Content uploaded by Andrew J Martin
Author content
All content in this area was uploaded by Andrew J Martin on Dec 21, 2016
Content may be subject to copyright.
© 2013 Martin et al
0
Martin, A.J., Nejad, H.G., Colmar, S., & Liem, G.A.D. (2013). Adaptability: How students’ responses to
uncertainty and novelty predict their academic and non-academic outcomes. Journal of Educational
Psychology, 105(3), 728-746. DOI: 10.1037/a0032794.
This article may not exactly replicate the authoritative document published in the journal. It is not the copy
of record. The exact copy of record can be accessed via the DOI: 10.1037/a0032794.
© 2013 Martin et al
1
Adaptability:
How Students’ Responses to Uncertainty and Novelty Predict their Academic and Non-Academic
Outcomes
Andrew J. Martin, Harry G. Nejad, Susan Colmar, Gregory Arief D. Liem
Faculty of Education and Social Work
University of Sydney
Requests for further information about this investigation can be made to Professor Andrew J.
Martin, Faculty of Education and Social Work, A35 – Education Building, University of Sydney,
NSW 2006, AUSTRALIA. E-Mail: andrew.martin@sydney.edu.au. Phone: +61 2 9351 6273. Fax:
+61 2 9351 2606.
The authors would like to thank the Australian Research Council for funding this research.
© 2013 Martin et al
2
Adaptability:
How Students’ Responses to Uncertainty and Novelty Predict their Academic and Non-Academic
Outcomes
Abstract
Adaptability is defined as appropriate cognitive, behavioral and/or affective adjustment in the face
of uncertainty and novelty. Building on prior measurement work demonstrating the psychometric
properties of an adaptability construct, the present study investigates dispositional predictors
(personality, implicit theories) of adaptability, and the role of adaptability in predicting academic
(motivation, engagement, disengagement) and non-academic (self-esteem, life satisfaction, meaning
and purpose, emotional instability) outcomes. This longitudinal study (two time points, one year
apart), involving 969 adolescents from nine high schools, found that personality (conscientiousness
and agreeableness – positively; neuroticism – negatively) and implicit theories (effort-related
beliefs about intelligence – positively) significantly predicted adaptability (beyond the effects of
socio-demographics and prior achievement). Further, adaptability significantly predicted academic
(class participation, school enjoyment, and positive academic intentions – positively; self-
handicapping and disengagement – negatively) and non-academic (self-esteem, life satisfaction, and
meaning and purpose – positively) outcomes beyond the effects of socio-demographic factors, prior
achievement, personality, implicit theories, and two cognate correlates (buoyancy and self-
regulation). These findings hold implications for researchers and practitioners seeking to understand
and address young people’s responses to their changing academic and non-academic worlds.
Running Head: Adaptability
Keywords: adaptability; adolescence; motivation; engagement; self-regulation; personality
© 2013 Martin et al
3
Adaptability:
How Students’ Responses to Uncertainty and Novelty Predict their Academic and Non-Academic
Outcomes
Uncertainty and Novelty in Daily Life – and Throughout Life
Across a human life span, the world will undergo substantial changes on economic, geo-
political, socio-cultural, technological, medical, and other fronts (Hofacker, Buchholz, & Blossfeld,
2010; Tomasik, Silbereisen, & Heckhausen, 2010). Indeed, individuals’ own lives will be
characterized by frequent uncertainty and novelty. These include beginning school, adjusting to
new year groups and subjects at school, moving out of home, starting and changing jobs,
marriage/partnership, child-rearing/care-giving, and retiring from work – to name just some major
transitional milestones. Minor transitional elements typically involve the changing nature of tasks
and conditions throughout the day. Such changes can disrupt routines and create new circumstances
to which individuals must habituate (Pinquart & Silbereisen 2004; Tomasik & Silbereisen 2009;
Tomasik et al., 2010). How they deal with uncertainty and novelty has been central to formal
philosophizing as far back as figures such as Lao Tzu and the Buddha. With a focus on adolescents,
this study examines ‘adaptability’ as one potentially relevant psychological construct that may
assist them in their academic and non-academic lives.
The American Psychological Association’s (APA) definition of adaptability is “the capacity
to make appropriate responses to changed or changing situations; the ability to modify or adjust
one’s behavior in meeting different circumstances or different people” (VandenBos, 2007, p. 17).
Recent research developed and validated a measure of adaptability (the Adaptability Scale) to
assess individuals’ capacity to appropriately adjust and modify psycho-behavioral functions in
response to uncertain and novel circumstances, conditions and situations (Martin, 2012; Martin,
Nejad, Colmar, & Liem, 2012). The present study represents a substantial extension of this prior
measurement work. It builds longitudinal data into its design. This is important because this enables
© 2013 Martin et al
4
adjustments for prior variance in outcomes and thereby examines how adaptability predicts upward
and downward shifts in these outcomes. We also examine appropriate covariates and appropriate
controls for overlapping variance. In so doing, we seek to gain a fuller sense of unique variance
attributable to adaptability. Thus, the present study is something of a substantive-methodological
synergy (Marsh & Hau, 2007) in that methodological extension and refinement on prior adaptability
measurement research enables new and powerful substantive research questions to be addressed.
Consistent with Martin and colleagues (2012), we do so in the developmental context of
adolescence and the academic outcomes (motivation, engagement, disengagement) and non-
academic outcomes (self-esteem, sense of meaning, life satisfaction, emotional instability) relevant
to this stage of development. We focus on adolescence because development through this stage of
life presents many experiences of uncertainty and novelty. These experiences require individuals to
adjust and modify appropriate functions to maintain healthy development (Heckhausen & Schulz,
1995); thus adaptability is particularly pertinent during adolescence.
Adaptability
As noted above, adaptability has been described as an individual’s behavioral adjustments and
modifications to uncertain and novel circumstances and conditions (VandenBos, 2007). Recently,
this concept was expanded to consider adaptability in terms of appropriate cognitive, behavioral
and/or affective adjustments in the face of uncertainty and novelty (Martin et al., 2012). Cognitive
adjustment refers to modifications in thinking to deal with new and uncertain demands. Behavioral
adjustment refers to modifications in the nature, level and degree of behavior to deal with new and
uncertain situations and conditions (Heckhausen & Schulz 1995; Heckhausen et al., 2010).
Affective adjustment is considered in terms of “emotional response-tendencies [that] may be
modulated” (Gross, 1998, p. 3; see also Pekrun, 2012) to respond to environmental uncertainty and
novelty. Based on this tripartite perspective, Martin and colleagues developed the Adaptability
Scale (Martin et al., 2012). The Scale comprised items that each met the following criteria: (a)
appropriate cognitive, behavioral, or affective adjustment in response to (b) uncertainty and/or
© 2013 Martin et al
5
novelty that has (c) a constructive purpose or outcome. Exploratory and confirmatory factor
analysis (EFA and CFA) identified adaptability as a higher order factor subsumed by a reliable first
order cognitive-behavioral factor and a reliable first order affective factor; however, for operational
purposes, a single global adaptability factor was also deemed appropriate.
Relevant and Related Conceptualizing and Constructs
Alongside our tripartite approach to adaptability is conceptualizing from numerous theoretical
traditions that are a basis for further consideration of the construct and its part in young people’s
academic and non-academic outcomes. Whilst not intended to span the corpus of work in this area,
we map out relevant terrain by briefly discussing some salient frameworks and perspectives that
have informed our thinking and operationalization, including: the life-span theory of control, self-
regulated learning, models of adaptation, and adversity-related conceptualizing.
Life-span Theory of Control
According to the life-span theory of control, a major part of development involves the
individual adaptively adjusting goals to the opportunities and constraints in their ecology
(Heckhausen, Wrosch, & Schulz, 2010; Wrosch, Schulz, & Heckhausen, 2002). Control is framed
in terms of primary control (viz. the behavioral element of goal pursuit), secondary control (viz. the
cognitive element of goal pursuit) and – of relevance to adaptability – compensatory control
comprising alternative courses of action (compensatory primary control) and reappraising goals,
regulating aspirations and altering expectations (compensatory secondary control) (Tomasik et al.,
2010). Somewhat lacking in these control approaches is an explicit focus on affective adjustments –
one of the cornerstones of adaptability. Further, much of life-span theory’s emphasis is on goal
disengagement – whereas adaptability focuses on situations and circumstances from which the
individual cannot disengage (see Martin et al., 2012). Thus, we consider the present
operationalization of adaptability adds to recent life-span theory approaches.
Self-regulated Learning
© 2013 Martin et al
6
Self-regulated learning encompasses monitoring, directing, and controlling actions toward
learning goals, building expertise, and improving one’s skills (Boekaerts & Corno, 2005;
Zimmerman, 2002). Of recent relevance to adaptability is Winne and Hadwin’s (2008) four phases
of self-regulation that culminate in the adaptation phase in which the learner evaluates his/her
performance and identifies the necessary modifications needed to improve next time. We seek to
extend such work through our adaptability framework – extending from ‘classic’ self-regulatory
models of cognition and behavior to also encompass affect. In addition, whereas self-regulation is
about monitoring, directing and managing thought and behavior, adaptability is specifically about
adjustments and modifications to thought and behavior (and affect). Further, whereas self-
regulatory models tends to focus broadly on learning tasks and academic demands, the adaptability
construct is focused squarely on uncertainty and novelty and the purposeful adjustments and
modifications to deal with these. We see adaptability as a special case of negotiating situational
uncertainty and novelty that is compatible with broad theories of developmental regulation. Hence,
the present research work complements self-regulation research with the aligned construct of
adaptability, empirically tests the separability of adaptability and self-regulation, and explores their
respective contributions to academic and non-academic outcomes.
Adversity Constructs: Resilience, Buoyancy, and Coping
We separate uncertainty and novelty from adversity, difficulty and setback. We argue that
adaptability addresses the former – and factors such as buoyancy, resilience and coping address the
latter. Resilience has been defined as the process of successful adaptation despite challenging or
threatening circumstances (Howard & Johnson, 2000). Such circumstances are not minor or
insubstantial; they are characterized in terms of ‘acute’ and ‘chronic’ adversities that are ‘major
assaults’ on the developmental process (e.g., see Masten, 2001;). Whereas resilience is framed in
chronic and acute terms, buoyancy has been developed to address ‘everyday’ challenges (see
Martin & Marsh, 2009 for a review). These include study deadlines, difficult schoolwork, and a
poor result. Coping is another adversity-related construct defined in terms of cognitive and
© 2013 Martin et al
7
behavioral efforts to deal with specific demands that are appraised as difficult or perceived as
beyond the individual’s resources (e.g., see Frydenberg, 2008; Lazarus & Folkman, 1984). Recent
research has shown buoyancy and coping to represent distinct adversity-related constructs
predicting different outcomes (Putwain, Connors, Symes, & Douglas-Osborn, 2012). All three
constructs are separable from adaptability in that they all purposefully and specifically target
adversity and difficulty, whereas adaptability purposefully and specifically targets uncertainty and
novelty. It may be that adaptability is a special case of negotiating situational uncertainty and
novelty that is compatible with broad theories addressing adversity. Given this, we complement
buoyancy research with the aligned construct of adaptability, empirically test the separability of
adaptability and buoyancy, and explore their respective contributions to academic and non-
academic outcomes.
Models of Adaptation
Theory and research relevant to subjective well-being have also investigated how people
adapt to positive and negative life circumstances. One salient theory in this area is the adaptation
theory of well-being (Diener, Lucas, & Scallon, 2006). The adaptation framework is founded on an
automatic habituation model in which the individual reacts to deviations from his/her current
adaptation level (see Diener et al., 2006 for a review). Diener and colleagues have outlined a
number of refinements to the theory that are relevant to the present study. Two important
dimensions to this particular refinement are of particular pertinence. The first is that individuals will
vary in the specific strategies they use to adapt. In the present investigation this is investigated by
way of differences in cognitive, behavioral and affective adjustment in the face of uncertainty and
novelty. The second is that a number of individual difference factors predict adaptation. In the
present investigation, this signals the need to explore dispositional predictors (personality and
implicit theories) of adaptability.
Summary of Theorizing Relevant to Adaptability
© 2013 Martin et al
8
There are numerous theories and concepts that inform and align with adaptability. Although
there are a number of ways in which adaptability can be considered separable from other factors,
there are grounds for considering adaptability as a special case of negotiating situational uncertainty
and novelty that is compatible with broad theories of developmental regulation and adversity.
