Special Section: Social in the Emotional and the Emotional in the Social

International Journal of
Behavioral Development
Predictors of changes in students’ 2017, Vol. 41(1) 15–29
ª The Author(s) 2016
Reprints and permissions:
motivational resilience over the sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0165025416642051
school year: The roles of teacher support, journals.sagepub.com/home/ijbd

self-appraisals, and emotional reactivity

Jennifer Pitzer1,2 and Ellen Skinner1

Abstract
Students perform better in school to the extent they are able to engage fully, cope adaptively, and bounce back from obstacles and setbacks
in their academic work. These three processes, which studies suggest are positively inter-connected, may comprise a self-sustaining system
that enables motivational resilience. Using self-determination theory to frame hypotheses about such a motivational system, this study
examined (1) whether a set of personal factors (self-perceptions of relatedness, competence, and autonomy), interpersonal resources
(perceptions of teacher warmth, structure, and autonomy support), and emotional reactivity predicted changes in motivational resilience
over the school year; (2) whether motivational resilience in turn predicted improvements in students’ achievement and also fed back to
increases in their personal and interpersonal resources; and (3) whether teacher support could shift established motivational patterns. A
latent path model depicting these processes showed a good fit with self-report data from 1020 students in Grades 3 through 6 collected in
fall and spring of the same school year (including achievement data from a random subset, n ¼ 365). Multiple regressions predicting changes
from fall to spring refined the proposed model. Moreover, teacher support was crucial: Students who began the year with at-risk profiles,
but also experienced high teacher support, ended the year on par with low-risk students; whereas students who began with resilient
profiles but experienced low levels of teacher support ended the year at-risk. Discussion focused on identifying levers for intervention and
the essential role teacher support plays in these dynamics.

Keywords
achievement, coping, emotional reactivity, engagement, motivational resilience, persistence, reciprocal effects, self-determination theory,
teacher support

Intuitively, it seems that students should learn more when they strategies students utilize when they encounter challenges and set-
engage actively in academic activities, deal constructively with backs in their schoolwork. This research points to advantages for
challenges they encounter while working on these tasks, and persist students who use more productive ways of coping (especially
in the face of obstacles and setbacks. Empirical evidence corrobo- problem-solving and help-seeking, but also reappraisal and emotion
rates this intuition: Separate lines of study show that students’ regulation strategies like self-encouragement, comfort-seeking, and
engagement (Fredricks, Blumenfeld, & Paris, 2004), their academic commitment) and to disadvantages for those who rely on more
coping (Hess & Copeland, 2001), and their persistence (Martin & maladaptive strategies (such as avoidance, escape, concealment,
Marsh, 2009) each predict academic success (i.e., learning, class- self-pity, rumination, or blaming others). Together, these adaptive
room grades, and achievement), starting in the elementary grades and maladaptive ways of coping, which can be considered a profile
and continuing through middle and high school. The most robust or repertoire of strategies that children draw upon when they
body of research examines the effects of classroom engagement, encounter academic stressors (Boekaerts, 1993), have been shown
that is, students’ ongoing, active, enthusiastic participation in aca- to predict better school performance during late elementary (Cau-
demic activities, as an important driver of learning, retention, and sey & Dubow, 1992; Lau & Nie, 2008; Tero & Connell, 1984),
academic performance, as well as a protective factor against neg- middle (Leung, & He, 2010; MacCann, Fogarty, Zeidner, &
ative outcomes such as gang involvement and school dropout Roberts, 2011), and junior high school (Swanson, Valiente,
(Christenson, Reschly, & Wylie, 2012; Fredricks et al., 2004; Lemery-Chalfant, & O’Brien, 2011). At least in part, students’
Upadyaya & Salmela-Aro, 2013). This research also shows that
students’ disengagement or disaffection from school (including
lack of effort, passivity, boredom, apathy, and frustration) can 1
Portland State University, Portland, OR, USA
exacerbate these risky behaviors and contribute to underachieve- 2
Institute for Research and Reform in Education, NJ, USA
ment (Blondal & Adalbjarnardottir, 2012; Henry, Knight, & Thorn-
berry, 2012; Li & Lerner, 2011; Morrison, Robertson, Laurie, & Corresponding author:
Kelly, 2002). Jennifer Pitzer, Portland State University, PO Box 751, Portland, OR
Although not as definitive as research on engagement, studies of 97207-0751, USA.
academic coping also link markers of school success to the Email: jpitzer@pdx.edu
16 International Journal of Behavioral Development 41(1)

Figure 1. Model of internal and external dynamics of motivational resilience. Internal dynamics describe a self-reinforcing cycle wherein students’ ongoing
engagement vs. disaffection in school influences their coping strategies and actions following challenges and setbacks. External dynamics demonstrate how
students’ personal and interpersonal resources and emotional reactivity can either support or hinder their motivational resilience, which in turn influences
their academic outcomes.

adaptive coping seems to benefit their academic performance Skinner et al., 2013a, 2013b). Conceptualizations and studies that
through its effects on persistence or re-engagement in the face of explicitly bring these constructs together may serve to fill gaps in
challenges (Lau & Nie, 2008; Lemos, 2002; Skinner, Pitzer, & both the research on engagement (which can be expanded by incor-
Steele, 2013a). porating coping’s accounts of what happens when engaged and
disaffected students encounter academic challenges and problems)
and the research on coping (which can be expanded by incorporat-
A systems perspective on motivational processes ing re-engagement and persistence as proximal outcomes of coping
Although motivational researchers have long posited connections and as pathways to subsequent learning).
between students’ engagement and their coping and persistence Models of motivational resilience (and its opposite, motiva-
(Boekaerts, 2002; Dweck, 2006; Lemos, 2002; Martin & Marsh, tional vulnerability) view processes of engagement, coping, and
2009; Thompson & Gaudreau, 2008), empirical evidence linking persistence as complementary parts of a dynamic motivational sys-
these processes has only begun to accumulate in recent years. For tem that work together to promote or undermine students’ learning,
example, elementary and middle-school students higher in engage- academic success, and retention. From this perspective (see
ment (or the kinds of positive emotions that characterize engage- Figure 1), students’ ongoing engagement in school can serve as
ment) also report utilizing a range of more adaptive strategies, an energetic resource, such that students who are more highly
including more positive coping (Kaplan & Midgely, 1999), more engaged in school are better able to utilize adaptive coping strate-
self-reliance/problem-solving and social support-seeking gies and ultimately to re-engage with challenging tasks. In contrast,
(Reschley, Huebner, Appleton, & Antaramian, 2008), and more students who are disaffected are less equipped to deal construc-
strategizing, help-seeking, comfort-seeking, self-encouragement, tively with such challenges, leading them to use more maladaptive
and commitment (Skinner, Pitzer, & Steele, 2013b). In contrast, coping strategies, and culminating in an increased likelihood of
students who show higher levels of behavioral or emotional disaf- giving up rather than persevering.
fection also report greater reliance on a range of unproductive ways Situated within the work on ‘‘everyday resilience’’ and ‘‘every-
of coping, including more denial, projection, and anxiety amplifi- day coping’’ (Martin, 2013; Martin & Marsh, 2009; Skinner &
cation coping (Kaplan & Midgely, 1999), more avoidance coping Pitzer, 2012; Wolchik & Sandler, 1997), notions of everyday moti-
(Lau & Nie, 2008), and more escape, confusion, concealment, self- vational resilience and vulnerability refer holistically to the idea of
pity, and blaming others (Skinner et al., 2013b). an integrated engagement-coping-persistence system, and are used
Recent theoretical efforts have begun to consider whether, to highlight the dynamic interactions among students’ ongoing
rather than studying engagement, coping, and persistence as sepa- engagement, coping, and re-engagement in the face of difficulties
rate motivational processes (as has largely been the case up to now), and setbacks in school. Such models are useful to the extent that
it might be useful to examine them as inter-connected components they can inspire researchers to examine these processes simultane-
of the same developing motivational system (Leung & He, 2010; ously, to explore whether they seem to function as complementary
Martin, 2013; Martin & Marsh, 2009; Skinner & Pitzer, 2012; components of the same system. For example, studies of the
Pitzer and Skinner 17

