be used to tap the more differentiated disaffected emotions.

 107

Appendix F

Patterns of Adaptive Learning Scales: Student Survey Items

1. I'm certain I can master the skills taught in class this year.

2. I would avoid participating in class if it meant that other students would think I know a

lot.

3. It’s important to me that I don’t look stupid in class.

4. Even if I do well in school, it will not help me have the kind of life I want when I grow

up.

5. If other students found out I did well on a test, I would tell them it was just luck even if

that wasn’t the case.

6. When I’ve figured out how to do a problem, my teacher gives me more challenging

problems to think about.

7. I would prefer to do class work that is familiar to me, rather than work I would have to

learn how to do.

8. It’s important to me that other students in my class think I am good at my class work.

9. It’s important to me that I learn a lot of new concepts this year.

10. My teacher presses me to do thoughtful work.

11. I'm certain I can figure out how to do the most difficult class work.

12. Some students fool around the night before a test. Then if they don’t do well, they can

say that is the reason. How true is this of you?

13. My chances of succeeding later in life don’t depend on doing well in school.

14. I sometimes annoy my teacher during class.

15. My teacher asks me to explain how I get my answers.

 108

16. Some students purposely get involved in lots of activities. Then if they don’t do well

on

their class work, they can say it is because they were involved with other things. How

true is this of you?

17. When I’m working out a problem, my teacher tells me to keep thinking until I really

understand.

18. Some students look for reasons to keep them from studying (not feeling well, having

to

help their parents, taking care of a brother or sister, etc.). Then if they don’t do well on

their class work, they can say this is the reason. How true is this of you?

19. My teacher doesn’t let me do just easy work, but makes me think.

20. I don’t like to learn a lot of new concepts in class.

21. I wouldn’t volunteer to answer a question in class if I thought other students would

think I

was smart.

22. I sometimes copy answers from other students during tests.

23. I prefer to do work as I have always done it, rather than trying something new.

24. If I did well on a school assignment, I wouldn’t want other students to see my grade.

25. One of my goals in class is to learn as much as I can.

26. One of my goals is to show others that I’m good at my class work.

27. It’s very important to me that I don’t look smarter than others in class.

28. Doing well in school doesn’t improve my chances of having a good life when I grow

up.
 109

29. One of my goals is to master a lot of new skills this year.

30. I sometimes get into trouble with my teacher during class.

31. I sometimes cheat on my class work.

32. Getting good grades in school won’t guarantee that I will get a good job when I grow

up.

33. One of my goals is to keep others from thinking I’m not smart in class.

34. I sometimes behave in a way during class that annoys my teacher.

35. I like academic concepts that are familiar to me, rather than those I haven’t thought

about before.

36. Even if I am successful in school, it won’t help me fulfill my dreams.

37. If I were good at my class work, I would try to do my work in a way that didn’t show

it.

38. It’s important to me that I thoroughly understand my class work.

39. I sometimes copy answers from other students when I do my class work.

40. I would choose class work I knew I could do, rather than work I haven’t done before.

41. One of my goals is to show others that class work is easy for me.

42. Some students let their friends keep them from paying attention in class or from doing

their homework. Then if they don’t do well, they can say their friends kept them from

working. How true is this of you?

43. Doing well in school won’t help me have a satisfying career when I grow up.

44. Some students purposely don’t try hard in class. Then if they don’t do well, they can

say

it is because they didn’t try. How true is this of you?

 110

45. One of my goals is to look smart in comparison to the other students in my class.

46. One of my goals in class is to avoid looking smarter than other kids.

47. Some students put off doing their class work until the last minute. Then if they don’t

do

well on their work, they can say that is the reason. How true is this of you?

48. It’s important to me that I look smart compared to others in my class.

49. It’s important to me that I improve my skills this year.

50. I sometimes don’t follow my teacher’s directions during class.

51. It’s important to me that my teacher doesn’t think that I know less than others in

class.

52. I can do almost all the work in class if I don't give up.

53. My teacher makes sure that the work I do really makes me think.

54. I sometimes disturb the lesson that is going on in class.

55. One of my goals in class is to avoid looking like I have trouble doing the work.

56. Even if the work is hard, I can learn it.

57. My teacher accepts nothing less than my full effort.

58. I can do even the hardest work in this class if I try.