Conceptually, adaptability may be helpful in describing functional versus dysfunctional reactions to
novelty and uncertainty. We seek to contribute to current understanding of this developmental
terrain and further ‘round out’ current operationalization of related constructs. Accordingly, the
present study seeks to examine the empirical terrain explained by adaptability with a view to
understanding its role with relevance to academic and non-academic outcomes.
We also argue that because adaptability is considered a special case of negotiating situational
uncertainty and novelty, we can draw on these theories to specify conceptual arguments why and in
which way adaptability predicts academic and non-academic outcomes. Consistent with the life-
span theory of control (e.g., Wrosch, Schulz, & Heckhausen, 2002), compensatory control via
alternative forms of action and regulation of cognition increases the likelihood of individuals
effectively functioning in the context of opportunities and constraints in their environment. Self-
regulation theories (e.g., Winne & Hadwin, 2008) articulate the direction and management of
action, thought, and emotion leading to adaptive outcomes. Similarly, models of adaptation (e.g.,
Diener et al., 2006) describe successful habituation to deviations in current adaptation levels. Based
on these regulation and adaptation theories, we would predict that the regulatory and habituation
aspects of adaptability lead to the promotion of academic and non-academic outcomes.
Adaptability and a Process of Youth Development
Consistent with Buss and Cantor (1989; see also McCrae & Costa, 1996) and more recent
applications of their framework in the educational context (Martin, Marsh, & Debus, 2001), we
explore an adaptability process in which (a) individuals’ dispositions or characteristic orientations
impact (b) the strategies they use to negotiate demands in their environment, that impact (c) their
outcomes in this environment. This approach to human functioning identifies strategies and tactics
© 2013 Martin et al
9
as mediating the link between personality and various outcomes (Kyl-Heku & Buss, 1996; see also
McCrae & Costa, 1996). It also addresses how dispositions can be adaptively expressed to solve
problems and respond to different stimuli, circumstances, situations and conditions to bring about
positive outcomes (Cantor, 1990).
In this study, we examine a model in which dispositions and characteristic orientations take
the form of personality and implicit beliefs about intelligence and performance; strategy takes the
form of adaptability; and, outcomes take the form of academic (motivation, engagement,
disengagement) and non-academic (self-esteem, sense of meaning and purpose, life satisfaction,
emotional instability) factors. As explained below, we include personal contextual factors in the
form of socio-demographic and prior achievement factors, and we include buoyancy and self-
regulation alongside adaptability to explore their respective contributions to academic and non-
academic outcomes.
The design is a longitudinal one (from one academic year to another) and this allows us to
adjust for prior variance in academic and non-academic outcomes and thus examine how
adaptability predicts upward and downward shifts in these outcomes. The hypothesized model is
presented in Figure 1. As shown, personality and implicit theories predict adaptability; buoyancy
and self-regulation are located alongside adaptability as cognate correlates; adaptability (and
buoyancy and self-regulation) predict academic and non-academic outcomes, as do personality and
implicit theories; and, socio-demographic and achievement covariates are included through the
model. The rationale for key factors in this process is now described.
Predictors of Adaptability: Personality and Implicit Theories
Two major factor sets are proposed to represent individuals’ characteristic dispositions and
orientations as relevant to adaptability: personality and implicit theories intelligence and
performance. In their five factor theory, McCrae and Costa (1996) describe how basic tendencies
such as personality give rise to individuals’ adaptations that take the form of, inter alia, regulatory
processes. In reviewing evidence on personality and regulatory control, McCrae and Löckenhoff
© 2013 Martin et al
10
(2010) found that conscientiousness (positively) and neuroticism (negatively) related to control.
They suggested that neuroticism comprises poor impulse control and poor self-management
whereas conscientiousness comprises persistence, self-control, and effective decision-making.
Similarly, Hoyle (2010) observes that there are logical connections between key facets of
personality and regulatory factors and processes. In terms of deliberate and purposeful adjustment
of cognition and behavior, Hoyle argues that conscientiousness ought to play a dominant role,
particularly because conscientiousness is concerned with the ways individuals characteristically
manage their behavior. In contrast, individuals low in conscientiousness are not able to effectively
control behavior (Costa & McCrae, 1992). Further, de Raad and Schouwenberg (1996) found that
extraversion, conscientiousness and openness were significant factors in the positive development
and adaptive adjustment of one’s personal resources. Inferring from this presented theory and
research, it seems reasonable to posit that personality factors have a role in predicting adaptability.
Implicit theories refer to individuals’ beliefs about ability and effort, the extent to which they
see intelligence as something that is fixed (an ‘entity’ or ‘ability’ view) or something that is
malleable (an ‘incremental’ or ‘effort’ view), and the perceived link between ability, effort, and
performance (Dweck, 2000; Stipek & Gralinksi, 1996). In recent applications of implicit theories,
Yeager and Dweck (2012) explain that implicit theories might also predict responses to adversity
and challenge. They found that a view that intelligence can be developed or that personality
characteristics can be changed leads to resilience in academic and social settings respectively. They
argued that these implicit theories shape students’ goals, attributions, and learning strategies to
affect outcomes. Other work has shown implicit theories of intelligence to predict academic
trajectories during times of academic transitions and change through school (Blackwell,
Trzesniewski, & Dweck, 2007) and implicit theories of emotion predict adjustment to change in the
form of transition from high school to college (Tamir, John, Srivastava, & Gross, 2007).
Along similar lines, it may be that adaptability is also shaped by students’ beliefs about ability
and effort, their perception of the links between ability and effort and performance, and by
© 2013 Martin et al
11
implication, the extent to which they may or may not invest (cognitive, behavioral, and emotional)
effort to deal with uncertain and novel situations and circumstances. Specifically, individuals with
an incremental or effort view may see academic and non-academic outcomes as something that can
be addressed through cognitive, emotional, and/or behavioral modification (i.e., effortful
regulation). In contrast, individuals holding an entity or ability view may see their competence as
fixed and difficult to address, leading to less inclination to make psycho-behavioral adjustments.
We adopt the constructs proposed by Stipek and Gralinksi (1996; see also Martin, Marsh, & Debus,
2001) that explore the extent to which ability (‘ability-performance beliefs’) and effort (‘effort-
related beliefs’) are seen as determinants of intelligence and performance. Harnessing their
particular operationalization of implicit theories enables us to explore the role of both ability and
effort beliefs in predicting adaptability. Although ability-performance and effort-related beliefs
predominantly operate as two independent constructs, there are some children and young people
who see intelligence and outcomes as determined by both ability and effort (Dweck, Chiu, & Hong,
1995; Martin, Marsh, & Debus, 2001; Stipek & Gralinksi, 1996) and thus inclusion of both
accounts for this and controls for shared variance to identify the unique effects of ability-
performance and effort-related beliefs. Additionally, in Method we assess individual items to show
how some students can endorse both ability and effort beliefs.
Cognate Correlates: Buoyancy and Self-Regulation
It is important to understand adaptability in the context of related factors when modeling the
proposed process. For the purposes of the present study, we do so by including buoyancy alongside
adaptability in all modeling. We choose buoyancy over resilience because resilience deals with
chronic and acute adversity that is relevant to a relative minority of students, whereas buoyancy is
relevant to everyday adversity and difficulty that is relevant to all students (Martin & Marsh, 2009)
– as is adaptability. In addition, buoyancy is operationalized as a unidimensional factor that is
parsimonious to include in modeling; coping tends to be multidimensional and thus less
parsimonious when not a focus of the study. We also recognize the importance of including self-
© 2013 Martin et al
12
regulation to better understand the role of adaptability. Earlier in our introduction we distinguished
adaptability from self-regulation. However, we also noted that the two share developmental
regulatory terrain and that adaptability is a special case of negotiating situational uncertainty and
novelty that is compatible with broad theories of developmental regulation. We therefore include
self-regulation in our model to ascertain its variance relative to that of adaptability.
Outcome Factors to Investigate
In the context of adolescence, we propose that psycho-educational development comprises
positive and negative academic and non-academic outcomes.
Academic outcomes
Positive academic outcomes. For positive academic outcomes, we investigate students’
cognitive, behavioral and affective motivation and engagement (see Fredricks et al., 2004) as
operationalized by positive intentions, class participation, and enjoyment of school respectively. In
terms of positive intentions, it is feasible to consider that students who are better able to deal with
novelty and uncertainty are more willing to consider more ambitious and positive future selves in
the academic context. With regards to class participation, given the speed at which lessons progress
and the amount of content to cover in a given lesson (Marzano, 2003), there is a need for students to
adapt as new tasks, new task demands, and new task formats are presented to them. Students’
capacity to keep up and participate in an ongoing way will in part depend on their capacity to
constructively adjust thought, behavior and/or emotion along the way. Enjoyment of school is
another outcome of interest. Research has found adaptive self-modulation predicts subjective well-
being (e.g., Wrosch & Scheier, 2003) and so it may be that it also predicts academically-oriented
subjective well-being in the form of school enjoyment.
Negative academic outcomes. Based on need achievement and self-worth models of
motivation, negative outcome typologies can be characterized in terms of failure-avoidant students
and failure-accepting students (Covington, 1992; Martin et al., 2001). We suspect that students low
in adaptability (and thus less capable of negotiating uncertainty and novelty) may anticipate low
© 2013 Martin et al
13
efficacy and a greater likelihood of poor performance – and thus be more inclined to maneuver
defensively (e.g., self-handicap; Martin et al., 2001) or give up trying altogether (disengage). Hence,
from need achievement and self-worth motivation perspectives, we propose self-handicapping and
disengagement as two negative academic outcome factors in the adaptability process.
Non-academic outcomes
Positive non-academic outcomes. From a life-span theory perspective, the sense of control
gained from constructively adjusting cognition and behavior and engaging in alternative paths and
goals lay a foundation for an enhanced sense of meaning and purpose (Wrosch & Scheier, 2003).
Furthermore, the enhanced capacity to modulate cognitive, behavioral and/or affective resources is
also likely to be associated with factors such as self-esteem. For example, effective adjustment
should result in goal realization and fewer failure experiences, leading to a higher sense of self-
esteem and perceived self-worth (Wrosch & Scheier, 2003). Research has also associated life
satisfaction with broadened cognitive capacity and resources (Fredrickson, 2001), and this
broadening of capacity is aligned with the adaptability concept.
Negative non-academic outcomes. Life-span control research argues and finds that failure to
adopt alternative approaches to unattainable goals and maladaptive self-regulation is associated
with psychological distress and poor mental health outcomes (Wrosch et al., 2003). The present
study explores poor mental health in the form of emotional instability. Emotional instability refers
to individuals’ moodiness, worry, emotional confusion, and tendency to be unsettled and upset
(Marsh, 2007).
Socio-demographic and achievement covariates
Although not central to the substantive issues under focus, it is important to understand
adaptability controlling for numerous socio-demographic and achievement covariates. To the extent
that there exists shared variance between these covariates and adaptability – or between these
covariates and academic and non-academic outcomes, it is important to account for their presence
in the modeling. There are also theoretical bases for their inclusion. Major models of personality
© 2013 Martin et al
14
processes (e.g., Buss & Cantor, 1989; McCrae & Costa, 1996) suggest antecedent roles for
biological and background factors affecting individuals’ dispositional tendencies and characteristic
adaptations. Although no research has examined the predictive role of these factors on adaptability,
we briefly infer from related evidence to argue for their inclusion. Research (e.g., Ferrer &
McArdle, 2004) identifies gender influencing the development and adjustment of behavioral,
cognitive, and emotional capacities. In relation to age, Garcia-Coll and colleagues (1996) have
found it a positive predictor of children’s adaptive capacity to manage life demands. Socio-
economic status (SES) can shape one’s (behavioral, cognitive, affective) personal resources and
how these resources are adjusted and regulated (Moffitt, Caspi, Rutter, & Silva, 2001). Language
background is also relevant in defining and framing how people think, feel, and behave
(Organisation for Economic Co-operation and Development, 2006) and so may also be relevant to
adaptability. Finally, recent psychometric work found academic achievement significantly
correlated with cognitive-behavioral and affective adaptability (Martin et al., 2012). Taken together,
prior research and theory suggest the importance of including socio-demographic and achievement
factors; factors important to control for in the present study.