internal dynamics of motivational resilience suggest that the posi- show that students’ positive self-appraisals predict multiple favor-
tive feed-forward and feedback effects among these components able motivational, coping, and academic outcomes (Green et al.,
may create a self-perpetuating feedback loop over time that can 2012; Niemiec & Ryan, 2009; Raufelder et al., 2014; Roeser, Midg-
sustain resilience for students who are initially high in engagement, ley, & Urdan, 1996; Ryan & Deci, 2000, 2009; Skinner et al., 1998;
but can also amplify motivational vulnerability for students initially Thompson, & Gaudreau, 2008).
high in disaffection (Kaplan & Midgley, 1999; Lau & Nie, 2008;
Reschly et al., 2008; Skinner, Furrer, Marchand, & Kindermann, Perceptions of teacher support. SDT focuses on three facets of
2008; Skinner et al., 2013a). teacher–student relationships that can directly support or under-
mine students’ needs for relatedness, competence, and autonomy.
Factors that promote motivational resilience. Building on this Supportive classroom interactions include warmth (i.e., caring
work, the purpose of the current study was to explore whether interpersonal interactions, as opposed to rejecting relationships;
outside forces can re-shape students’ otherwise self-sustaining Wentzel, 2009), structure (i.e., predictable and consistent class-
motivational systems. In systems terminology, this study tests a room contexts, in contrast to chaotic environments; Skinner
model of the external dynamics of students’ everyday motivational et al., 1998), and autonomy support (i.e., providing choice and
resilience (see Figure 1). Grounded in self-determination theory explanations of relevance, versus coercive interactions; Reeve,
(SDT; Connell & Wellborn, 1991; Deci & Ryan, 1985), this inves- 2009). These contextual supports are strong predictors of students’
tigation utilized data from students in Grades 3 through 6 collected personal resources as well as their motivational states and academic
in fall and spring of the same school year (as part of a larger project success (e.g., Hughes & Kwok, 2007; Klem & Connell, 2004; Nie-
on motivational development during late elementary and early mid- miec & Ryan, 2009; Reeve, Jang, Carrell, Jeon, & Barch, 2004;
dle school; Skinner, Zimmer-Gembeck, & Connell, 1998) to exam- Reyes, Brackett, Rivers, White, & Salovey, 2012; Skinner et al.,
ine: (1) whether students’ self-appraisals and experiences of 2008). Because interpersonal resources seem to exert such a perva-
teacher support could predict increases in their motivational resi- sive positive effect throughout students’ motivational systems, we
lience (captured as the aggregate of their engagement, coping, and were interested in examining whether teacher support could pro-
persistence) over the school year; and whether emotional reactivity mote motivational resilience both directly, and by bolstering stu-
is one pathway through which both personal and interpersonal dents’ self-system appraisals (i.e., relatedness, competence, and
resources contribute to such increases; (2) whether students’ moti- autonomy).
vational resilience in turn, not only predicts improvements in their
achievement, but also feeds back to enhance their perceptions of Emotional reactivity. An important part of this first goal was to
personal and interpersonal resources; and (3) whether teacher sup- explore the role of emotional reactivity in students’ motivational
port can, over time, shift already established motivational patterns, systems. In keeping with studies suggesting that positive emotions
exploring (a) whether, with high levels of teacher support, at-risk predict adaptive coping and negative emotions maladaptive coping
students caught in ongoing vicious cycles could begin to participate in school (Kaplan & Midgley, 1999; Spangler, Pekrun, Kramer, &
in more virtuous feedback loops, and (b) whether low levels of Hofmann, 2002), emotional reactivity was examined as a predictor
teacher support could contribute to the emergence of motivational of increases in students’ motivational vulnerability. Moreover, it
vulnerabilities, even for students who began the year with few seemed possible that one way in which personal and interpersonal
indicators of risk. Previous research provides evidence supporting resources might promote motivational resilience would be through
each of these proposed links, but to date, no published studies have their effects on dampening reactivity. For example, anxiety ampli-
considered this motivational system as a whole. fication, and other forms of emotional reactivity, have been shown
to be lower in students with appraisals of high competence or con-
trollability (Dweck, 2006; Tero & Connell, 1984), a strong sense of
Supporting motivational resilience autonomy (Lemos, 2002), and solid relatedness to teachers (Lynch
The first goal of the study was to determine whether students’ & Cicchetti, 1997). Hence, it seemed that one way such personal
personal and interpersonal resources can bolster their motivational and interpersonal resources might contribute to greater resilience
resilience over time. According to SDT, students who appraise would be by preventing children from becoming overly distressed
themselves as belonging in the classroom, as competent and auton- in the face of academic obstacles and difficulties. Conversely, lack
omous in their work at school, and who perceive their interactions of these same personal and interpersonal resources could contribute
with teachers to be supportive and well-calibrated to their current to increases in motivational vulnerability by triggering students’
needs will show more resilient responses in the face of setbacks and distress and emotional reactivity.
challenges, whereas students who lack this confidence and support
will be at greater risk for developing motivational liabilities over Motivational resilience and academic achievement
time (Reeve, 2012).
The second goal of the study was to investigate the hypothesized
Self-appraisals. A central assertion of SDT is that individuals’ well- reciprocal link between students’ motivational resilience and their
being is optimized when their contexts support (rather than thwart) academic achievement. Because engagement is well-established as
their innate psychological needs for relatedness, competence, and a significant contributor to students’ academic success (Christenson
autonomy (Connell & Wellborn, 1991; Deci & Ryan, 1985). That et al., 2012; Fredricks et al., 2004; Jimerson, Campos, & Greif,
is, students will best thrive when they feel they belong and are cared 2003; Upadyaya & Salmela-Aro, 2013; Wigfield et al., 2015), it
for, are effective in their interactions with their environments, and follows that motivational resilience, which brings students back to
view their actions as originating from their own authentic core self this essential engaged state, could promote students’ achievement.
and desires. Consistent with this notion, long histories of research in Therefore, we expected that each component of motivational resi-
the areas of perceived control, self-determination, and attachment lience would be positively and significantly correlated with
18 International Journal of Behavioral Development 41(1)

academic performance at each time point. In addition, we expected higher teacher support and more positive self-appraisals predict
motivational resilience to predict increases in students’ academic improvements in students’ motivational resilience over the school
achievement across the school year, such that students who began year, as well as (a) whether self-appraisals are an important path-
the year highly motivationally resilient would show improvements way through which interpersonal resources exert their effects, and
in their grades, whereas achievement would decline for those who (b) whether emotional reactivity is one way in which both kinds
were more motivationally vulnerable. Moreover, we also expected of resources shape motivational resilience; (2) whether motiva-
to see feedback effects of achievement on changes in motivational tional resilience not only predicts gains in students’ academic
resilience, based on the notion that when a student learns more, this performance across the school year, but also enhances students’
will add fuel to their ongoing engagement and persistence, whereas personal and interpersonal resources; and whether achievement
continued academic struggles could add discouragement and frus- in turn predicts gains in motivational resilience over time; and
tration to disaffection, increasing motivational vulnerability. (3) whether teacher support is a particularly important factor—
one that can help students break out of vicious motivational
cycles or, when lacking, one that can put students at risk for the
Feedback from motivational resilience to teacher development of motivational vulnerabilities. The primary contri-
support and self-system processes butions of the present study are in its short-term longitudinal
design, which allows analyses to examine whether each proposed
As part of the proposed links between external supports and moti-
antecedent measured in fall can predict changes in its hypothe-
vational resilience, this study also examined whether students’
sized consequences over the same school year; and in its
motivational resilience exerted reciprocal effects on teachers’ pro-
framework, which allows all of these potential feedforward and
vision of support across the school year and on their own self-
feedback effects to be considered simultaneously as parts of
system processes. Student–teacher relationships may exist in a
the same complex and dynamic motivational system during late
dynamic feedback loop, such that students who are highly engaged
elementary and early middle school.
tend to elicit teacher more support which in turn leads to further
increases in motivation, whereas disaffected students attract more
unsupportive teacher behaviors over time and consequently exacer- Method
bate their own motivational vulnerabilities (Connell, Spencer, &
Aber, 1994; Jang, Kim, & Reeve, 2012; Nurmi & Kiuru, 2015; Participants
Skinner & Belmont, 1993; Van Ryzin, 2011). Moreover, higher Data for this study were drawn from an existing longitudinal dataset
levels of motivational resilience were also predicted to enhance that was part of a large, district-wide evaluation of a rural-suburban
students’ feelings of competence, autonomy, and relatedness over school district in upstate New York in which 1608 elementary and
time (Van Ryzin, Gravely, & Roseth, 2009). middle-school students (Grades 3 through 7) completed surveys
about their engagement and coping in school. Fifty-three of their
At-risk motivational systems and teacher support teachers also participated by completing questionnaires about their
observations of and interactions with students. Data were collected
Finally, a special focus of the study was the empirical examination using a cohort-sequential design, with data collected in fall
of the extent to which teachers can shape the motivational dynamics (October) and spring (May) for four consecutive years. Achieve-
of the classroom by interrupting existing detrimental feedback ment scores were also obtained from school records for a random
loops and reestablishing positive motivational pathways. We subset of the participants. For a complete description of the larger
expected that some students would begin the year motivationally study, see Skinner et al. (1998).
at-risk, as marked by a pattern of appraisals in which they interpret For this study, information from a subset of students from the
the stressful events they encounter at school as having devastating third year of the project (n ¼ 1020) was used, because these two
implications for their ability to meet their needs for relatedness, assessments included measures of academic coping. Participants
competence, and autonomy. These ‘‘catastrophizing’’ appraisals, were a sample of students in Grades 3 through 6, including
because they magnify the negative consequences of stressful 138 third-grade students (66 boys and 72 girls), 342 fourth-grade
events, can intensify students’ emotional reactions and increase students (172 boys and 170 girls), 170 fifth-grade students (78 boys
their reliance on maladaptive coping strategies (Brown, O’Keefe, and 92 girls), and 368 sixth-grade students (192 boys and 176 girls);
Sanders, & Baker, 1986; Friedel, Cortina, Turner, & Midgley, two students were missing grade and/or gender data. The majority
2007; Kaplan & Midgley, 1999; Mantzicopoulos, 1997; Skinner of students were Caucasian, with less than 5% identifying as non-
et al., 2013b; Tero & Connell, 1984). Nevertheless, we hypothe- white, and their families’ socioeconomic status (determined by
sized that even for motivationally at-risk students, high teacher parent occupation and education level) were primarily working to
support would be able to lift them off their expected downward middle class.
motivational trajectories. In contrast, we expected that students who
were not at-risk at the beginning of the school year (i.e., who
reported low levels of catastrophizing) could still lose their motiva- Procedures
tional advantage if they experienced low levels of teacher support Pairs of trained interviewers administered questionnaires to stu-
across the year. dents during three 40-minute class sessions. In each session, one
interviewer read the questions aloud to students as they marked
their answers on the questionnaire, while the second circulated
Summary of hypotheses
around the classroom to answer students’ questions. The students’
In order to investigate the external dynamics of students’ motivational teachers were not present in the classroom during the data collec-
resilience across the school year, this study examined: (1) whether tion; most used the time to complete their own questionnaires.
Pitzer and Skinner 19

Table 1. Summary of descriptive statistics for each construct in fall and spring.

Fall Spring

Construct No. of Items a Mean SD a Mean SD

Interpersonal resources: Teacher support 66 .95 2.98 .43 .96 2.93 .46
Warmth 16 .86 3.01 .50 .88 2.92 .53
Structure 29 .87 3.04 .42 .89 3.01 .44
Autonomy support 21 .86 2.90 .47 .88 2.87 .49
Personal resources: Self-system processes 44 .91 3.09 .41 .92 3.05 .41
Relatedness 8 .82 3.15 .59 .83 3.14 .58
Competence 23 .88 3.35 .44 .90 3.32 .44
Autonomy 13 .80 2.76 .53 .83 2.68 .55
Emotional reactivity 11 .86 2.45 .63 .87 2.35 .61
Motivational resilience 89 .82 3.19 .38 .84 3.13 .40
Engagement vs. Disaffection 25 .88 3.17 .44 .90 3.14 .45
Coping profile 55 .86 2.98 .35 .87 2.94 .38
Re-engagement vs. Giving up 9 .81 3.43 .48 .82 3.33 .49
Catastrophizing appraisals 27 .94 2.02 .59 .94 2.00 .58
Cat of relatedness 9 .88 1.88 .65 .89 1.86 .64
Cat of competence 9 .84 2.11 .65 .86 2.09 .64
Cat of autonomy 9 .79 2.07 .59 .81 2.05 .58

Note. N ¼ 1020 students in Grades 3 through 6. All scales ranged from 1 (not at all true for me) to 4 (very true for me).

Measures assessed using 21 items tapping the extent to which teachers pro-
vided students with choices, exerted control over them, offered
Students completed sets of items tapping their experiences of inter- respect for their ideas and opinions, and explained the relevance
personal resources, personal resources, emotional reactivity, moti- of learning activities (e.g., ‘‘My teacher gives me a lot of choices
vational resilience, and, as a measure of existing motivational risk, about how I do my schoolwork’’).
catastrophizing appraisals. Students rated all items using a 4-point
Likert scale to indicate whether each item was (1) Not at all true for Personal resources: Self-system processes. Students also responded
me, (2) Not very true for me, (3) Sort of true for me, or (4) Very true to measures of their perceptions of relatedness, competence, and
for me. All negatively worded items were reverse coded, and items autonomy. Students’ sense of relatedness was measured using eight
were averaged within constructs to create composite scale scores. items that described their feelings of connectedness and belonging
These scale scores could range from 1 to 4, with higher numbers to their teachers and classmates (Furrer & Skinner, 2003) via four
indicating more of the respective construct. items for each social partner (e.g., ‘‘When I am with my teacher,
All the measures used in this study capture multi-dimensional I feel accepted’’). Perceived competence was measured using
constructs, as demonstrated by confirmatory structural analyses of 23 items from the Student Perceptions of Control Questionnaire
most of the scales used (e.g., Skinner, Kindermann, & Furrer, (Skinner, Wellborn, & Connell, 1990; e.g., ‘‘If I decide to learn
2009). At the same time, the dimensions of all the measures have something hard, I can’’). Perceptions of autonomy were measured
also been found to be highly inter-correlated enough to allow them using 13 items depicting reasons for participating in academic
to be usefully combined into aggregate overall scale scores, that are activities (Ryan & Connell, 1989), that varied on a continuum of
both multidimensional and internally consistent (see Table 1). In self-regulation from external (e.g., ‘‘Because the teacher says we
general, latent variables corresponding to the higher-order target have to’’) to identified (e.g., ‘‘Because I want to learn new things’’),
constructs were used in structural analyses. However, in order to to intrinsic (e.g., ‘‘Because it’s fun’’). Summary scores averaged the
examine whether individual dimensions were differentially three autonomy subscales, with external reverse coded.
involved in each process link, inter-correlations and multiple
regressions involving both aggregates and subscales were also
Emotional reactivity. Students reported on 11 items measuring the
conducted.
extent to which they experience negative emotional responses when
they encounter obstacles and setbacks in school (Skinner et al.,
Interpersonal resources: Teacher support. Students completed 2013b; e.g., ‘‘I get really upset when something bad happens in
measures tapping their experiences of support from their classroom school’’).
teachers along three dimensions (Skinner & Belmont, 1993):
(1) warmth versus rejection, measured via 16 items tapping whether Motivational resilience. Students responded to measures of motiva-
teachers spent time with students, showed them affection, and were tional resilience, including their engagement in school, academic
available, knowledgeable, and dependable (e.g., ‘‘My teacher is coping, and re-engagement. The three components were combined
always there for me’’); (2) structure versus chaos, captured by to form a summary score, with negative items reverse coded.
29 items tapping whether teachers offered clear expectations, con-
tingent responses, help and support, and attuned teaching strategies Engagement versus disaffection. Students responded to 25 items
(e.g., ‘‘Every time I do something wrong, my teacher acts differ- tapping their ongoing engagement versus disaffection in the class-
ently,’’ reverse coded); and (3) autonomy support versus coercion, room (Skinner et al., 2009): (1) five items measured behavioral
20 International Journal of Behavioral Development 41(1)

engagement (e.g., ‘‘I work hard when we start something new in Results
class’’); (2) five tapped behavioral disaffection (e.g., ‘‘When I’m in
class, I just act like I’m working’’); (3) six measured emotional Missing data
engagement (e.g., ‘‘When we start something new in school, I feel As is inevitable in longitudinal studies, a strategy was required to
interested’’); and (4) nine items tapped emotional disaffection, address missing data. Data were assumed to be Missing at Random
including boredom, frustration, or anxiety (e.g., ‘‘When I’m doing (MAR; Schafer & Graham, 2002), and were handled using the Mul-
my work in class, I feel worried’’). tiple Imputation function in SPSS version 23 (IBM Corp., 2015), with
results pooled across five iterations. The percentage of missing values
Academic coping. Students responded to 55 items tapping their for each variable ranged from nearly 0% for the demographic vari-
academic coping in school (Skinner et al., 2013b). Items were ables to 25% for a few items within coping subscales (Comfort-
divided into 11 subscales consisting of five items each. Each sub- seeking and Projection). Achievement data were not imputed for stu-
scale prompted students to describe their responses to stressful dents who were not part of the randomly selected subset; we used Full
events in school, utilizing one of four different item stems (e.g., Information Maximum Likelihood (FIML) estimation to deal with the
‘‘When I have difficulty learning something . . . ’’). Five of the sub- missingness for the structural equation model (Graham, 2009).
scales measured students’ adaptive ways of coping, including
(1) Strategizing (e.g., ‘‘I try to figure out what I did wrong so that
it won’t happen again’’); (2) Help-seeking (e.g., ‘‘I ask the teacher Descriptive information
to explain what I didn’t understand’’); (3) Comfort-seeking (e.g., ‘‘I
Means, standard deviations, and internal consistencies for each
discuss it with someone who will help me feel better about it’’);
variable at each time point are presented in Table 1. In general,
(4) Self-encouragement (e.g., ‘‘I tell myself I’ll do better next
students reported high levels of personal and interpersonal
time’’); and (5) Commitment (e.g., ‘‘I remind myself that it’s some-
resources, moderate levels of emotional reactivity, and relatively
thing that I really want to do’’). The six maladaptive ways of coping
low levels of catastrophizing. They were actively engaged in class-
included (1) Confusion (e.g., ‘‘It’s difficult for me to think’’);
room activities, and, when faced with challenges, they tended to
(2) Escape or avoidance (e.g., ‘‘I say I didn’t care about it’’);
utilize more adaptive than maladaptive coping strategies, and to
(3) Concealment (e.g., ‘‘I don’t tell anyone about it’’); (4) Self-pity
persist after encounters with obstacles or setbacks. In school, stu-
(e.g., ‘‘I ask myself, ‘Why is this always happening to me?’’’);
dents earned above-average marks, typically between a B and a B.
(5) Rumination (e.g., ‘‘I can’t get it out of my head’’); and (6) Pro-
Correlations among all variables at both time points and their
jection, or blaming others (e.g., ‘‘I say it was the teacher’s fault’’).
cross-time stabilities are presented in Table 2. As expected, engage-
Profile scores were computed which averaged the sets of adaptive
ment, coping, and re-engagement were positively and significantly
and maladaptive coping scores, with the maladaptive scores reverse
related to one another, while emotional reactivity was negatively cor-
coded, indicating the balance of overall coping that was adaptive
related with these constructs at both time points. Likewise, students’
versus maladaptive (Skinner et al., 2013b).
interpersonal resources, personal resources, and catastrophizing
appraisals were all significantly related in both fall and spring as
Re-engagement vs. giving up. Students reported on nine items tap-
expected. For all constructs, as expected, cross-time stabilities were
ping their reactions to encounters with challenges in school (Skin-
high, making it difficult to predict change over time due to the limited
ner et al., 2013b). Four items tapped persistence or re-engagement
variance remaining after controlling for students’ scores in fall.
(e.g., ‘‘If a problem is really hard, I keep working at it’’), and five
items tapped giving up (e.g., ‘‘If I don’t understand something right
away, I stop trying’’). Items were averaged to form a summary Structural model of the external dynamics
score, with giving up items reverse-coded.
of motivational resilience
Outcomes: Achievement. For a random subset of students Latent structural modeling was used to examine the predicted links
(n ¼ 365), achievement data were available, including students’ among students’ personal and interpersonal resources, emotional
report card grades for reading, language arts, spelling, and math. reactivity, motivational resilience, and academic achievement
Scores were converted from letter grades to numbers ranging from simultaneously. For each latent variable, items were distributed
1 (F or U) to 12 (A or V), and composite scores were calculated by among three parcels (Little, Cunningham, Shahar, & Widaman,
averaging students’ grades across these subjects. 2002). Students’ personal and interpersonal resources (i.e., teacher
support and self-system processes, respectively) were hypothesized
Motivational risk: Catastrophizing appraisals. Students also to predict both their emotional reactivity and their motivational
reported on three kinds of catastrophizing appraisals (Skinner resilience, which in turn predicted their academic achievement.
et al., 2013b). Nine items tapped catastrophizing of relatedness, We compared two alternative models: one in which interpersonal
in which appraisals of stressful events magnified their negative and personal resources made independent contributions, versus one
implications for interpersonal relationships (e.g., ‘‘When something in which interpersonal resources also exerted an effect through
bad happens to me in school [like not doing well on a test or not personal resources. Although both models showed a good fit with
being able to answer an important question in class], I feel like I let the data, the second model was a significantly better fit than the
everybody down’’). Nine items targeted catastrophizing of compe- first. The final structural model, presented in Figure 2, also included
tence, in which appraisals focused on negative events as demon- cross-time stabilities for all constructs (e.g., interpersonal resources
strating low ability and forecasting future problems (e.g., ‘‘I worry in fall to interpersonal resources in spring), but for the sake of
that I won’t do well on anything’’). Finally, nine items measured clarity, they are not shown; see Table 2 for the zero-order correla-
catastrophizing of autonomy, in which appraisals emphasized guilt, tions. Fit to the data was good, CFI ¼ .988; RMSEA ¼ .034 (Hu &
self-blame, or loss of self-worth (e.g., ‘‘I feel like it’s all my fault’’). Bentler, 1999). As can be seen, three of the predicted links
Pitzer and Skinner 21