Aims of the Present Study
Building on prior measurement work demonstrating the psychometric properties of
adaptability, the present study investigates dispositional predictors (personality, implicit theories) of
adaptability and the role of adaptability in predicting academic (motivation, engagement,
disengagement) and non-academic (life satisfaction, self-esteem, meaning and purpose, emotional
instability) outcomes. The study also controls for variance attributable to buoyancy, self-regulation,
socio-demographics, and prior achievement. It is conducted among adolescents across two
academic years (2010-2011) at high school, allowing us to adjust for prior variance in academic and
non-academic outcomes and thus ascertain the role of adaptability in predicting upward and
downward shifts in outcomes over this time. Figure 1 demonstrates the major factors and processes.
Following from the review of relevant research and theory, it is hypothesized that conscientiousness
© 2013 Martin et al
15
will predict adaptability (no clear pattern has previously emerged on other personality factors), as
will ability-performance beliefs (negatively) and effort-related beliefs. It is hypothesized that
adaptability will positively predict class participation, school enjoyment and positive academic
intentions, and negatively predict self-handicapping and disengagement. Adaptability will also
positively predict self-esteem, life satisfaction and meaning and purpose, and negatively predict
emotional instability. We further hypothesize that these effects will be invariant across key sub-
groups (e.g., gender, ability, language background) and that adaptability will at least partially
mediate effects between personality and implicit theories and outcomes.
Method
Sample and Procedure
The sample comprises 969 high school students in junior high 11-14 years (54%) and senior
high 15-19 years (46%) from nine high schools in four major cities on the east coast of Australia.
Schools in the sample comprised students of mixed ability (but slightly higher in achievement and
socio-economic status than the national average). Four of the schools were co-educational, three
schools comprised girls only, and two schools comprised boys only. Just under half (48%) of the
respondents were female and 52% were male. The mean age of respondents was 14.40 (SD = 1.55)
years. A total of 16% of the sample spoke a language other than English at home. Ethics approval
was provided by the researchers’ university and parental consent was required. With few
exceptions, targeted students in attendance on the day of the testing participated in the survey.
Teachers administered the instrument to students during class. The rating scale was first explained
and a sample item presented. Students were asked to complete the instrument on their own and to
return the completed instrument at the end of class. Students completed the instrument twice, once
in Term 1 2010 and again in Term 1 2011, one year apart.
Taking into account students who could not have completed both surveys (i.e., students in
year 12 at Time 1 who had graduated by Time 2; students in year 7 at Time 2 who were new to the
school and not part of Time 1; new students joining the school in any given year group; students
© 2013 Martin et al
16
leaving the school in any given year group; and, students absent for any reason at either Time 1 or
Time 2), we estimated the response rate at 58%. This is 58% of the eligible sample at Time 2 (note
that N = 2,731 is the sample at Time 1 – not Time 2 – and which was reported in Martin et al.,
2012). To check that there were no significant differences between students participating at both
times and students participating only at one time, we performed tests of invariance that compared
the factor structure (factor loadings, correlations, residuals, and latent means) for unmatched and
matched students at 2010 and 2011. Comparing a model where all parameters were freely estimated
and one where all parameters were constrained across the unmatched and matched groups, there
was support for invariance (based on a change in CFI of no greater than .01, Cheung & Rensvold,
2002, and an RMSEA no greater than .015, Chen, 2007): Time 1 unconstrained, χ² (df = 5370) =
13066.884, p<.001; CFI = .91; RMSEA = .032, Time 1 constrained, χ² (df = 5738) = 13980.410,
p<.001; CFI = .90; RMSEA = .032 and Time 2 unconstrained, χ² (df = 5370) = 11247.807, p<.001;
CFI = .92; RMSEA = .031, Time 2 constrained, χ² (df = 5738) = 12448.022, p<.001; CFI = .91;
RMSEA = .032. Based on the comparable measurement properties for the two groups, we conclude
that the matched students in the present study can be considered broadly representative of students
at the nine schools.
Materials
Descriptive and psychometric statistics for each of the measures are detailed in Results and
Table 1.
Adaptability. Adaptability is defined as individuals’ adjustments of psycho-behavioral
functions in response to novel and/or uncertain circumstances, conditions and situations. The
Adaptability Scale comprises nine items, each item reflecting the following criteria: (a) appropriate
cognitive, behavioral, or affective adjustment in response to (b) uncertainty and/or novelty that has
(c) a constructive purpose or outcome. Items are rated on a 1 (‘Strongly Disagree’) to 7 (‘Strongly
Agree’) continuum. All nine items are presented in the Appendix. Prior cross-sectional
psychometric work with over 2,700 high school students has confirmed the psychometric status of
© 2013 Martin et al
17
the Adaptability Scale on the basis of distribution properties, reliability, factor loadings, invariance
as a function of key sub-groups (e.g., by gender and ethnicity), and correlations with external
validity constructs (Martin et al., 2012). Martin et al. (2012) advised that adaptability can be
operationalized as a higher order factor (indicated by cognitive-behavioral and affective first order
factors each indicated by six and three items respectively) or as a first order factor (indicated by
nine items). For parsimony, we adopt the latter operationalization.
Personality. Extraversion, openness to experience, neuroticism, conscientiousness, and
agreeableness (8 items per factor) were assessed using the 40-item International English Big-Five
Mini-Markers instrument (IEBM; Thompson, 2008). Participants rated the extent to which 40 trait
adjectives were accurate descriptors of themselves. Items for the IEBM are each represented by one
word in which the respondent rates themselves 1 (‘Very Inaccurate’) to 7 (‘Very Accurate’).
Sample words for each factor are as follows: ‘talkative’ (extraversion), ‘creative’ (openness),
‘moody’ (neuroticism), ‘efficient’ (conscientiousness), and ‘warm’ (agreeableness). Thompson
(2008) has previously demonstrated the reliability and predictive validity of the five factors
amongst adolescents.
Implicit theories (ability and effort beliefs). In the present study, implicit theories are
operationalized using Martin et al.’s (2001) adaptation of Stipek and Gralinksi’s (1996) ability-
performance beliefs and effort-related beliefs factors. Martin et al.’s adaptation involved changing
words such as “kids” to “people” and selecting the highest loading items. The ability-performance
beliefs factor holds ability is the determinant of intelligence irrespective of effort (e.g., “There isn’t
much some people can do to make themselves smarter”; “People can learn new things but how
smart they are doesn’t change”). The effort-related beliefs factor holds that intelligence can be
developed through the application of effort (e.g., “A person who works really hard can be very
smart”; “Any person could get smarter if they worked hard”). Five items comprise the ability-
performance beliefs scale and five items comprise the effort-related beliefs scale. Items were rated
on a 1 (‘Strongly Disagree’) to 7 (‘Strongly Agree’) scale. Close inspection of the items suggests
© 2013 Martin et al
18
students will tend to endorse one factor or the other. However, it is possible that some students can
endorse both. For example, a student agreeing with an effort view, can conceivably agree that there
is not “much” that “some” students can do to improve intelligence. Similarly, a student agreeing
with effort items can conceivably agree with an item that recognizes “people can learn new things”.
Indeed, Dweck et al., (1995) reported that about 15% of individuals may hold something of a mix
of entity and incremental beliefs. Based on the difference in chi square, CFI and RMSEA, the one-
factor approach to implicit theories yielded a significantly poorer overall model fit than the two-
factor (ability and effort beliefs) approach (χ²/df = 479.29/43, CFI = .03, RMSEA = .02).
Thus, we operationalized implicit theories as two factors. Importantly, however, by including them
in the one model, we control for their shared variance and therefore partial out empirical overlap to
identify unique variance attributable to ability and effort beliefs.
Buoyancy. Buoyancy is assessed using the Academic Buoyancy Scale (ABS; Martin & Marsh,
2008). Academic buoyancy (e.g., “I think I'm good at dealing with schoolwork pressures”) refers to
students’ ability to effectively deal with ‘everyday’ setback, challenge, adversity, and pressure in
the academic setting. The ABS is assessed through four items, rated from 1 (‘Strongly Disagree’) to
7 (‘Strongly Agree’).
Self-regulation. Self-regulation is assessed through planning and task management items from
the Motivation and Engagement Scale (MES; Martin, 2010b), an instrument that measures school
students’ motivation and engagement. Two of the eleven factors in the MES address self-regulation:
planning (e.g., “Before I start an assignment I plan out how I am going to do it”) and task
management (e.g., “When I study, I usually organize my study area to help me study best”). Each is
assessed through four items, rated from 1 (‘Strongly Disagree’) to 7 (‘Strongly Agree’), and for the
purposes of the present study is estimated as a single self-regulation factor – in line with previous
work showing they can be aggregated into a higher order factor (Martin, 2009).
Academic outcomes. Positive academic outcome factors were enjoyment of school, class
participation, and positive intentions. Enjoyment of school (e.g., “I enjoy being a student”), class
© 2013 Martin et al
19
participation (e.g., “I get involved in things we do in class”), and positive intentions (e.g., “I’d like to
continue studying or training after I complete school”) each comprised 4 items, which were rated on a
scale of 1 (‘Strongly Disagree’) to 7 (‘Strongly Agree’). These factors have been validated in previous
motivation and engagement research (Martin, 2009). Negative academic outcome factors are self-
handicapping (e.g., “I sometimes don’t study very hard before exams so I have an excuse if I don’t do so
well”) and disengagement (e.g., “I’ve pretty much given up being involved in things at school”). These
are drawn from the MES (Martin, 2010b). To each item, students rate themselves on a scale of 1
(‘Strongly Disagree’) to 7 (‘Strongly Agree’).
Non-academic outcomes. Non-academic measures comprised self-esteem, sense of meaning
and purpose, satisfaction with life, and emotional instability. To all measures, students were asked
to rate each statement on a 1 (‘Strongly Disagree’) to 7 (‘Strongly Agree’) scale. Self-esteem (e.g.,
“Overall, most things I do turn out well”) examined students’ overall evaluation of their self-worth.
The items were drawn from the general self-esteem scale of the Self-Description Questionnaire II
(SDQ-II; see Marsh, 2007). The general self-esteem scale has previously demonstrated high
reliability (Marsh, 2007). Sense of meaning and purpose (e.g., “My personal beliefs give meaning
to my life”) items were drawn from the World Health Organisation’s Quality of Life Instrument
(WHOQOL, 1998). It has previously shown sound reliability (WHOQOL, 1998). Satisfaction with
life (e.g., “In most ways my life is close to my ideal”) assesses participants’ satisfaction with their
life in general. The items were derived from the Satisfaction with Life Scale (Diener, Emmons,
Larsen, & Griffin, 1985). The scale has previously demonstrated good reliability (Pavot & Diener,
1993). Emotional instability (e.g., “I worry more than I need to”) examines respondents’ emotional
instability in the forms of worry, moodiness, and stress. The items are from the SDQ-II and
previously shown sound psychometric properties (Marsh, 2007).
Socio-demographics and prior achievement. Data were also collected on socio-demographic
characteristics including: gender, age, language background, and socio-economic status (SES). On
language background, participants were asked if they spoke English (0) or another language (1 –
© 2013 Martin et al
20
non-English speaking background, NESB) at home. Gender was coded 0 for females and 1 for
males. Age was retained as a continuous variable. Students’ SES was scored on the basis of their
home postcode using the Australian Bureau of Statistics (ABS) relative advantage/disadvantage
index, with higher scores reflecting higher SES. Academic achievement in this study is based on
students’ results in annual nation-wide assessment of literacy and numeracy (National Assessment
Program in Literacy and Numeracy, NAPLAN) administered by the Australian Curriculum and
Assessment and Reporting Authority (ACARA). It is a nationally standardized test for which school
students receive a score for each of literacy and numeracy. In the present study, an achievement
factor was formed through the average of literacy and numeracy scores.
Data Analysis
Descriptive statistics, reliability, and factor structure
Distributional and psychometric properties were assessed with confirmatory factor analysis
(CFA) to test factor structure, reliability (Cronbach’s alpha) to assess internal consistency, and
skewness and kurtosis to examine distribution properties. CFA was performed with Mplus version
7.0 (L. K. Muthén & B. O Muthén, 2012). In evaluating model fit, the root mean square error of
approximation (RMSEA) and the comparative fit index (CFI) are emphasized. For RMSEAs, values
at or less than .08 and .05 are taken to reflect close and excellent fits respectively (see Schumacker
& Lomax, 1996). For CFI, values at or greater than .90 and .95 are typically taken to reflect
acceptable and excellent fits respectively (McDonald & Marsh, 1990). Maximum likelihood with
robustness to non-normality and non-independence of observations (MLR; L. K. Muthén & B. O.