Table 2. Intercorrelations among components of motivational resilience, among personal resources, among interpersonal resources, and among
catastrophizing appraisals in fall and spring.

Interpersonal Personal Emotional Motivational Achievement Catastrophizing

resources resources reactivity resilience (n ¼ 365) appraisals

Fall Spring Fall Spring Fall Spring Fall Spring Fall Spring Fall Spring

Interpersonal resources .64 .75 .76 .25 .24 .73 .72 .17** .22 .49 .49
Warmth – – .71 .71 .22 .21 .67 .64 .12* .20 .44 .43
Structure – – .71 .72 .21 .22 .73 .73 .19 .22 .47 .48
Autonomy support – – .68 .72 .27 .26 .66 .68 .18** .22 .47 .48
Personal resources .75 .76 .66 .34 .29 .82 .84 .19 .28 .58 .54
Relatedness .65 .67 – – .26 .22 .56 .59 .14** .20 .44 .41
Competence .61 .62 – – .33 .30 .80 .83 .27 .36 .65 .64
Autonomy .51 .51 – – .23 .18 .61 .60 .06ns .14** .33 .27
Emotional reactivity .25 .24 .34 .29 .56 .37 .35 .09^ .02ns .67 .70
Motivational resilience .73 .72 .82 .84 .37 .35 .71 .19 .29 .67 .66
Engagement .69 .70 .80 .83 .42 .40 – – .17** .25 .65 .64
Coping profile .70 .69 .76 .75 .39 .37 – – .16** .22 .68 .66
Re-engagement .61 .60 .67 .71 .21 .21 – – .17** .31 .51 .52
Achievement .17** .22 .19 .28 .09^ .02ns .19 .29 .75 .17** .17**
Catastrophizing appraisals .49 .49 .58 .54 .67 .70 .67 .66 .17** .17** .67
Cat of relatedness .47 .47 .55 .51 .56 .57 .63 .63 .16** .17** – –
Cat of competence .44 .44 .53 .49 .69 .72 .60 .60 .19 .16** – –
Cat of autonomy .47 .47 .56 .52 .64 .67 .65 .64 .14** .13* – –

Note. N ¼ 1020. Cross-time stabilities are reported in bold. All correlations are significant at p < .001 except as noted.
**p < .01; *p < .05; ^p < .10; ns¼ not significant.

Figure 2. Time-ordered latent structural model depicting the external dynamics of motivational resilience, in which students’ personal and interpersonal
resources predict their emotional reactivity and motivational resilience, which in turn predicts achievement. N ¼ 1020 students in Grades 3 through 6. All
coefficients significant at p < .001 except as noted, **p < .01. *p < .05. Dashed lines were not significant. 2(264) ¼ 575.168, CFI ¼ .988, RMSEA ¼ .034, 90%
CI (.030, .038). Model included indicators for each latent construct and cross-time stabilities, but they are not shown for the sake of clarity.

involving emotional reactivity were not significant—from interper- markers of proposed antecedents in fall to predict changes in poten-
sonal resources to emotional reactivity, in both fall and spring, and tial consequences from fall to spring. These analyses were designed
from grades in fall to emotional reactivity in spring. Additionally, to allow us to refine the proposed model by examining each link in
two of the predicted feedback loops involving interpersonal the system using both aggregated and individual dimensions of
resources were not significant: from motivational resilience and each component. As a whole, these analyses could either show that
grades in fall to interpersonal resources in spring. These refine- all the components of the aggregate constructs showed the same
ments suggested that students’ self-appraisals were more central pattern of effects, or they could suggest that specific components of
to their emotional reactivity and motivational resilience than were the aggregate constructs were more likely to be the ‘‘active ingre-
their perceptions of teacher support. dients’’ with respect to particular consequences.
The first set of analyses focused on the proposed antecedents of
motivational systems. Correlations between students’ motivational
Predictors of motivational resilience resilience and each of the proposed predictors from the model (see
The foregoing latent structural analyses of all the main components Table 2) showed that, as expected, in both fall and spring, motiva-
of the proposed motivational system were supplemented by a series tional resilience was positively and significantly related to each
of more differentiated multiple regression analyses which used interpersonal and personal resource (average r ¼ .73 and .83,
22 International Journal of Behavioral Development 41(1)

Table 3. Multiple regressions in which students’ personal resources, interpersonal resources, and emotional reactivity in fall predict changes in emotional
reactivity and motivational resilience from fall to spring.

Changes from fall to spring in

Engagement vs. Re-engagement vs.

Emotional reactivity Motivational resilience Disaffection Coping profile Giving up

b (SE) b (SE) b (SE) b (SE) b (SE)

Predictor in Fall 95% CI ß 95% CI ß 95% CI ß 95% CI ß 95% CI ß

Personal resources .10 (.04) .07* .17 (.04) .18 .29 (.04) .26 .16 (.03) .18 .24 (.04) .20
[.18, .02] [.10, .25] [.21, .37] [.10, .23] [.16, .32]
Relatedness .08 (.03) .07** .03 (.02) .05^ .08 (.02) .10 .04 (.02) .06* .06 (.02) .07*
[.13, .02] [.01, .07] [.04, .12] [.00, .07] [.01, .10]
Competence .08 (.06) .06* .07 (.03) .08* .13 (.03) .13 .08 (.03) .09** .23 (.04) .20
[.16, .01] [.01, .14] [.06, .20] [.02, .14] [.15, .30]
Autonomy .02 (.03) .02ns .09 (.02) .12 .13 (.02) .15 .10 (.02) .14 .13 (.03) .15
[.08, .04] [.05, .13] [.08, .17] [.06, .13] [.08, .18]
Interpersonal resources .06 (.04) .04ns .08 (.03) .09** .12 (.03) .11 .12 (.03) .14 .17 (.04) .15
[.13, .02] [.02, .14] [.06, .18] [.07, .18] [.10, .24]
Warmth .04 (.03) .03ns .06 (.02) .07* .08 (.03) .08** .09 (.02) .12 .12 (.03) .12
[.11, .02] [.01, .10] [.02, .13] [.05, .13] [.06, .18]
Structure .04 (.04) .03ns .06 (.03) .07* .10 (.03) .09** .10 (.03) .11** .17 (.04) .15
[.12, .04] [.01, .12] [.04, .16] [.04, .15] [.10, .25]
Autonomy support .07 (.04) .05* .07 (.03) .08** .10 (.03) .11 .10 (.02) .12 .13 (.03) .12
[.14, .00] [.02, .12] [.05, .16] [.05, .14] [.07, .18]
Emotional reactivity – – .04 (.02) .07** .05 (.02) .07** .03 (.02) .06* .08 (.02) .10
[.07, .02] [.09, .02] [.06, .00] [.12, .04]

Note. N ¼ 1020 students in grades three through six. All regressions significant at p < .001 except as noted, **p < .01; *p < .05; ^p < .10.

respectively). Also as expected, students’ emotional reactivity was achievement. Correlations among students’ report card grades and
negatively and significantly related to motivational resilience in components of their motivational resilience at both time points for
both fall and spring (average r ¼ .36). Of all the antecedents, the subset of students for whom achievement data were available
students’ perceptions of competence and teachers’ provision of (n ¼ 365; see Table 2) revealed that, as expected, motivational
structure seemed to have the strongest concurrent relationships resilience was positively and significantly related to academic per-
with motivational resilience. formance both in fall and in spring, slightly higher in spring
Most interesting were analyses examining whether these pro- (r ¼ .29) than in fall (r ¼ .19).
posed antecedents, both individually and in combination, could pre- Of greatest interest were multiple regressions examining
dict changes in students’ motivational resilience over the school whether students’ motivational resilience in fall predicted changes
year. Despite the high stability of motivational resilience, multiple in their academic achievement from fall to spring. As expected,
regression analyses revealed support for each of these antecedents despite the high stability in achievement from fall to spring
and their sub-components as significant predictors of changes in (r ¼.75), motivational resilience in fall did predict students’
students’ motivational resilience (see Table 3). Students who were achievement in spring, even when controlling for fall achievement
high in personal or interpersonal resources in fall showed increases in scores, b ¼ .10, B ¼ .40, SE ¼ .14, p < .01, 95% CI (.12, .67).
their motivational resilience across time, whereas students who Additionally, to examine whether reciprocal effects on students’
reported high initial levels of emotional reactivity tended to decrease motivational resilience were evident, we conducted a second mul-
in motivational resilience from fall to spring. There appeared to be tiple regression analysis in which academic achievement was used
some specificity in the feedforward effects; students’ personal as a predictor of students’ motivational resilience in spring, con-
resources primarily predicted their engagement and re-engagement, trolling for their previous levels of motivational resilience; this
while interpersonal resources most strongly predicted changes in regression also approached significance in the predicted direction,
coping profiles and re-engagement. Students’ personal resources pre- b ¼ .06, B ¼ .02, SE ¼ .01, p ¼ .06, 95% CI (.00, .05), despite the
dicted changes in their emotional reactivity from fall to spring. At the high stability in motivational resilience from fall to spring (r ¼ .71).
same time, however, teacher support did not predict changes in emo-
tional reactivity, indicating that students’ interpersonal resources did
not protect them from feeling bad when things went wrong. Reciprocal effects of motivational resilience on
changes in personal and interpersonal resources
Reciprocal relationship between motivational Because of the dynamic relationship between teachers and students
in the classroom, we expected to find feedback effects from stu-
resilience and academic achievement
dents’ motivational resilience and emotional reactivity to changes
The next set of reciprocal relationships examined more closely in their personal and interpersonal resources. In line with these
were between students’ motivational resilience and their academic expectations, and despite high stabilities in the dependent variables,
Pitzer and Skinner 23

Table 4. Multiple regressions in which motivational resilience and emotional reactivity in fall predict changes in students’ personal and interpersonal
resources from fall to spring.

Changes from fall to spring in:

Overall personal resources Relatedness Competence Autonomy

b (SE) b (SE) b (SE) b (SE)

Predictor in fall 95% CI ß 95% CI ß 95% CI ß 95% CI ß

Motivational resilience .26 (.04) .24 .36 (.05) .24 .27 (.04) .24 .09 (.04) .06*
[.18, .35] [.27, .46] [.19, .36] [.00, .18]
Emotional reactivity .05 (.02) .08** .09 (.03) .10 .05 (.02) .06** .05 (.02) .05*
[.08, .02] [.14, .04] [.08, .01] [.09, .01]

Overall interpersonal resources Warmth Structure Autonomy support

b (SE) ß b (SE) ß b (SE) ß b (SE) ß

95% CI 95% CI 95% CI 95% CI

Motivational resilience .05 (.02) .09* .15 (.05) .11** .15 (.04) .13 .20 (.04) .16
[.09, .02] [.06, .24] [.08, .23] [.12, .29]
Emotional reactivity .05 (.02) .07** .07 (.02) .08** .05 (.02) .07** .07 (.02) .09**
[.09, .02] [.11, .02] [.09, .02] [.11, .03]

Note. N ¼ 1020 students in Grades 3 through 6. All coefficients significant at p < .001 except as noted, **p < .01; *p < .01.

multiple regression analyses showed that both motivational resili- correlations and multiple regressions were consistent with the
ence and emotional reactivity predicted changes in students’ per- notion of catastrophizing as a marker of risk: Students who reported
sonal and interpersonal resources from fall to spring (see Table 4). higher levels of catastrophizing also showed significantly higher
Students who began the school year high in motivational resilience levels of emotional reactivity (average r ¼ .69) and lower levels of
experienced increases in each of their personal resources over time, motivational resilience (average r ¼ .67) than students who did
including higher levels of perceived relatedness, competence, and not have this motivational risk factor (see Table 2). Moreover, high
autonomy, while their more motivationally vulnerable peers catastrophizing was associated with increases in emotional reactiv-
showed the opposite pattern. Likewise, students who began the year ity from fall to spring, b ¼ .28, B ¼ .29, SE ¼ .04, p < .001, 95% CI
high in emotional reactivity experienced decreases in personal and (.22, .36) and decreases in motivational resilience, b ¼ .09,
interpersonal resources as the year progressed, whereas students B ¼ .06, SE ¼ .02, p < .01, 95% CI (.10, .02) across the school
who were less emotionally reactive reported increases in both year.
resources from fall to spring. And (in contrast to the structural To examine the effects of teacher support on changes in stu-
model in which this link was not significant), students who reported dents’ established motivational systems, we compared the subset of
high levels of motivational resilience in fall experienced increased students who began the school year reporting high (i.e., above
warmth, structure, and autonomy support from their teachers as the median) levels of catastrophizing appraisals with those who were
year progressed, whereas students low in motivational resilience less motivationally at risk (i.e., who reported low—below med-
attracted fewer of these interpersonal resources. ian—levels of catastrophizing), to determine whether changes in
There appeared to be some specificity in these feedback loops students’ motivational resilience over the school year differed as a
indicating that students who took more initiative and bounced back function of the level of teacher support they received. Specifically,
were in turn granted additional freedoms by teachers (i.e., auton- for students who were motivationally at-risk at the beginning of the
omy support). Moreover, students’ motivational resilience in fall year, we looked to see whether those who received consistently
had a particularly strong effect on changes in their feelings of high or increasing levels of teacher support (i.e., who had suppor-
competence as the school year progressed; students who showed tive teachers) would be able to recover their motivational resilience,
a greater capacity to rebound from struggles in the fall subsequently whereas those who received low or decreasing levels of support
experienced higher perceptions of control, whereas students who (i.e., who had unsupportive teachers) would stay caught in the
were initially less motivationally resilient reported experiencing negative motivational space. Conversely, for students who began
increases in helplessness. the school year without such motivational risk, we examined
whether those with unsupportive teachers would become increas-
ingly more vulnerable, while those with supportive teachers would
Effects of teacher support on at-risk motivational maintain their motivational resilience. Ultimately, we wondered if
teacher support could intervene in these self-amplifying systems to
systems help pull students out of detrimental feedback loops.
In order to examine the role of teachers in re-shaping the trajec- As can be seen in Figure 3, students who began the year moti-
tories of students who were at-risk for the development of motiva- vationally at-risk (i.e., high in catastrophizing) reported signifi-
tional vulnerabilities, students’ reports of catastrophizing appraisals cantly lower levels of motivational resilience than students who
were used as an indicator of existing motivational risk. Although showed fewer risk factors in fall, t(1018) ¼ 28.48, p < .001, 95%
students reported relatively low levels of catastrophizing, both CI (0.46, 0.53). Repeated measures analyses of variance
24 International Journal of Behavioral Development 41(1)

4
Low Catastrophizing,
Supportive Teacher

Motivational Resilience
3.5 (n = 346)
Low Catastrophizing,
Unsupportive Teacher
3 (n = 179)
High Catastrophizing,
Supportive Teacher
2.5 (n = 164)
High Catastrophizing,
2 Unsupportive Teacher
Fall Spring (n = 331)

Figure 3. Mean levels of students’ motivational resilience across the school year according to initial vulnerability status and level of teacher support over the
school year. Responses could range from (1) Not at all true for me to (4) Very true for me.

(ANOVAs) showed significant interaction effects between level of with increasing feelings of relatedness, autonomy, and especially
teacher support (supportive vs. unsupportive) and time point (fall competence. They also reported experiencing increases in warmth,
vs. spring) for each risk group, F(1, 523) ¼ 49.29, p < .001, for the structure, and especially autonomy support from their teachers. In
low-catastrophizing group, and F(1, 493) ¼ 54.26, p < .001 for the contrast, students who began the year with greater motivational
high-catastrophizing group, indicating that changes in students’ vulnerabilities were likely to show small declines in their achieve-
motivational resilience across time depended on which type of ment from fall to spring accompanied by decreases in their positive
teacher support they experienced. Specifically, for the high-risk self-perceptions; they also experienced their teachers as withdraw-
(i.e., high catastrophizing) students (n ¼ 495), those with suppor- ing from them and becoming more controlling over time.
tive teachers increased in motivational resilience from fall to Together, these feedforward and feedback effects may form
spring, paired t(163) ¼ 7.83, p < .001, 95% CI (0.20, 0.12), dynamic, potentially self-perpetuating cycles, such that students who
whereas those who reported unsupportive teachers remained low in start the school year high in personal and interpersonal resources are
motivational resilience, paired t(330) ¼ 1.71, ns, 95% CI (0.00, likely to exhibit higher levels of motivational resilience, which in
0.05). For the low-risk (i.e., low-catastrophizing) students turn elicits increases in those resources. In contrast, students who are
(n ¼ 525), also consistent with expectations, those with supportive initially low in motivational resilience tend to experience erosion of
teachers stayed high in motivational resilience across the school their existing resources over time, which has ever increasing deleter-
year, paired t(345) ¼ .82, ns, 95% CI (0.02, 0.04), whereas those ious effects. Taken together with evidence that the internal dynamics
who reported unsupportive teachers decreased from fall to spring, of motivational resilience are also self-sustaining (Skinner et al.,
paired t(178) ¼ 8.09, p < .001, 95% CI (0.14, 0.23). Moreover, 2013a), it seems that, without outside intervention, these virtuous
by the end of the school year, students who reported high- and vicious feedback cycles are likely to persist, in the classic ‘‘rich
catastrophizing appraisals in fall but received consistently high or get richer, poor get poorer’’ dynamic. However, findings suggested
increasing levels of teacher support from fall to spring actually that teacher support might be a good candidate as an external lever,
reported significantly higher levels of motivational resilience than because (compared to personal resources) it is not as tightly coupled
their classmates who began the school year with few risk factors but to students’ previous motivational resilience and academic perfor-
received consistently low or decreasing levels of teacher support, mance. In fact, even though multiple regressions suggested a feed-
t(341) ¼ 2.76, p < .01, 95% CI (0.02, 0.14). back effect from motivational resilience to increases in teacher
support from fall to spring, no such feedback effect was apparent
in the overall structural model, when other important factors, such as
Discussion students’ self-appraisals, were taken into consideration.
Consistent with other research that documents reciprocal feedback Hence, a key interest of this study was to examine whether teacher
among the factors contributing to students’ engagement (e.g., support can reshape these otherwise self-sustaining motivational sys-
Green et al., 2012; Jang et al., 2012; Van Ryzin, 2011), the findings tems. Comparisons of students who were motivationally at-risk (as
of this study reveal the dynamic relationships that exist among marked by high levels of catastrophizing appraisals) with students
students’ motivational resilience and their social contexts, personal who showed less risky profiles demonstrated the important role teach-
resources, emotional reactivity, and achievement outcomes. This ers play in classroom dynamics: Students who began the school year
study confirmed the typical strong feedforward effects from stu- high in catastrophizing appraisals but received high levels of teacher
dents’ self-system processes (i.e., personal resources) and also support were able to bounce back such that they ended the year with
documented effects of teacher support (i.e., interpersonal resources) higher levels of motivational resilience than even students who began
on changes in their motivational resilience. Students’ motivational with less risky profiles but received low levels of teacher support.
resilience, in turn, predicted changes in their academic achievement
over the school year. Perhaps surprisingly, feedback effects were
Study strengths and limitations
also found for many of the links in the proposed model. Students
who evinced high motivational resilience in fall showed small Of course, these findings must be interpreted in light of the study’s
improvements in their achievement as the year progressed along strengths and limitations. Although it is a significant strength of this
Pitzer and Skinner 25