Muthén, 2012) was the method of estimation used. Due to the large number of personality items
and parameters relative to sample size, we created item parcels for personality factors, as suggested
by Schweizer (2012).
Although we did not have enough schools for multilevel modeling, and the study is focused
on intra-psychic constructs not expected to vary at class and school levels, we do adjust for the
clustering of students within schools through the Mplus ‘cluster’ command using the ‘complex’
© 2013 Martin et al
21
method. This procedure provides adjusted standard errors and so does not bias tests of statistical
significance due to clustering of students within schools (L. K. Muthén & B. O Muthén, 2012).
Using composite scores for structural equation modeling
Modeling longitudinal data using SEM can lead to a lack of stability of parameter estimation
and model fit statistics when the ratio of the sample size relative to the parameters to be estimated is
large (Holmes-Smith & Rowe, 1994). In the present study, our sample comprised 969 cases and the
hypothesized model is a relatively complex and longitudinal one. That is, given that there are more
than 200 observed variables involved in the hypothesized model, the number of parameters to be
estimated can be more than 20,000 if the traditional SEM is performed (based on formula p(p +
1)/2, with p = the number of observed variables; see Byrne, 2012). To address this problem, and
consistent with similar recent approaches to this issue (Liem, Ginns, Martin, Stone, & Herrett,
2012), we performed composite score-based SEM (Holmes-Smith & Rowe, 1994). Essentially,
composite score-based SEM is multivariate path analysis corrected for measurement error. In this
technique, the number of parameters is reduced because, instead of being predicted by its
constituent observed variables, each latent variable is represented by a weighted composite score
derived from a one-factor CFA. Proportional factor score regression weights (κ) generated from a
congeneric model solution are used to modify the weight of each item indicator before a composite
score is calculated (Holmes-Smith & Rowe, 1994). Factor score regression weights are particularly
important because they take into account individual item measurement error and their unique
(unequal) contributions to the composite score. The number of parameters in composite score-based
SEM are further reduced as the factor loading (λ) and measurement error variance (θ) of each latent
variable in the model are fixed with the values calculated using the weighted composite score
reliability of the factor under consideration. The factor loading is calculated through the square-root
of ρ and the measurement error variance is calculated by subtracting ρ from 1 (see Holmes-Smith &
Rowe, 1994). In this study, one factor CFAs were performed using Mplus 7.0 (L. K. Muthén & B.
O Muthén, 2012) with syntax provided by Raykov (2009).
© 2013 Martin et al
22
These composite scores represented latent factors in the central SEM. Consistent with Buss
and Cantor (1989) and more recent applications of their framework in the educational context
(Martin et al., 2001), we explore an SEM in which individuals’ dispositions or characteristic
orientations impact the strategies they use to negotiate demands in their environment, which in turn
impact their outcomes in this environment. Specifically, (a) personality and implicit theories
predict (b) adaptability (and also buoyancy and self-regulation), (c) personality, implicit theories,
and adaptability (and also buoyancy and self-regulation) predict academic and non-academic
outcomes, (d) all these factors are adjusted for prior variance (auto-regression; MacCallum &
Austin, 2000; Martin, 2011), and (e) socio-demographics and prior achievement are covariates,
predicting all factors at each point in the process. Figure 1 demonstrates this process.
In all parts of the process model, we sought to understand predictive effects beyond prior
variance in the predicted factors. In various forms (e.g., longitudinal models, experimental designs,
intervention research), this is a well-established technique to more conclusively establish the effects
of independent variables on dependent variables (MacCallum & Austin, 2000; Martin, 2011). Thus,
because we assessed the effects of socio-demographic and achievement variables on personality and
implicit theories and then the effects of all these on adaptability (and self-regulation and buoyancy)
and then the effects of all these on outcomes, we opted to partial out prior variance (i.e., account for
auto-regression) for all but the exogenous socio-demographic and achievement factors.
Subsidiary analyses
To further understand findings, three subsidiary analyses are conducted. The first is a test of
moderation. This involves constraining parallel beta parameters across groups. For example, by
constraining parallel beta parameters for boys and girls, we can ascertain if there is significant
model fit decline (based on a change in CFI greater than .01, Cheung & Rensvold, 2002; and a
change in RMSEA greater than .015, Chen, 2007), which would suggest that the beta parameters
for boys and girls are different. We also tested for differences in individual structural parameters as
a function of group membership (i.e., test for the statistical significance between two β paths). A
© 2013 Martin et al
23
second subsidiary analysis explored indirect effects of personality and implicit theory factors on
outcomes via adaptability. Indirect effects are based on bootstrapped standard errors (with 1000
draws) (MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002). MLR is not appropriate for
indirect bootstrapping models and so the present study implemented maximum likelihood (ML) as
the method of estimation here (L. K. Muthén & B. O Muthén, 2012). The third subsidiary analysis
explores alternative positioning of adaptability in our model, with adaptability operationalized as
(a) a dispositional attribute alongside personality and implicit theories predicting buoyancy, self-
regulation and outcomes and (b) buoyancy and self-regulation predicting adaptability and all three
predicting outcomes. Difference in model fit between the hypothesized model and the alternative
models is of interest here.
Missing data
Missing data pose problems for data analysis, particularly when it exceeds five percent (e.g.,
Graham & Hoffer, 2000). Research has identified potential problems with listwise, pairwise and
similar substitution methods (Graham & Hoffer, 2000), leading to recommendation of the
Expectation Maximization (EM) Algorithm, as operationalized in our study using LISREL 8.80
(Jöreskog & Sörbom, 2006). In fact, less than five percent of the data were missing and so the EM
Algorithm was employed as the approach to missing data.
Results
Descriptive Statistics, Reliability, and Confirmatory Factor Analysis
Table 1 present scale means (and standard deviations), distributions (skewness, kurtosis),
reliability, and mean factor loadings. Means and standard deviations are consistent with prior
research (Martin, 2009; Martin & Marsh, 2008). Skewness and kurtosis values indicate
approximately normal distributions. Reliability for all factors range between .75 and .92 (mean
reliability of .83), suggesting sound internal consistency. Confirmatory factor analysis yielded a
good fit to the data at Time 1, χ² = 6398.92, df = 2968, CFI = .91, RMSEA = .04 and Time 2, χ² =
© 2013 Martin et al
24
6369.74, df = 2968, CFI = .91, RMSEA = .03. Mean factor loadings are acceptable, ranging from
.67 to .90 (grand mean loading of .77), as shown in Table 1.
Additional CFAs were run to test model fit for CFA in which adaptability and buoyancy are
estimated as a single factor and model fit for CFA in which adaptability and self-regulation are
estimated as a single factor. To the extent that adaptability is deemed a special case of negotiating
situational demands, it is aligned with factors such as buoyancy and self-regulation, but must be
shown to be sufficiently separate from them. When integrating adaptability and buoyancy into one
factor, the fit is poorer than the hypothesized structure at Time 1 (χ² = 7227.84, df = 2991, CFI =
.88, RMSEA = .04) and Time 2 (χ² = 7076.96, df = 2991, CFI = .89, RMSEA = .04). When
integrating adaptability and self-regulation into one factor, the fit is worse than the hypothesized
structure at Time 1 (χ² = 7869.20, df = 2991, CFI = .87, RMSEA = .04) and Time 2 (χ² = 7723.12,
df = 2991, CFI = .87, RMSEA = .04). Taken together, these results provide a sound measurement
basis for the separation of adaptability from buoyancy and self-regulation and for assessing the
hypothesized process model.
Correlations Relevant to Adaptability
Table 2 presents correlations. For brevity, correlations central to hypothesized parameters
(i.e., between personality and implicit theory predictors and adaptability; between adaptability and
outcomes; between adaptability and cognate correlates) are reported here – all other correlations are
available in Table 2. Correlations show extraversion, openness, neuroticism (negatively),
conscientiousness, and agreeableness are all significantly correlated with adaptability. Ability-
performance beliefs (negatively) and effort-related beliefs are also significantly correlated with
adaptability. Adaptability is significantly correlated with buoyancy and self-regulation. In terms of
academic outcomes, adaptability is significantly correlated with: class participation, school
enjoyment, positive intent, self-handicapping (negatively), and disengagement (negatively). For
non-academic outcomes, adaptability is significantly correlated with: self-esteem, life satisfaction,
meaning and purpose, and emotional instability (negatively).
© 2013 Martin et al
25
Structural Equation Modeling
SEM was used to test Figure 1. The model fit the data well, χ² = 693.29, df = 342, CFI = .98,
RMSEA = .03. Table 3 presents all findings and Figure 2 presents central findings. Here, beta
coefficients central to the hypothesized model are summarized (i.e., between personality and
implicit theory predictors and adaptability; between adaptability and outcomes; between cognate
correlates and outcomes). Table 3 presents all other coefficients and explained variance.
Beyond variance explained by socio-demographics, prior achievement and auto-regression
(prior variance), three personality factors significantly predicted adaptability, as follows:
neuroticism (negatively), conscientiousness, and agreeableness. Adaptability was also predicted by
implicit theories by way of effort-related beliefs, but not by ability-performance beliefs.
After controlling for cognate correlates (buoyancy and self-regulation), adaptability explained
unique variance in outcomes. Specifically, beyond the variance explained by buoyancy, self-
regulation, personality, implicit theories, socio-demographics, prior achievement and auto-
regression (prior variance), adaptability significantly predicted eight of the nine outcome factors:
class participation, school enjoyment, positive intent, self-handicapping (negatively),
disengagement (negatively), self-esteem, life satisfaction, and meaning and purpose.
In different ways, the two cognate correlates significantly predicted outcomes after
controlling for adaptability, confirming that the three factors (adaptability, buoyancy, self-
regulation) are separable factors explaining unique variance in academic and non-academic
outcomes. Specifically, buoyancy significantly predicted four of the nine outcome factors: school
enjoyment, self-esteem, meaning and purpose (negatively), and emotional instability (negatively).
Meanwhile, self-regulation significantly predicted three of the nine outcome factors: school
enjoyment, self-handicapping (negatively), and disengagement (negatively).
Subsidiary Analyses
Investigating moderation: Invariance in predictive parameters
© 2013 Martin et al
26
We examined moderation with multi-group SEMs to test for invariance between parallel
adaptability parameters across groups of interest. In order to do so, it was important to first establish
invariance in measurement as a function of these groups. In all cases, there was no departure in
RMSEA greater than .015 (based on Chen, 2007) when all parameters were constrained. In relation
to CFI, the minimum criterion for invariance (invariance in factor loadings; Byrne, 2012) was
established in all models. Invariance in CFI was also established when correlations and loadings
were constrained and when residuals and loadings were constrained; but in four of ten models
(related to gender and achievement) there was minor departure in CFI greater than .01 (Cheung &
Rensvold, 2002) in the final model when all parameters were constrained at once. Thus, we
conclude predominant invariance, but advise that the reader is mindful of minor departures on some
fit indices for some sets of constrained parameters.
For our focal invariance tests, we then constrained 17 adaptability beta parameters in each of
five sets of invariance tests based on gender (boys, girls), age (junior high, senior high), language
background (English speaking, non-English speaking), socio-economic status (mean split of low
and high), and prior achievement (mean split of low and high). In comparison to when all
parameters were freely estimated, these invariance tests predominantly showed no reduction in CFI
> .01 and no change in RMSEA > .015 when the adaptability parameters were constrained across
groups: gender (unconstrained CFI = .97, RMSEA = .042; constrained CFI = .96, RMSEA = .050),
language background (unconstrained CFI = .97, RMSEA = .041; constrained CFI = .96, RMSEA =
.050), age (unconstrained CFI = .96, RMSEA = .045; constrained CFI = .96, RMSEA = .050),
socio-economic status (unconstrained CFI = .97, RMSEA = .044; constrained CFI = .95, RMSEA =
.052), and prior achievement (unconstrained CFI = .96, RMSEA = .041; constrained CFI = .95,
RMSEA = .049). Hence, notwithstanding a minor departure in CFI (but not RMSEA) for socio-
economic status, central adaptability parameters are not moderated by gender, age, language
background, SES, and prior achievement.