study that it is embedded in the larger SDT framework, the con- findings suggest that the role of emotional reactivity is more com-
ceptualization does not encompass all the constructs that potentially plex (Lemos, 2002). Based on students’ reports of how upset they
are relevant to motivational resilience. For example, students’ goal become following setbacks in school, previous studies showed that
orientations and mindsets (Dweck, 2006) likely play key roles in students who were disaffected from and doing poorly in school
how they appraise and respond to challenges and setbacks in school, could indeed be highly emotionally reactive, but so too could stu-
as would their self-regulatory strategies (Schunk & Zimmerman, dents who were highly engaged and doing well. And, although
2012) and other factors important to academic buoyancy, such as emotional reactivity did not prevent students from re-engaging with
high levels of planning or support from parents and the community challenging academic tasks, it did make them more likely to give up
(Martin & Marsh, 2008). (Skinner et al., 2013a). Thus, emotional reactivity, at least as mea-
In terms of measures, all of the information in the current study sured in this study, seems to contain not only elements of risk, but
was acquired via surveys, relying on students’ own self-reports of also elements of how much students care (Lemos, 2002): If things
their experiences. Although students’ perspectives are crucial to are going well, emotional reactivity seems to be a marker for com-
access internal processes, many constructs are also observable in mitment or investment, but if things are going poorly it can exacer-
the classroom, and future studies would benefit from the inclusion bate ongoing negative cycles and undermine students’ motivational
of observational methods and other reporters’ perspectives on stu- resources.
dent engagement and teacher–student interactions. Moreover, in Additional complexities in the functioning of emotional reactiv-
terms of design, it will be important in future studies to examine ity were uncovered in the current study. Neither academic success
how motivational resilience and its reciprocal relationships with nor teacher support prevented students from feeling bad after a
students’ personal and interpersonal resources operate over longer stressful event. Students who reported high levels of teacher sup-
periods, potentially accumulating in their effects as self-amplifying port were just as likely to report being upset by difficulties as the
cycles play out over time. Experimental designs will also be essen- year progressed as students who reported less supportive relation-
tial to assess issues of causality more directly. ships. And, although emotional reactivity seemed to have some
In terms of sampling, having the participation of an entire school impact on changes in students’ motivational resilience, it did not
district is a significant strength of this study. However, the district mediate the stronger effects of students’ personal and interpersonal
consisted of predominantly working-class, Caucasian families. resources on their motivational resilience. Instead, reactivity seems
Classrooms themselves are also changing, with policy changes and to be buffered only by students’ own personal resources, suggesting
increasing integration of technology in learning environments. that teacher support or academic accomplishments will be effective
According to SDT, these should be universal motivational princi- in reducing emotional reactivity only to the extent they help
ples (Deci & Ryan, 1985), but, it will be necessary to replicate this strengthen students’ own internal assets, such as a sense of belong-
study on more diverse samples in order to assess their generaliz- ingness, confidence, or autonomy. Moreover, results from the mul-
ability over time and populations. tiple regressions suggested that students who reported higher levels
of emotional reactivity in the fall also experienced losses over the
school year in all three kinds of teacher support.
Implications and future research More work is needed to understand the role students’ emotional
This study adds to a growing understanding of how the dynamics of reactions play in classroom dynamics. Because it provides obser-
motivational resilience function. Previous research provided initial vable information to teachers, emotional reactivity may be a par-
evidence suggesting that students’ ongoing engagement fuels their ticularly important target for future study. In such work, it may be
reactions to challenges: Students who are enthusiastic and actively instructive to further examine what types of reactivity are being
involved in academic tasks tend to use adaptive coping strategies to expressed. A teacher may respond differently, for example, to a
bounce back from difficulties, contributing to a virtuous feedback student who is worried or sad than one who is angry or frustrated,
loop that sustains engagement. In contrast, students who begin the and the same type of support offered for varying emotional
school year relatively more disaffected show increasing emotional responses would likely have differential effects. Most importantly,
reactivity, maladaptive coping, and eventually, giving up, which future investigations may benefit from the use of person-centered
together form a detrimental self-reinforcing cycle that can be dif- analysis to identify different profiles of motivational resilience,
ficult to escape (Skinner et al., 2013a). Building on this earlier distinguishing students who are high versus low on emotional reac-
work, the current study focused on how these internal dynamics tivity and on the components of resilience (engagement, coping,
can be influenced by external contributors such as students’ own and re-engagement) in order to better elucidate how these features
self-system processes and supports from their classroom teachers. function in combination (Luo, Hughes, Liew, & Kwok, 2009; Wang
Taken together, these studies provide evidence for almost all of the & Peck, 2013).
links in the proposed model (see Figure 1); and although findings Ideally, students would be motivationally resilient and exhibit
suggest that neither students’ achievement nor their close relation- low levels of emotional reactivity, bouncing back from setbacks
ships with teachers protected them from increasing emotional reac- without being derailed by their emotions, and instead just busily
tivity, results from the current study reveal that internal dynamics, learning from each experience. Indeed, such students may display
which are otherwise self-amplifying, can be reduced, and in some very different patterns of functioning than students who are gener-
cases reversed, by factors external to this system. ally motivationally resilient but also highly reactive. In the same
vein, students who are low in motivational resilience but who nev-
Emotional reactivity. In early iterations of the model of motiva- ertheless get very upset when they run into difficulties may at least
tional resilience, we assumed that high levels of emotional reactiv- still show the ‘‘spark’’ of caring about their academic work, and so
ity signaled a motivational vulnerability, and would interfere with their enthusiasm may be more easily re-ignited with well-calibrated
students’ capacities to cope adaptively and bounce back after fail- support. The toughest combination to rekindle may be a profile that
ure. However, taken together with previous research, current is low on motivational resilience and low in emotional reactivity,
26 International Journal of Behavioral Development 41(1)

perhaps manifest as apathy or amotivation, which is particularly the extent to which motivational resilience is either preserved or
detrimental (Ratelle, Guay, Vallerand, Larose, & Senécal, 2007). undermined by the kinds of support they subsequently receive.
Future studies could also examine profiles that incorporate stu- Such studies, especially if they include at least three time points,
dents’ achievement—emotional reactivity may influence students’ may be well suited for following up on some of the findings from
motivational resilience differently for students with consistently the present investigation that suggest multiple meditational pro-
high achievement compared to those with persistent academic cesses within students’ motivational systems.
difficulties. Certainly, supporting students’ motivational resilience is not an
It is also possible that, compared to emotional reactivity, a more easy task. Without mindful intention to positively intervene in these
important element to incorporate into the model of motivational processes, it is all too easy for teachers to participate in ways that
resilience would be recovery from emotional distress. It seems sustain or amplify existing negative motivational dynamics. It is
plausible that it may not be whether (or how far) students fall understandable that one default for teachers may be to recipro-
emotionally, but rather how quickly they recover or bounce back cate—providing more support for motivated students while at the
that truly matters to their motivational resilience. To more thor- same time withdrawing support and increasing pressure on students
oughly investigate this idea, future studies would benefit from the who are actively disaffected in class, emotionally reactive, cope
inclusion of a measure of emotional bounceback, or what Davidson maladaptively, or give up in the face of challenges. After all, teach-
(1998) refers to as affective chronometry. ers too have needs to feel related, competent, and autonomous
(Klassen, Perry, & Frenzel, 2012; Niemiec & Ryan, 2009; Spilt,
Reciprocal effects. In research on coping and resilience, examina- Koomen, & Thijs, 2011), and dealing with demands such as student
tion of feedforward effects is standard practice. Researchers typi- disruption, emotional outbursts, or helpless behaviors can directly
cally attempt to determine the kinds of factors in students’ school or undermine each of these needs (Furrer, Skinner, & Pitzer, 2014).
family contexts that encourage them to try hard, cope well, and However, the knowledge that these vulnerabilities will otherwise
bounce back. However, the findings from this study underscore the multiply over time can motivate the urgency of early intervention
importance of looking for both feedforward and feedback effects: efforts.
Both directions of effects were evident in the proposed model. It Teachers will themselves need support if they are to participate
seems that when students perform well in school, this learning and in these dynamics in ways that counteract vulnerability and sustain
success naturally feed back into their motivational resilience. In the resilience. Educators may need training to be able to simultane-
same vein, the experience of motivational resilience enhances their ously monitor all the components of the complex motivational
personal resources and may even influence the availability of their system, vigilantly watching for multiple indicators of vulnerability
interpersonal resources. To more thoroughly elucidate the function- and attempting to provide students with appropriately well-tuned
ing of such complex dynamics over time, future studies would support. It is essential for schools and administrators to recognize
benefit from designs that allow researchers to measure these con- the important role students’ social contexts play both in bringing
structs on time scales more closely aligned with the actual timing of students out of vulnerability and in helping those who are already
how these reciprocal loops likely play out and stabilize (i.e., across doing well to maintain their momentum. Facilitating teachers’
days or weeks, or even moment by moment, rather than months). capacities to provide students with optimally calibrated support can
have powerful effects on students’ motivational systems. The pres-
Importance of self-appraisals and of teacher support. Results ent study demonstrated that teachers can indeed provide compen-
from the current study highlight the centrality of students’ self- satory dynamics within this motivational system, which
appraisals to the functioning of their motivational systems. In this encouragingly, once re-calibrated, can become fueled by its own
sample of third- through sixth-graders, self-system processes were self-sustaining nature. Teachers are in a unique position to inten-
the only kinds of resources that buffered students’ emotional reac- tionally intervene in this process, and their support has the capacity
tivity and they were the strongest predictors of motivational resi- to have a lasting impact on students’ profiles of motivational
lience. In addition, they represented a key pathway through which resilience.
interpersonal resources exerted their effects. Self-appraisals were
also more tightly coupled in feedback effects: They registered the
effects of both previous academic performance and motivational Funding
resilience, suggesting an amplifying dynamic. For students who are This study was supported by the National Institutes of Mental
already doing well, the self-reinforcing feedback loops are benefi- Health (Training Grant No. 527594), the National Institute of Child
cial, but that same cyclical dynamic can also prevent students from Health and Human Development (Research Grant No. HD19914),
escaping an existing adverse feedback loop organized around moti- and the W. T. Grant Foundation.
vational vulnerability and self-deprecating appraisals.
Findings from this study also underscore the crucial role played
by teachers in the dynamics of their students’ motivational resili- References
ence. Importantly, these typically self-sustaining dynamics should Blondal, K. S., & Adalbjarnardottir, S. (2012). Student disengagement
not be taken as evidence that the system is fixed or impervious to in relation to expected and unexpected educational pathways. Scan-
intervention. In fact, the present study suggests that this system may dinavian Journal of Educational Research, 56(1), 85–100. doi:10.
indeed be open to influence—teachers are not as likely to be drawn 1080/00313831.2011.568607
into vicious cycles, and they may even disrupt maladaptive Boekaerts, M. (1993). Being concerned with well-being and with learn-
dynamics by providing compensatory well-calibrated support. Next ing. Educational Psychologist, 28(2), 149–167. doi:10.1207/
steps for examining this process may include studies that are more s15326985ep2802_4
longitudinal in nature, where researchers could observe students as Boekaerts, M. (2002). Coping with challenge. Stress, Anxiety, and Cop-
they encounter new teachers across different school years and note ing, 15(4), 321–326. doi:10.1080/1061580021000056492
Pitzer and Skinner 27

Brown, J. M., O’Keefe, J., Sanders, S. H., & Baker, B. (1986). Devel- Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covar-
opmental changes in children’s cognition to stressful and painful iance structure analysis: Conventional criteria versus new alterna-
situations. Journal of Pediatric Psychology, 11, 343–357. doi:10. tives. Structural Equation Modeling, 6(1), 1–55. doi:10.1080/
1093/jpepsy/11.3.343 10705519909540118
Causey, D. L., & Dubow, E. F. (1992). Development of a self-report Hughes, J., & Kwok, O.-M. (2007). Influence of student-teacher and
coping measure for elementary school children. Journal of Clinical parent-teacher relationships on lower achieving readers’ engage-
Child Psychology, 21, 47–59. ment and achievement in the primary grades. Journal of Educa-
Christenson, S., Reschly, A., & Wylie, C. (Eds.). (2012). Handbook of tional Psychology, 99(1), 39–51. doi:10.1037/0022-0663.99.1.39
research on student engagement. New York, NY: Springer Science. IBM Corp. (2015). SPSS Statistics for Windows, Version 23.0 [Com-
Connell, J. P., Spencer, M. B., & Aber, J. L. (1994). Educational risk puter software]. Armonk, NY: IBM Corp.
and resilience in African-American youth: Context, self, action, and Jang, H., Kim, E. J., & Reeve, J. (2012). Longitudinal test of
outcomes in school. Child Development, 65, 493–506. doi:10.1111/ self-determination theory’s motivation mediation model in a natu-
j.1467-8624.1994.tb00765.x rally occurring classroom context. Journal of Educational Psychol-
Connell, J. P., & Wellborn, J. G. (1991). Competence, relatedness, ogy, 104(4), 1175–1188. doi:10.1037/a0028089
and autonomy: A motivational analysis of self-system processes. Jimerson, S. R., Campos, E., & Greif, J. L. (2003). Towards an under-
In M. R. Gunnar & L. A. Sroufe (Eds.), Self processes in devel- standing of definitions and measures of school engagement and
opment: Minnesota symposium on child psychology (pp. 43–77). related terms. The California School Psychologist, 8, 7–28.
Hillsdale, NJ: Erlbaum. Kaplan, A., & Midgley, C. (1999). The relationship between per-
Davidson, R. J. (1998). Affective style and affective disorders: Perspec- ceptions of the classroom goal structure and early adolescents’
tives from affective neuroscience. Cognition and Emotion, 12(3), affect in school: The mediating role of coping strategies. Learn-
307–330. doi:10.1080/026999398379628 ing and Individual Differences, 11(2), 187–212. doi:10.1016/
Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and S1041-6080(00)80005-9
self-determination in human behavior. New York, NY: Plenum Klassen, R. M., Perry, N. E., & Frenzel, A. C. (2012). Teachers’ relat-
Press. edness with students: An underemphasized component of teachers’
Dweck, C. S. (2006). Mindset: The new psychology of success. New basic psychological needs. Journal of Educational Psychology,
York, NY: Random House. 104(1), 150–165. doi:10.1037/a0026253
Fredricks, J. A., Blumenfeld, P. C., & Paris, A. H. (2004). School Klem, A. M., & Connell, J. P. (2004). Relationships matter: Linking
engagement: Potential of the concept, state of the evidence. Review teacher support to student engagement and achievement. The Jour-
of Educational Research, 74(1), 59–109. doi:10.3102/ nal of School Health, 74(7), 262–273. doi:10.1111/j.1746-1561.
00346543074001059 2004.tb08283.x
Friedel, J. M., Cortina, K. S., Turner, J. C., & Midgley, C. (2007). Lau, S., & Nie, Y. (2008). Interplay between personal goals and class-
Achievement goals, efficacy beliefs and coping strategies in mathe- room goal structures in predicting student outcomes: A multilevel
matics: The roles of perceived parent and teacher goal emphases. analysis of person-context interactions. Journal of Educational Psy-
Contemporary Educational Psychology, 32(3), 434–458. doi:10. chology, 100(1), 15–29. doi:10.1037/0022-0663.100.1.15
1016/j.cedpsych.2006.10.009 Lemos, M. S. (2002). Social and emotional processes in the classroom
Furrer, C. J., & Skinner, E. A. (2003). Sense of relatedness as a factor in setting: A goal approach. Anxiety, Stress & Coping, 15(4), 383–400.
children’s academic engagement and performance. Journal of doi:10.1080/1061580021000056537
Educational Psychology, 95(1), 148–162. doi:10.1037/0022-0663. Leung, G. S. M., & He, X. (2010). Resourcefulness: A protective
95.1.148 factor buffer against the academic stress of school-aged children.
Furrer, C. J., Skinner, E. A., & Pitzer, J. R. (2014). The influence of Educational Psychology, 30(4), 395–410. doi:10.1080/
teacher and peer relationships on students’ classroom engagement 01443411003682574
and everyday resilience. In D. J. Shernoff & J. Bempechat (Eds.), Li, Y., & Lerner, R. M. (2011). Trajectories of school engagement
Engaging youth in schools: Empirically-based models to guide during adolescence: Implications for grades, depression, delin-
future innovations (pp. 101–123). New York, NY: Teachers College quency, and substance use. Developmental Psychology, 47(1),
Press. 233–247. doi:10.1037/a0021307
Graham, J. W. (2009). Missing data analysis: Making it work in the real Little, T. D., Cunningham, W. A., Shahar, G., & Widaman, K. F.
world. Annual Review of Psychology, 60, 549–576. doi:10.1146/ (2002). To parcel or not to parcel: Exploring the question, weighing
annurev.psych.58.110405.085530 the merits. Structural Equation Modeling, 9(2), 151–173. doi:10.
Green, J., Liem, G. A. D., Martin, A. J., Colmar, S., Marsh, H. W., & 1207/S15328007SEM0902_1
McInerney, D. (2012). Academic motivation, self-concept, engage- Luo, W., Hughes, J. N., Liew, J., & Kwok, O. (2009). Classifying
ment, and performance in high school: Key processes from a long- academically at-risk first graders into engagement types: Associa-
itudinal perspective. Journal of Adolescence, 35(5), 1111–1122. tion with long-term achievement trajectories. The Elementary
doi:10.1016/j.adolescence.2012.02.016 School Journal, 109(4), 380–405. doi:10.1086/593939
Henry, K. L., Knight, K. E., & Thornberry, T. P. (2012). School disen- Lynch, M., & Cicchetti, D. (1997). Children’s relationships with adults
gagement as a predictor of dropout, delinquency, and problem sub- and peers: An examination of elementary and junior high school
stance use during adolescence and early adulthood. Journal of Youth students. Journal of School Psychology, 35(1), 81–99. doi:10.1016/
and Adolescence, 41(2), 156–166. doi:10.1007/s10964-011-9665-3 S0022-4405(96)00031-3
Hess, R. S., & Copeland, E. P. (2001). Students’ stress, coping strate- MacCann, C., Fogarty, G. J., Zeidner, M., & Roberts, R. D. (2011).
gies, and school completion: A longitudinal perspective. School Coping mediates the relationship between emotional intelligence
Psychology Quarterly, 16(4), 389–405. doi:10.1521/scpq.16.4. (EI) and academic achievement. Contemporary Educational Psy-
389.19899 chology, 36(1), 60–70. doi:10.1016/j.cedpsych.2010.11.002
28 International Journal of Behavioral Development 41(1)