© 2013 Martin et al
27
In a further test, we compared each specific structural parameter value (transforming the β to
r using Peterson and Brown’s (2005) formula and comparing the derived coefficients using
Preacher’s (2002) formula) across the groups involved. Of 85 tests comparing parallel structural
parameters, only three were significantly different as a function of group membership: effort-related
beliefs → adaptability as a function of gender (female β = .20; male β = .42; difference in β, p <
.001), effort-related beliefs → adaptability as a function of ability (low β = .36; high β = .18;
difference in β, p < .01), and adaptability → self-esteem as a function of gender (female β = .35;
male β = .22; difference in β, p < .05). Thus, in addition to model invariance (above), there is
predominant invariance in specific parameters across groups.
Testing for indirect effects via adaptability
The second set of subsidiary analyses involved testing for indirect effects of personality and
implicit theories on outcomes via adaptability using bootstrapping (1000 draws). We found
adaptability significantly mediated the relationship between the following factors: neuroticism and
self-esteem (β = -.06; p < .001), life satisfaction (β = -.05; p < .01), meaning and purpose (β = -.07;
p < .001), enjoyment of school (β = -.03; p < .05), positive intentions (β = -.04; p < .01), class
participation (β = -.04; p < .01), self-handicapping (β = .03; p < .05), and disengagement (β = .03; p
< .05); between conscientiousness and self-esteem (β = .06; p < .001), life satisfaction (β = .06; p <
.001), meaning and purpose (β = .08 p < .001), enjoyment of school (β = .03; p < .05), positive
intentions (β = .04; p < .01), class participation (β = .04; p < .01), self-handicapping (β = -.03; p <
.05), and disengagement (β = -.03; p < .05); between ability-performance beliefs and self-esteem (β
= .03; p < .05) and meaning and purpose (β = .04; p < .05); and, between effort-related beliefs and
self-esteem (β = .09; p < .001), life satisfaction (β = .08; p < .001), meaning and purpose (β = .11; p
< .001), enjoyment of school (β = .04; p < .05), positive intentions (β = .06; p < .01), class
participation (β = .05; p < .01), self-handicapping (β = -.04; p < .01), and disengagement (β = -.04;
p < .05). In sum, not only did adaptability directly predict the bulk of outcome factors, it also
significantly mediated the relationship between many personality and implicit theory factors and
© 2013 Martin et al
28
outcomes. In fact, there were about the same number of significant indirect paths to outcomes via
adaptability at p < .001 as there were direct paths from personality and implicit theories to
outcomes at p < .001.
Exploring alternative model ordering
The final subsidiary analyses involved examining alternative model ordering. Although our
process model was ordered on conceptual grounds (see Buss & Cantor, 1989; McCrae & Costa,
1996), the data are available to explore alternative ordering of adaptability. The two conceptually
viable alternative orderings are where adaptability is considered a dispositional and characteristic
orientation (and thus modeled alongside personality and implicit theories) and where adaptability is
dependent of buoyancy and self-regulation (and thus regressed on buoyancy and self-regulation).
The first model yielded a higher chi square value than the hypothesized model (Δ χ² = 38.65) and
the second model also yielded a higher chi square value than the hypothesized model (Δ χ² =
10.79), providing some support for our positioning of adaptability in the process.
Discussion
Adaptability and Outcomes
For academic outcomes, it was hypothesized that adaptable students would be more ambitious
in their future plans (positive intent), more able to keep up with the rapid pace and variable nature
of lessons (class participation), experience more positive academic outcomes (school enjoyment),
and be less inclined to maneuver defensively (self-handicapping) or give up (disengagement)
(Martin et al., 2012). For non-academic outcomes, we drew on life-span and adaptation frameworks
to argue that adaptability should predict subjective well-being (Diener et al., 2006), sense of
purpose (Wrosch & Scheier, 2003), and psychological distress (negatively; Wrosch et al., 2003).
Consistent with hypotheses, adaptability uniquely predicted these outcomes in the expected
direction. Furthermore, adaptability significantly mediated the relationship between personality and
outcomes and between effort-related beliefs and outcomes. Importantly also, it explained unique
variance beyond the cognate academic factors of buoyancy and self-regulation. Thus, beyond the
© 2013 Martin et al
29
regulatory aspects of self-regulation and the adversity-related aspects of buoyancy, individuals’
cognitive, behavioral and/or affective adjustments to uncertainty and novelty play a distinct role in
their outcomes.
It is also noteworthy that these effects were beyond the effects of personality and implicit
theories. By including these trait factors, we could show that adaptability is not simply a proxy for
other well-known and well-established dispositional constructs. More broadly, by including a very
broad range of factors alongside adaptability in a multivariate set-up, we can argue against the
classic jingle (scales with the same label reflecting the same construct) and jangle (scales with
different labels measuring different constructs) fallacy (Marsh, 1994) by showing that adaptability
represents a distinct construct in academic and non-academic development.
It is also worth commenting on explained variance attributable to adaptability. Beyond the
effects of auto-regression, adaptability explained large variance in outcomes (up to 22%) and
beyond the effects of all factors, adaptability explained up to 6% additional variance. It is therefore
a significant factor, but clearly resides alongside other factors that also significantly predict
academic and non-academic outcomes. For example, self-regulation and buoyancy also explained
significant variance in outcomes. These findings might further support our contention that
adaptability may be a special case of negotiating situational uncertainty and novelty that is
compatible with broad theories of developmental regulation (and their component constructs). As
such, this study contributes to further understanding of this regulatory constellation and may help
further ‘round out’ the operationalization of regulatory constructs.
In terms of adaptability intervention, we suggest adaptability might be addressed in similar
ways to efforts addressing adversity-related constructs such as resilience. For example, Morales
(2000; see also Martin & Marsh, 2009) has proposed a resilience cycle that is aimed at sustaining
individuals’ ability to deal with risk on an ongoing basis. Adapting this framework, adaptability
intervention might comprise the following steps: 1) the individual is taught how to realistically and
effectively recognize uncertainty and novelty that might require adaptability, 2) he/she is taught
© 2013 Martin et al
30
how to make appropriate adjustments to behavior, cognition and/or affect, 3) these adjustments
assist the individual to deal with uncertainty and novelty, 4) he/she is encouraged to recognize the
value of these adjustments and then refine and/or progress them, and 5) this continuous refinement
and implementation of behavioral, cognitive and/or affective adjustment sustains the individual’s
ability to deal with ongoing uncertainty and novelty in academic and non-academic life. There is a
long line of cognitive-behavioral and affective intervention research demonstrating that students can
change cognition, behavior and affect to more effectively function in relevant performance domains
(e.g., Hattie, 2009; McInerney, McInerney, & Marsh 1997; O’Mara, Marsh, Craven, & Debus,
2006). These targeted interventions may be a basis for bringing about the type of adjustments
required to constructively respond to uncertainty and novelty.
Juxtaposition with Cognate Correlates
When considering adaptability in the context of buoyancy and self-regulation, it was clear that
adaptability accounted for significant variance. Notwithstanding this, buoyancy and self-regulation
did explain unique variance in outcomes as well as predict outcomes in notably different ways. For
example, buoyancy clearly mapped onto emotional instability in ways that adaptability did not: it
was the sole predictor of this outcome factor. Thus, it seems that when mental health (as indicated
by emotional instability) is more a focus, adversity-related constructs (such as buoyancy) are
perhaps more important. Indeed, along these lines, it was interesting to note that neuroticism (a
major mental health personality indicator) significantly predicted buoyancy and yielded larger paths
to buoyancy than to adaptability and self-regulation.
In terms of academic outcomes, it seems that both adaptability and self-regulation are
important. It may be that self-regulation is important for the ongoing direction and control of
executive and meta-cognitive functions for daily schoolwork (Boekaerts & Corno, 2005;
VandenBos, 2007; Zimmerman, 2002) – whereas adaptability may be relevant when new demands
and tasks are presented to students (Martin et al., 2012). Thus, across the two factors of adaptability
© 2013 Martin et al
31
and self-regulation, students are better placed to become, and then stay engaged in the course of the
school day and week.
Predictors of Adaptability
Personality
Salient personality predictors of adaptability were conscientiousness, agreeableness, and
neuroticism (negatively). Moreover, adaptability significantly mediated the relationship between
these three personality constructs and outcomes. In fact, there were about the same number of
significant indirect paths to outcomes via adaptability at p < .001 as direct paths from personality to
outcomes at p < .001. Further, our multivariate modeling (that controlled for shared variance among
personality factors) extended the bivariate correlational work by Martin et al. (2012) that found
adaptability correlated with all personality factors. The present study clarified unique personality
predictors after accounting for shared personality variance. Thus, consistent with Cantor (1990; see
also McCrae & Costa, 1996), dispositions can be adaptively expressed (in the case of
conscientiousness and agreeableness) to respond to different stimuli, circumstances, situations and
conditions to bring about positive outcomes. Conversely, dispositions may be maladaptively
expressed (in the case of neuroticism) to lead to negative outcomes.
These findings are consistent with what might be predicted by theory. McCrae and Costa’s
(1996) five-factor theory involves the regulatory and control processes that are shaped by
personality. Subsequently, researchers have shown that personality factors predict various
regulatory and control factors. Theory and research tends to converge on conscientiousness as a
major factor relevant to regulation and control (e.g., de Raad and Schouwenberg, 1996; Hoyle,
2010; McCrae & Löckenhoff, 2010). Conscientiousness is conceptualized as the personality factor
giving rise to adaptive self-management, persistence, effective decision-making, and control
(McCrae & Löckenhoff, 2010). These characteristics are clearly aligned with adaptability and this is
in keeping with our framing of adaptability as a special case of personal adjustments associated
with situational uncertainty and novelty. Consistent with McCrae and Löckenhoff (2010),
© 2013 Martin et al
32
neuroticism was also found to be a salient (negative) predictor of adaptability. Interestingly,
conceptualizing about neuroticism notes poor impulse control as a major attribute (McCrae &
Löckenhoff, 2010). The significant negative association between neuroticism and adaptability
suggests that adaptability is not a function of impulsive cognitive, behavioral and affective
modulation that might be characteristic of the neurotic individual. In combination with the positive
association between conscientiousness and adaptability, the negative neuroticism effect suggests
that adaptable students’ adjustments in cognition, behavior, and affect may be well considered,
deliberate, and purposeful. Including personality in our hypothesized model was important in
shedding this light on the specific nature of adaptability.
It therefore appears to be the case that some individuals are dispositionally (or
temperamentally) better placed for adaptability and others are not. This is important to know
because it will impact intervention designed to promote and sustain adaptability. For individuals
who may be low in conscientiousness and agreeableness or high in neuroticism, we point to the
review by Ginns and colleagues (2011) who describe how individuals can be taught to change
behavior, cognition and affect in the face of traits that might otherwise leave them ‘stuck’.
Practitioners, then, would do well to understand individuals’ trait-like profile as they direct
intervention at adaptability.
Implicit theories
Consistent with hypotheses, effort-related beliefs significantly predicted adaptability, whereas
ability-performance beliefs did not. Moreover, adaptability significantly mediated the relationship
between effort-related beliefs and outcomes. Individuals with effort-related beliefs see intelligence
and performance as malleable through effort whereas students with an ability-performance view see
intelligence and performance as relatively fixed (Dweck, 2000). We hypothesized that individuals
with an effort-related view would see academic and non-academic outcomes as something that can
be addressed through cognitive, emotional, and/or behavioral modification – and thus would be
more adaptable than individuals who see their competence as fixed and difficult to address (i.e., see
© 2013 Martin et al
33
less point in attempting cognitive, emotional and/or behavioral adjustment). It is also interesting to
note that the effort-related beliefs factor was the only substantive predictor of adaptability
moderated by students’ background characteristics. Specifically, we found that ability (a
significantly stronger path for low ability students) and gender (a significantly stronger path for
males) significantly moderated the effects of effort-related beliefs on adaptability.
Findings on students’ effort-related beliefs hold implications for intervention. Two lines of
work are relevant here. First, research into growth mindsets (Dweck, 2006) informs practical
approaches aimed at promoting effort-related beliefs about ability that help individuals see that
adjustment is possible and how to make such adjustments. Second, recent work has emphasized the
utility of growth goals and growth assessment (Martin & Liem, 2010). This growth perspective on
student academic and non-academic development is consistent with the adaptability construct and
thus adaptability may be an important factor to include in growth-related conceptual and applied
frameworks. Research into growth goals (Martin & Liem, 2010) has shown that personal best (PB)
goals are positively associated with academic outcomes. Similar growth approaches have been
recently proposed in the assessment domain (Anderman, Anderman, Yough, & Gimbert, 2010).
Finding that the effects of effort-related beliefs on adaptability were moderated by ability and
gender also provides intervention direction. For example, when seeking to target individual
students’ adaptability, intervention around growth mindsets, growth goals, and growth assessment
might be particularly useful for low achievers and males.
Socio-demographics and prior achievement
Although not central to hypothesizing, it is important to recognize socio-demographic and
achievement factors relevant to adaptability. Inclusion of these factors was important for three
reasons. First, it extends prior correlational work that did not control for shared variance among
these factors (Martin et al., 2012) (thus, we gain a sense of their unique effects). Second, inclusion
of these factors enables an understanding of adaptability purged of socio-demographic and
© 2013 Martin et al
34
achievement variance. Third, socio-demographic and achievement findings hold implications for
intervention by identifying the types of students who are likely or not likely to be adaptable.
In correlational findings, age was inversely associated with adaptability; younger adolescents
reported higher adaptability than older adolescents. In relation to regulation, the literature reports
mixed contentions regarding age. Some research suggests greater capacity to regulate personal
functions among older students (e.g., Garcia-Coll et al., 1996; Locke, 1996) whereas some research
suggests stability in the self-system among older students (e.g., Marsh, 2007). The present findings
suggest that older students are less capable of adjusting their cognition, behavior and affect. Perhaps
by this age they are solidifying their characteristic way of negotiating uncertainty and novelty. In
any case, educators might look to sustain students’ adaptability from early adolescence through to
later adolescence. Particularly given the uncertainties and novelties in the transition from school to
post-school life (Martinez, Martin, Liem, & Colmar, 2012), sustaining previously higher levels of
adaptability may be important. Prior achievement was also significantly associated with
adaptability. Achievement in high school requires an aggregation of numerous cognitive,
behavioral, and affective skills important for managing multiple demands, diffuse subject matter,
new teachers, different classes, diverse performance requirements, academic fear, and the like
(Martin, 2010a; Marzano, 2003). Viewed from this perspective, it is perhaps not surprising that
students who are able to develop these skills are also building their adaptability which also
comprises an aggregation of cognitive, behavioral, and affective skills to deal with situational
uncertainty and novelty. From an intervention perspective, low achievers might be identified and
supported in the cognitive, behavioral and affective factors and processes relevant to management
of multiple academic demands that will concomitantly assist their adaptability.
Summary of an Adaptable Profile
Based on the present findings, we can posit a profile of the adaptable student. In terms of
socio-demographics and prior achievement, this secondary school student is likely to be higher in
prior achievement and younger. Drawing on a previous study (Martin et al., 2012), this student is
© 2013 Martin et al
35
also likely to have parents/caregivers with higher levels of education. In terms of dispositional and
characteristic orientations, an adaptable student is likely to hold effort-related beliefs of intelligence
and performance (with the positive effects of effort-related beliefs more salient for low achievers
and males), be agreeable and conscientious, and unlikely to be neurotic. In relation to other
regulatory and adversity-based constructs, an adaptable student is more likely to self-regulate and
be buoyant in the face of everyday academic challenge and difficulty. Finally, this student’s
adaptability is likely to be demonstrated through higher levels of mental health in the form of life
satisfaction, self-esteem, and sense of meaning and purpose and higher levels of academic
motivation and engagement in the form of class participation, enjoyment of school, positive
academic intentions and low self-handicapping and disengagement. This profile represents a first
step in enabling practitioners to identify the types of students who are likely to be adaptable, assist
students who may not reflect some or all of these factors, and assess the effectiveness of their
efforts by examining academic and non-academic outcomes to which adaptability intervention
should ultimately connect. As discussed below, a second step is now to formally profile adaptability
using person-centered analytic approaches (e.g., cluster analysis) and to investigate intervention
relevant to the derived profiles.
Future Directions
There are some limitations to the present study that provide direction for future research. Data
were self-reported and thus there is a need to include other data sources, such as teacher and parent
observations and reports of students’ responses to uncertainty and novelty. There might also be
value including coping measures in future research. Although the present study included both
buoyancy and self-regulation as adaptability correlates, further research might seek to disentangle
any outstanding variance relevant to coping – particularly adaptive and maladaptive coping
techniques. Other personal characteristics (e.g., tolerance for ambiguity, need for closure, risk
aversion) relevant to uncertain and novel conditions might also be worth consideration. In line with
adversity-related research, it might also be important to understand the cumulative effect of
© 2013 Martin et al
36
uncertainty or novelty. For example, is there a critical point where too much uncertainty or novelty
represents adversity – and would this signal the need for resilience, buoyancy or coping? In recent
research, it seems the presence of two risk factors is sufficient to predict academic failure (Lucio,
Hunt, & Bornovalova, 2012) – how does this compare with cumulative instances of uncertainty and
novelty? Further, it might be important to understand the limits of adaptability. There may be
uncertain and novel situations and conditions where some level of stability and steadfastness is
required. If individuals are too often or markedly adjusting their cognition, affect and behavior, is
there a cost in terms of a stable sense of self and character?
Further, relevant to limitations and future directions, it must be noted that the adaptability
measure is domain general and not specific to a particular context. Perhaps the more focused the
measure is on a specific domain or situation the more it may impact outcomes in that domain.
Importantly, though, our domain general adaptability measure predicted academic outcomes as well
or better than the academically-oriented buoyancy and self-regulation measures predicted academic
outcomes. Hence, we suggest that even as a domain general measure, adaptability is noteworthy.
Our research is currently exploring adaptability, buoyancy, self-concept and academic achievement
in reading and mathematics among a sample of elementary students. These data will shed light on
domain specificity and the operation of adaptability (and its further juxtaposition with buoyancy
and self-regulation) with younger children. Additionally, because our measure of achievement was
taken before the survey period, we could only use it as a covariate. A post-survey measure of
achievement would enhance future research.
In terms of methodology, this study was a quantitative one leading to limits to what can be
understood through such data. Future research might involve qualitative data to better understand
how and when adaptability may operate. Another direction might be to collect data in the context of
a novel situation (e.g., in the laboratory) or at a time of transition (e.g., beginning a school year) to
assess the extent to which individuals who score higher on adaptability exhibit more effective
adjustments than those scoring lower on adaptability. Indeed, this might entail collecting real-time
© 2013 Martin et al
37
information from students which would enable contemporaneous quantitative and qualitative data at
specific times of uncertainty and novelty (e.g., when beginning a new school year). Recently,
Malmberg and colleagues (2012) demonstrated the efficacy of Personal Digital Assistants (PDAs)
as a means of collecting real-time data on learning and instruction from students. Finally, whereas
we have adopted a variable-centered approach to adaptability, future work might consider person-
centered approaches. This would involve identifying groups of students deemed as adaptable (or
not) with a view to identifying factors that determine their group membership. This has the
advantage of studying patterns of adaptability occurring ‘naturally’ and may also provide
opportunities for in-depth case study research.
Conclusion
Adaptability refers to appropriate adjustments in cognition, behavior and/or affect in response
to uncertain and novel circumstances, conditions and situations. Building on prior measurement
work demonstrating the psychometric properties of the Adaptability Scale, the present study
demonstrated dispositional predictors (effort-related beliefs, conscientiousness, agreeableness, and
neuroticism) of adaptability and ways in which adaptability predicted academic outcomes (class
participation, school enjoyment, positive academic intentions, self-handicapping, disengagement)
and non-academic outcomes (self-esteem, life satisfaction, meaning and purpose). Together, these
findings are suggestive of the dispositional profile of students who are adaptable and those who are
not. They are also indicative of the types of developmental outcomes that adaptability is likely to
predict, underscoring its importance on the developmental landscape. Findings hold theoretical and
empirical implications for researchers and practitioners seeking to better understand responses to
the uncertainty and novelty that are a reality of the world ahead for children and young people.
© 2013 Martin et al
38
References
Anderman, E., Anderman, L., Yough, M., & Gimbert, B. (2010). Value-added models of assessment:
Implications for motivation and accountability. Educational Psychologist, 45, 123-137.
Blackwell, L. S., Trzesniewski, K. H., & Dweck, C. S. (2007). Implicit theories of intelligence predict
achievement across an adolescent transition: A longitudinal study and an intervention. Child
development, 78, 246-263.
Boekaerts, M. & Corno, L. (2005). Self-regulation in the classroom: A perspective on assessment and
intervention. Applied Psychology: An International Review, 54, 199-231.
Buss, D.M., & Cantor, N. (1989). Personality psychology: Recent trends and emerging directions.
New York: Springer-Verlag.
Byrne, B. M. (2012). Structural equation modeling with Mplus: Basic concepts, applications and
programming. New York: Taylor & Francis/Routledge.
Cantor, N. (1990). From thought to behavior: ‘Having’ and ‘doing’ in the study of personality and
cognition. American Psychologist, 45, 735-750.
Chen, F.F. (2007). Sensitivity of goodness of fit indices to lack of measurement invariance. Structural
Equation Modeling, 14, 464-504.
Cheung, G. W., & Rensvold, R.B. (2002). Evaluating goodness-of-fit indexes for testing measurement
invariance (2002). Structural Equation Modeling, 9, 233-255.
Costa, P.T., & McCrae, R.R. (1992). Revised NEO Personality Inventory and NEO Five-Factor
Inventory professional manual. Odessa, FL: Psychological Assessment Resources.
Covington, M.V. (1992). Making the grade: A self-worth perspective on motivation and school reform.
Cambridge: Cambridge University Press.
de Raad, B., & Schouwenburg, H. C. (1996). Personality in learning and education: A review.
European Journal of Personality, 10, 303-336.
Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The Satisfaction with Life Scale.
Journal of Personality Assessment, 49, 71-75.
© 2013 Martin et al
39
Diener, E., Lucas, R., & Scollon, C. N. (2006). Beyond the hedonic treadmill: Revising the adaptation
theory of well-being. American Psychologist, 61, 305-314.
Dweck, C.S. (2000). Self-theories: Their role in motivation, personality and development.
Philadelphia: Psychology Press.
Dweck, C.S. (2006). Mindset: The new psychology of success. New York: Random House.
Dweck, C. S., Chiu, C. Y., & Hong, Y. Y. (1995). Implicit theories and their role in judgments and
reactions: A word from two perspectives. Psychological Inquiry, 6, 267-285.
Ferrer, E., & McArdle, J. (2004). An experimental analysis of dynamic hypotheses about cognitive
abilities and achievement from childhood to early adulthood. Developmental Psychology, 40, 935-952.
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.
Fredrickson, B. L. (2001). The role of positive emotions in positive psychology. American
Psychologist, 56, 218-226.
Frydenberg, E. (2008). Adolescent coping: Advances in theory, research and applications. London:
Routledge.
Garcia-Coll, C., Lamberty, G., Jenkins, R., McAdoo, H. P., Crnic, K., Wasik, B. H., & Vasquez-
Garcia, H. (1996). An interactive model for the study of developmental competencies in minority
children. Child Development, 67, 1891-1914.
Ginns, P., Liem, G.A.D., & Martin, A.J. (2011). The role of personality in learning processes and
learning outcomes in applied settings. In S. Boag & N. Tiliopoulos (Eds). Personality and
individual differences: Theory, assessment, and application. New York: Nova Science Publishers.
Graham, J. W., & Hoffer, S. M. (2000). Multiple imputation in multivariate research. In T. D. Little,
K. U. Schnable, & J. Baumert (Eds.), Modeling longitudinal and multilevel data: Practical issues,
applied approaches, and specific examples (pp. 201–218). Mahwah, NJ: Erlbaum.
Gross, J. J. (1998). The emerging field of emotion regulation: An integrative review. Review of
General Psychology, 2, 271-299.
© 2013 Martin et al
40
Hattie, J. (2009). Visible learning: A synthesis of over 800 meta-analyses relating to achievement.
London: Routledge.
Heckhausen, J., & Schulz, R. (1995). A life-span theory of control. Psychological Review, 102, 284-304.
Heckhausen, J., Wrosch, C., & Schulz, R. (2010). A motivational theory of life-span development.
Psychological Review, 117, 32-60.
Hofäcker, D., Buchholz, S., & Blossfeld, H.-P. (2010). Globalization, institutional filters and changing
life course patterns in modern societies. A summary of the results from the GLOBALIFE-project. In
R. K. Silbereisen & X. Chen (Eds.), Social change and human development: Concept and results
(pp. 101-124). London: Sage Publications.
Holmes-Smith, P., & Rowe, K.J. (1994). The development and use of congeneric measurement models
in school effectiveness research: Improving the reliability and validity of composite and latent
variables for fitting multilevel and structural equation models. Paper presented at the International
Congress for School Effectiveness and Improvement, Melbourne, Australia.
Howard, S., & Johnson, B. (2000). Resilient and non-resilient behavior in adolescents. In A. Graycar
(Ed.), Trends and issues in crime and criminal justice series (pp. 1-6). Canberra ACT: Australian
Institute of Criminology.
Hoyle, R. H. (2010). Personality and self-regulation. In R.H. Hoyle (Ed). Handbook of personality and
self-regulation. (pp. 1-18). Oxford: Wiley-Blackwell.
Jöreskog, K.G. & Sörbom, D. (2006). LISREL 8.80 for Windows. Lincolnwood, IL: Scientific Software
International, Inc.
Kyl-Heku, L.M., & Buss, D.M. (1996). Tactics as unit of analyses in personality psychology: An
illustration using tactics of hierarchy negotiation. Personality & Individual Differences, 21, 497-519.
Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer.
Liem, G.A., Ginns, P., Martin, A.J. Stone, B., & Herrett, M. (2012). Personal best goals and academic
and social functioning: A longitudinal perspective. Learning and Instruction, 22, 222-230.
© 2013 Martin et al
41
Locke, E. A. (1996). Motivation through conscious goal setting. Applied & Preventive Psychology, 5,
117-124.
Lucio, R., & Hunt, E., & Bornovalova, M. (2012). Identifying the necessary and sufficient number of
risk factors for predicting academic failure. Developmental Psychology, 48, 422-428.
MacCallum, R. C., & Austin, J. T. (2000). Applications of structural equation modelling in
psychological research. Annual Review of Psychology, 51, 201-226.
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison
of methods to test the significance of the mediated effect. Psychological Methods, 7, 83-104.
Malmberg, L-E., Little, T.D., Walls, T.A, & Martin, A.J. (2012). Primary school students’ intra- and
interpersonal beliefs about competence, effort, and difficulty. Submitted for Publication.
Marsh, H. W. (1994). Sport motivation orientations: Beware of jingle-jangle fallacies. Journal of Sport
and Exercise Psychology, 16, 347-380.
Marsh, H.W. (2007). Self-concept theory, measurement and research into practice: The role of self-
concept in educational psychology. Leicester, UK: British Psychological Society.
Marsh, H. W., & Hau, K-T. (2007). Applications of latent-variable models in educational psychology:
The need for methodological-substantive synergies. Contemporary Educational Psychology, 32,
151-171.
Martin, A.J. (2009). Motivation and engagement across the academic lifespan: A developmental
construct validity study of elementary school, high school, and university/college students.
Educational and Psychological Measurement, 69, 794-824.
Martin, A.J. (2010a). Building classroom success: Eliminating academic fear and failure. London:
Continuum.
Martin, A.J. (2010b). Motivation and Engagement Scale. Sydney: Lifelong Achievement Group
(www.lifelongachievement.com).
Martin, A.J. (2011). Prescriptive statements and educational practice: What can Structural Equation
Modeling (SEM) offer? Educational Psychology Review, 23, 235-244.
© 2013 Martin et al
42
Martin, A.J. (2012). Adaptability and learning. In N.M. Seel (Ed.). Encyclopedia of the Sciences of
Learning. Heidelberg: Springer.
Martin, A.J., & Liem, G.A. (2010). Academic Personal Bests (PBs), engagement, and achievement: A
cross-lagged panel analysis. Learning and Individual Differences, 20, 265-270.
Martin, A. J., & Marsh, H. W. (2008). Academic buoyancy: Towards an understanding of students’
everyday academic resilience. Journal of School Psychology, 46, 53-83.
Martin, A. J., & Marsh, H. W. (2009). Academic resilience and academic buoyancy: Multidimensional
and hierarchical conceptual framing of causes, correlates and cognate constructs. Oxford Review of
Education, 35, 353-370.
Martin, A.J., Marsh, H.W., & Debus, R.L. (2001). Self-handicapping and defensive pessimism:
Exploring a model of predictors and outcomes from a self-protection perspective. Journal of
Educational Psychology, 93, 87-102.
Martin, A.J., Nejad, H., Colmar, S., & Liem, G.A.D. (2012). Adaptability: Conceptual and empirical
perspectives on responses to change, novelty and uncertainty. Australian Journal of Guidance and
Counselling, 22, 58-81.
Martin, J. (2006). The meaning of the 21st Century. London: Transworld.
Martinez, C., Martin, A.J., Liem, G.A.D. & Colmar, S. (2012). A longitudinal analysis of physical and
psychological well-being amongst late adolescents: Exploring the transition from school to post-
school life. Australian Educational and Developmental Psychologist, 29, 17-43.
Marzano, R.J. (2003). What works in schools: Translating research into action. Alexandria, VI: ASCD.
Masten, A. S. (2001). Ordinary magic: Resilience processes in development. American Psychologist,
56, 227–238.
McCrae, R. R., & Costa, P. T., (1996). Toward a new generation of personality theories: Theoretical
contexts for the Five-Factor Model. In J. S. Wiggins (Ed.), The Five-Factor Model of personality:
Theoretical perspectives (pp. 51–87). New York: Guilford Press.
© 2013 Martin et al
43
McCrae & Löckenhoff (2010). Self-Regulation and the Five-Factor Model of personality traits. In
R.H. Hoyle (Ed). Handbook of personality and self-regulation. Oxford: Wiley-Blackwell.
McDonald, R. P., & Marsh, H. W. (1990). Choosing a multivariate model: Noncentrality and goodness
of fit. Psychological Bulletin, 107, 247-255.
McInerney, V., McInerney, D.M., & Marsh, H.W. (1997). Effects of metacognitive strategy training
within a cooperative group learning context on computer achievement and anxiety: An aptitude-
treatment interaction study. Journal of Educational Psychology, 89, 686-695.
Moffitt, T. E., Caspi, A., Rutter, M., & Silva, P. A. (2001). Sex differences in antisocial behavior:
Conduct disorder, delinquency, and violence in the Dunedin Longitudinal Study. Cambridge, UK:
Cambridge University Press.
Morales, E. E. (2000). A contextual understanding of the process of educational resilience: High
achieving Dominican American students and the ‘Resilience Cycle’. Innovative Higher Education,
25, 7–22.
Muthén, L.K. & Muthén, B.O. (2012). Mplus user’s guide. Los Angeles: Muthén & Muthén.
O’Mara, A. J., Marsh H. W., Craven, R. G., & Debus, R. (2006). Do self-concept interventions make a
difference? A synergistic blend of construct validation and meta-analysis. Educational
Psychologist, 41, 181-206.
Organisation for Economic Co-operation and Development, OECD (2006). Where immigrant students
succeed. Paris: OECD.
Pavot, W. & Diener, E. (1993). Review of the Satisfaction with Life Scale. Psychological Assessment,
5, 164-172.
Pekrun, R. (2012). Emotion regulation. In N. Seel (Ed.), Encyclopedia of the Sciences of Learning.
New York: Springer.
Peterson, R.A., & Brown, S.P. (2005). On the use of beta coefficients in meta-analysis. Journal of
Applied Psychology, 90, 175-181.
© 2013 Martin et al
44
Pinquart, M., & Silbereisen, R.K. (2004). Human development in times of social change: Theoretical
considerations and research needs. International Journal of Behavioral Development, 28, 289-298.
Preacher, K. J. (2002). Calculation for the test of the difference between two independent correlation
coefficients [Computer software]. Available from http://quantpsy.org/
Putwain, D.W., Connors, L., Symes, W., & Douglas-Osborn, E. (2012). Is academic buoyancy
anything more than adaptive coping? Anxiety, Stress and Coping: An International Journal, 25,
349-358.
Raykov, T. (2009). Evaluation of scale reliability for unidimensional measures using latent variable
modeling. Measurement and Evaluation in Counseling and Development, 42, 223-232.
Schumacker, R. E., & Lomax, R. G. (1996). A beginner's guide to structural equation modeling.
Mahwah, NJ: Erlbaum.
Schweizer, K. (2012). Editorial: On correlated errors. European Journal of Psychological Assessment,
28, 1-2.
Stipek, D., & Gralinski, H. J. (1996). Children’s beliefs about intelligence and school performance.
Journal of Educational Psychology, 88, 397-407.
Tamir, M., John, O. P., Srivastava, S., & Gross, J. J. (2007). Implicit theories of emotion: Affective
and social outcomes across a major life transition. Journal of Personality and Social Psychology,
92, 731-744.
Thompson, E.R. (2008). Development and validation of an international English Big-Five Mini-
Markers. Personality and Individual Differences, 45, 542-548.
Tomasik, M. J., & Silbereisen, R. K. (2009). Demands of social change as a function of the political
context, institutional filters, and psychosocial resources. Social Indicators Research, 94, 13–28.
Tomasik, M.J., Silbereisen, R.K., & Heckhausen, J. (2010). Is it adaptive to disengage from demands
of social change? Adjustment to developmental barriers in opportunity-deprived regions.
Motivation and Emotion, 34, 384-398.
© 2013 Martin et al
45
VandenBos, G.R. (2007)(Ed). American Psychological Association Dictionary of Psychology.
Washington DC: American Psychological Association.
WHOQOL Group. (1998). The World Health Organization Quality of Life Assessment (WHOQOL):
Development and general psychometric properties. Social Science and Medicine, 46, 1569-1585.
Winne, P.H., & Hadwin, A.F. (2008). The weave of motivation and self-regulated learning. In D.H.
Schunk & B.J. Zimmerman (Eds), Motivation and self-regulated learning: Theory, research, and
application (pp. 297-314). New York: Routledge.
Wrosch, C., & Scheier, M. F. (2003). Personality and quality of life: The importance of optimism and
goal adjustment. Quality of Life Research, 12, 59-72.
Wrosch, C., Scheier, M. F., Miller, G. E., Schulz, R., & Carver, C.S. (2003). Adaptive self-regulation
of unattainable goals: Goal disengagement, goal reengagement, and subjective well-being.
Personality and Social Psychology Bulletin, 29, 1494-1508.
Wrosch, C., Schulz, R., & Heckhausen, J. (2002). Health stresses and depressive symptomatology in the
elderly: The importance of health engagement control strategies. Health Psychology, 21, 340- 349.
Yeager, D.S., & Dweck, C.S. (2012). Mindsets that promote resilience: When students believe that
personal characteristics can be developed. Educational Psychologist, 47, 302-314.
Zimmerman, B.J. (2002). Achieving self-regulation: The trial and triumph of adolescence. In F.
Pajares & T. Urdan (Eds). Academic motivation of adolescents. Greenwich, CT: Information Age
Publishing.
© 2013 Martin et al
46
Figure 1. Hypothesized model of adaptability: Personality and implicit theory predictors and academic and non-academic outcomes
Notes. Bold lines represent paths of central interest. Dashed lines represent auto-regression paths. Dashed double-headed arrows represent correlations among factors.
Dashed ellipse represents cognate factors to disentangle adaptability variance from buoyancy and self-regulation.
Personality
Dispositions and
Characteristic Orientations
Academic and Non-
Academic Outcomes
Adaptability
(and Cognate Buoyancy and Self-
Regulation)
Implicit
Theories
Adaptability
Buoyancy and
Self-Regulation
Academic Outcomes
- Class participation
- Enjoy school
- Positive intentions
- Self-handicapping
- Disengagement
Non-Academic Outcomes
- Self-esteem
- Life satisfaction
- Meaning and purpose
- Emotional instability
Prior Variance in
Academic
Outcomes
Prior Variance in
Non-Academic
Outcomes
Prior Variance in
Adaptability
Prior Variance in
Buoyancy and
Self-Regulation
Prior Variance in
Personality
Prior Variance in
Implicit
Theories
© 2013 Martin et al
47
Table 1. Descriptive Statistics, Cronbach’s Alphas, and Mean CFA Loadings
Mean
SD
Skewness
Kurtosis
Cronbach’s α
Mean CFA Loading
T1
T2
T1
T2
T1
T2
T1
T2
T1
T2
T1
T2
PERSONALITY
Extraversion
4.91
4.96
1.09
1.09
-0.32
-0.34
-0.24
-0.27
.83
.84
.83
.85
Agreeableness
5.68
5.67
0.83
0.86
-0.94
-0.92
1.60
1.51
.80
.80
.82
.82
Neuroticism
3.71
3.76
1.01
1.06
-0.04
0.09
0.04
0.12
.75
.75
.82
.83
Openness
5.09
5.14
0.89
0.92
0.48
-0.41
0.83
0.19
.75
.75
.77
.78
Conscientiousness
4.87
4.92
1.18
1.18
-0.37
-0.37
-0.21
-0.19
.86
.86
.90
.89
IMPLICIT THEORIES
Ability
2.68
2.62
1.25
1.31
0.63
0.58
-0.11
-0.41
.79
.84
.67
.72
Effort
5.81
5.73
0.99
1.08
-0.99
-0.94
1.15
0.98
.84
.87
.72
.75
ADAPTABILITY AND COGNATES
Adaptability
4.98
5.09
.99
.99
-.16
-.30
.11
.11
.90
.92
.72
.75
Buoyancy
4.69
4.63
1.22
1.19
-.46
-.33
.08
-.01
.81
.77
.72
.68
Self-regulation
4.81
4.89
1.11
1.09
-.32
-.33
-.15
-.15
.86
.87
.67
.68
ACADEMIC OUTCOMES
Enjoy school
5.60
5.54
1.23
1.24
-1.06
-.95
.93
.71
.92
.91
.86
.85
Positive intention
6.00
6.06
0.95
0.99
-1.43
-1.56
2.61
2.85
.80
.83
.72
.74
Class participate
5.41
5.41
1.19
1.18
-.85
-.77
.65
.46
.91
.90
.85
.84
Self-handicapping
2.45
2.43
1.24
1.26
.82
.73
.09
-.21
.81
.84
.72
.76
Disengagement
2.09
2.13
1.15
1.18
1.31
1.21
1.39
1.02
.79
.82
.71
.73
NON-ACADEMIC OUTCOMES
Self-esteem
5.35
5.33
1.11
1.08
-.79
-.61
.73
.10
.80
.79
.73
.75
Life satisfaction
4.89
4.96
1.17
1.15
-.50
-.51
.02
.01
.80
.81
.68
.68
Meaning & purpose
4.90
5.05
1.37
1.36
-.61
-.69
.23
.35
.84
.87
.80
.83
Emotional instability
3.94
3.94
1.44
1.43
-.02
-.08
-.76
-.63
.84
.84
.72
.72
© 2013 Martin et al
48
Table 2. Correlations
Personality
Implicit
Theories
Adaptability
and Cognates
Academic
Outcomes
Non-Academic
Outcomes
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
PRIOR VARIANCE
(2010-2011)
77***
72***
69***
64***
71***
54***
50***
59***
62***
59***
61***
66***
66***
53***
59***
62***
65***
51***
58***
COVARIATES
Age
-05
-08
12*
-08**
-05
13**
-15***
-13**
-12
-12*
-19**
-05
-13**
12**
22***
-14**
-14*
-03
06
Language
-15***
01
05
-01
05
-04
01
01
02
05
-03
-03
-10**
04
-01
-01
-03
06**
01
Gender
-14**
-22***
-11***
-09***
-06
11
-01
01
13***
-12*
02
-10**
-05
12
-01
-07
-01
01
-16***
SES
09*
-03
03
06
-10*
-05
-08
-02
-10**
-03
01
04
09
-04
01
-03
01
-12
05
Prior achieve
-01
15*
-02
14***
11
-12*
-04
20***
01
07
10*
23***
18***
-24***
-14**
31***
10
-01
06*
PERSONALITY
1. Extraversion
23***
-27***
28***
05
-07
02
23***
17***
06*
21***
15**
42***
-07
-10
19***
21***
06*
-20***
2. Agreeableness
-
-21***
47***
50***
-37***
39***
54***
33***
39***
42***
52***
47***
-37***
-45***
48***
41***
36***
02
3. Neuroticism
-
-17***
-07
05
-01
-31***
-52***
-05
-25***
-14***
-19***
16***
24***
-27***
-35***
-07
65***
4. Openness
-
23***
-28***
17***
36***
20***
17***
22***
25***
27***
-26***
-23***
31***
18***
24***
-11**
5. Conscientiousness
-
-19***
26***
40***
26***
52***
31***
35***
35***
-34***
-43***
40***
37***
31***
04
IMPLICIT THEORIES
6. Ability
-
-59***
-23***
-06
-19***
-24***
-38***
-22***
35***
39***
-16**
-19***
-20***
01
7. Effort
-
41***
25***
37***
39***
48***
31***
-22***
-41***
33***
39***
40***
03
ADAPTABILITY AND COGNATES
8. Adaptability
-
59***
54***
53***
57***
55***
-38***
-50***
61***
58***
53***
-21***
9. Buoyancy
-
32***
48***
41***
39***
-22***
-39***
47***
47***
35***
-46***
10. Self-regulation
-
38***
48***
43***
-38***
-42***
38***
40***
41***
-01
ACADEMIC OUTCOMES
11. Enjoy school
-
69***
61***
-31***
-62***
48***
58***
34***
-09***
12. Positive intentions
-
62***
-40***
-67***
58***
50***
39***
-01
13. Class participation
-
-36***
-51***
50***
42***
36***
-04
14. Self-handicapping
-
59***
-41***
-35***
-21***
10**
15. Disengagement
-
-49***
-54***
-30***
13**
NON-ACADEMIC OUTCOMES
16. Self-esteem
-
66***
35***
-15**
17. Life satisfaction
-
40***
-17***
18. Meaning & purpose
-
04
19. Emotional instability
-
Notes: Decimals omitted; * p < .05 ** p < .01 *** p < .001
© 2013 Martin et al
49
Table 3. Standardized beta coefficients and explained variance for full process model
Personality
Implicit
Theories
Adaptability and
Cognates
Academic
Outcomes
Non-Academic
Outcomes
1
2
3
4
5
6
7
8
9
10
PART
ENJ
POS
SH
DIS
EST
SAT
MEAN
EMOT
AUTO-REGRESSION (2010-2011)
.77***
.66***
.66***
.65***
.69***
.48***
.44***
.34***
.41***
.39***
.47***
.49***
.45***
.37***
.42***
.36***
.46***
.35***
.23***
COVARIATES
Age
.01
.02
.06**
-.05*
-.02
.11*
-.11**
.01
.01
-.02
-.05*
-.05*
-.04
.09**
.11**
-.06*
-.08**
-.03
-.01
Language
-.05*
-.01
-.01
.02
-.05
-.04
.01
.03
.04
.02
.01
-.05**
.07**
.03
-.05
.01
.02
.05*
.04*
Gender
-.06
-.18***
-.07**
-.06
-.06
.09*
-.05
.02
.03
-.02
.01
.01
-.05
.06
.01
-.08
.01
.02
.03
SES
-.01
-.02
.01
.01
.01
.03
-.03
.05
.06
.01
-.01
-.01
.05*
.01
-.03
.01
.02
-.05
.04
Prior achieve
-.03
.10***
-.05
.05
.14***
-.11**
.06
.09*
.06*
.05
.04
.03
.09***
-.08**
-.03
.23***
.07*
-.06**
-.01
PERSONALITY
1. Extraversion
-.01
.03
.05**
.19***
.04
.03
-.02
-.03
.03
.03
-.01
-.04*
2. Agreeableness
.10**
-.07
.03
.04
.04
.03
.01
-.09
.01
-.02
.05
.17**
3. Neuroticism
-.19***
-.37***
-.01
.06
.06
.13**
.06
.01
.02
-.12**
.03
.49***
4. Openness
.05
-.01
.05
.05
.01
.04
.01
.02
.08*
.01
.08
-.03
5. Conscientiousness
.21***
.18***
.34***
.01
-.03
.01
-.04
.08**
.08*
.06
-.01
.01
IMPLICIT THEORIES
6. Ability
.10
.17**
.11***
-.01
-.07**
-.05
.32***
.19**
.12***
.12*
.06*
.16*
7. Effort
.29***
.23***
.19***
.06
.03
.17**
.15***
-.03
.22***
.22***
.13***
.05
ADAPTABILITY AND COGNATES
8. Adaptability
.18***
.15***
.20**
-.15**
-.15*
.29***
.26***
.38***
.05
9. Buoyancy
.05
.11**
.06
.07
.07
.11**
-.04
-.10*
-.29***
10. Self-regulation
.05
.11***
.01
-.15***
-.18***
-.07
.04
.06
.01
EXPLAINED VARIANCE
62%
45%
46%
52%
54%
30%
23%
51%
53%
49%
58%
51%
58%
46%
52%
68%
60%
45%
66%
* p < .05 ** p < .01 *** p < .001
Note 1. Explained variance in outcomes without adaptability: Self-esteem=64%; Life Satisfaction=56%; Meaning & Purpose=39%; Emotional Instability=65%; Class
Participation=57%; School Enjoyment=49%; Positive Intentions=56%; Self-handicapping=45%; Disengagement=51%
Note 2. Additional variance in outcomes explained by adaptability beyond auto-regression: Self-esteem=22%; Life Satisfaction=19%; Meaning & Purpose=16%; Emotional
Instability=4%; Class Participation=7%; School Enjoyment=13%; Positive Intentions=14%; Self-handicapping=8%; Disengagement=12%
© 2013 Martin et al
50
Neuroticism
Conscientiousness
Ability Beliefs
Effort Beliefs
Buoyancy
Adaptability
Self-Regulation
Class Participation
Enjoy School
Self-handicapping
Disengagement
Self-esteem
Life Satisfaction
Meaning and
Purpose
Emotional
Instability
-.37
.18
.23
.21
.29
.34
.11
.19
-.29
.18
.15
.11
-.15
-.18
.29
.38
.26
-.19
Figure 2. Hypothesized standardized beta paths significant at p < .001.
Notes: See Table 3 for direct paths between personality and outcomes, between implicit theories and outcomes, and for all other effects significant at p < .01 and p <
.05. Bold paths represent parameters leading to and from adaptability. Dashed paths represent parameters leading to and from cognate factors (self-regulation and
buoyancy). Dashed ellipses represent cognate factors to disentangle adaptability variance from buoyancy and self-regulation.
Dispositions and
Characteristic Orientations
Adaptability
(and Cognate Buoyancy and Self-
Regulation)
Academic and Non-
Academic Outcomes
© 2013 Martin et al
51
Appendix: The Adaptability Scale
(Martin, Nejad, Colmar, & Liem, 2012)
1. I am able to think through a number of possible options to assist me in a new situation
2. I am able to revise the way I think about a new situation to help me through it
3. I am able to adjust my thinking or expectations to assist me in a new situation if necessary
4. I am able to seek out new information, helpful people, or useful resources to effectively deal with
new situations
5. In uncertain situations, I am able to develop new ways of going about things (eg. a different way
of asking questions or finding information) to help me through
6. To assist me in a new situation, I am able to change the way I do things if necessary
7. I am able to reduce negative emotions (eg. fear) to help me deal with uncertain situations
8. When uncertainty arises, I am able to minimize frustration or irritation so I can deal with it best
9. To help me through new situations, I am able to draw on positive feelings and emotions (eg.
enjoyment, satisfaction)
Notes. Based on prior factor analyses (Martin et al., 2012), Items 1-6 load on a first order
‘cognitive-behavioral’ factor and Items 7-9 load on a first order ‘affective’ factor. Importantly,
however, the two first order factors are highly correlated and thus a global factor (i.e., single 9-item
factor) or higher order factor (i.e., 9 items represented by 2 first order factors, that are the indicators
of a higher order factor) can also represent Adaptability in statistical analyses – particularly when
using Adaptability as a predictor, to avoid issues related to collinearity.
A preview of this full-text is provided by American Psychological Association.
Content available from Journal of Educational Psychology
This content is subject to copyright. Terms and conditions apply.