Mantzicopoulos, P. (1997). How do children cope with school failure? Ryan, R. M., & Connell, J. P. (1989). Perceived locus of causality and
A study of social/emotional factors related to children’s coping internalization: Examining reasons for acting in two domains. Jour-
strategies. Psychology in the Schools, 34(3), 229–237. doi:10. nal of Personality and Social Psychology, 57, 749–761. doi:10.
1002/(SICI)1520-6807(199707)34:3<229::AID-PITS5>3.0.CO;2-J 1037/0022-3514.57.5.749
Martin, A. J. (2013). Academic buoyancy and academic resilience: Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the
Exploring ‘‘everyday’’ and ‘‘classic’’ resilience in the face of aca- facilitation of intrinsic motivation, social development, and well-
demic adversity. School Psychology International, 34(5), 488–500. being. American Psychologist, 55(1), 68–78. doi:10.1037/
doi:10.1177/0143034312472759 0003-066X.55.1.68
Martin, A. J., & Marsh, H. W. (2008). Academic buoyancy: Towards an Ryan, R. M., & Deci, E. L. (2009). Promoting self-determined
understanding of students’ everyday academic resilience. Journal of school engagement: Motivation, learning, and well-being. In K. R.
School Psychology, 46(1), 53–83. doi:10.1016/j.jsp.2007.01.002 Wentzel & A. Wigfield (Eds.), Handbook on motivation at school
Martin, A. J., & Marsh, H. W. (2009). Academic resilience and aca- (pp. 171–196). New York, NY: Routledge.
demic buoyancy: Multidimensional and hierarchical conceptual Schunk, D. H., & Zimmerman, B. J. (2012). Motivation and
framing of causes, correlates and cognate constructs. Oxford Review self-regulated learning: Theory, research, and applications. New
of Education, 35, 353–370. doi:10.1080/03054980902934639 York, NY: Routledge.
Morrison, G. M., Robertson, L., Laurie, B., & Kelly, J. (2002). Protec- Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view of the
tive factors related to antisocial behavior trajectories. Journal of state of the art. Psychological Methods, 7(2), 147–177. doi:10.
Clinical Psychology, 58(3), 277–290. doi:10.1002/jclp.10022 1037//1082-989X.7.2.147
Niemiec, C. P., & Ryan, R. M. (2009). Autonomy, competence, and Skinner, E. A., & Belmont, M. J. (1993). Motivation in the classroom:
relatedness in the classroom: Applying self-determination theory to Reciprocal effects of teacher behavior and student engagement
educational practice. Theory and Research in Education, 7(2), across the school year. Journal of Educational Psychology, 85(4),
133–144. doi:10.1177/1477878509104318 571–581. doi:10.1037/0022-0663.85.4.571
Nurmi, J. E., & Kiuru, N. (2015). Introduction to the special section on Skinner, E. A., Furrer, C. J., Marchand, G., & Kindermann, T. (2008).
student effects on teachers’ behaviors and attitudes. International Engagement and disaffection in the classroom: Part of a larger
Journal of Behavioral Development, 39(5), 389–390. doi:10.1177/ motivational dynamic? Journal of Educational Psychology,
0165025415591231 100(4), 765–781. doi:10.1037/a0012840
Ratelle, C. F., Guay, F., Vallerand, R. J., Larose, S., & Senécal, C. Skinner, E. A., Kindermann, T. A., & Furrer, C. J. (2009). A motiva-
(2007). Autonomous, controlled, and amotivated types of academic tional perspective on engagement and disaffection: Conceptualiza-
motivation: A person-oriented analysis. Journal of Educational tion and assessment of children’s behavioral and emotional
Psychology, 99(4), 734–746. doi:10.1037/0022-0663.99.4.734 participation in academic activities in the classroom. Educational
Raufelder, D., Kittler, F., Braun, S. R., Lätsch, A., Wilkinson, R. P., & and Psychological Measurement, 69(3), 493–525. doi:10.1177/
Hoferichter, F. (2014). The interplay of perceived stress, 0013164408323233
self-determination and school engagement in adolescence. School Skinner, E. A., & Pitzer, J. R. (2012). Developmental dynamics of
Psychology International, 35(4), 405–420. doi:10.1177/ student engagement, coping, and everyday resilience. In S. L. Chris-
0143034313498953 tenson, A. L. Reschly & C. Wylie (Eds.), Handbook of research on
Reeve, J. (2009). Why teachers adopt a controlling motivating style student engagement (pp. 21–45). New York, NY: Springer Science.
toward students and how they can become more autonomy suppor- Skinner, E. A., Pitzer, J. R., & Steele, J. S. (2013a, April). Academic
tive. Educational Psychologist, 44(3), 159–175. doi:10.1080/ coping in elementary school: The dynamics of motivational resili-
00461520903028990 ence. Poster presented at the biennial meeting of the Society for
Reeve, J. (2012). A self-determination theory perspective on student Research in Child Development, Seattle, WA.
engagement. In S. L. Christenson, A. L. Reschly & C. Wylie (Eds.), Skinner, E. A., Pitzer, J. R., & Steele, J. S. (2013b). Coping as part of
Handbook of research on student engagement (pp. 149–172). New motivational resilience in school: A multidimensional measure of
York, NY: Springer. families, allocations, and profiles of academic coping. Educational
Reeve, J., Jang, H., Carrell, D., Jeon, S., & Barch, J. (2004). Enhancing and Psychological Measurement, 73(5), 803–835. doi:10.1177/
students’ engagement by increasing teachers’ autonomy support. 0013164413485241
Motivation and Emotion, 28(2), 147–169. doi:10.1023/B: MOEM. Skinner, E. A., Wellborn, J. G., & Connell, J. P. (1990). What it takes to
0000032312.95499.6f do well in school and whether I’ve got it: A process model of
Reschly, A. L., Huebner, E. S., Appleton, J. J., & Antaramian, S. perceived control and children’s engagement and achievement in
(2008). Engagement as flourishing: The contribution of positive school. Journal of Educational Psychology, 82(1), 22–32. doi:10.
emotions and coping to adolescents’ engagement at school and with 1037//0022-0663.82.1.22
learning. Psychology in the Schools, 45(5), 419–431. doi:10.1002/ Skinner, E. A., Zimmer-Gembeck, M. J., & Connell, J. P. (1998). Indi-
pits.20306 vidual differences and the development of perceived control. Mono-
Reyes, M. R., Brackett, M. A., Rivers, S. E., White, M., & Salovey, P. graphs of the Society for Research in Child Development, 63(2–3),
(2012). Classroom emotional climate, student engagement, and aca- 1–220. doi:10.2307/1166220
demic achievement. Journal of Educational Psychology, 104(3), Spangler, G., Pekrun, R., Kramer, K., & Hofmann, H. (2002). Students’
700–712. doi:10.1037/a0027268 emotions, physiological reactions, and coping in academic exams.
Roeser, R. W., Midgley, C., & Urdan, T. C. (1996). Perceptions of the Anxiety, Stress & Coping, 15(4), 413–432. doi:10.1080/
school psychological environment and early adolescents’ psycholo- 1061580021000056555
gical and behavioral functioning in school: The mediating role of Spilt, J. L., Koomen, H. M. Y., & Thijs, J. T. (2011). Teacher wellbeing:
goals and belonging. Journal of Educational Psychology, 88(3), The importance of teacher–student relationships. Educational Psy-
408–422. doi:10.1037//0022-0663.88.3.408 chology Review, 23(4), 457–477. doi:10.1007/s10648-011-9170-y
Pitzer and Skinner 29

Swanson, J., Valiente, C., Lemery-Chalfant, K., & O’Brien, T. C. Van Ryzin, M. J., Gravely, A. A., & Roseth, C. J. (2009). Autonomy,
(2011). Predicting early adolescents’ academic achievement, social belongingness, and engagement in school as contributors to adoles-
competence, and physical health from parenting, ego resilience, and cent psychological well-being. Journal of Youth and Adolescence,
engagement coping. The Journal of Early Adolescence, 31(4), 38(1), 1–12. doi:10.1007/s10964-007-9257-4
548–576. doi:10.1177/0272431610366249 Wang, M. T., & Peck, S. C. (2013). Adolescent educational success and
Tero, P. F., & Connell, J. P. (1984, April). When children think they’ve mental health vary across school engagement profiles. Developmen-
failed: An academic coping inventory. Paper presented at the Amer- tal Psychology, 49, 7, 1266–1276. doi:10.1037/a0030028
ican Educational Research Association, New Orleans, LA. Wentzel, K. R. (2009). Students’ relationships with teachers as moti-
Thompson, A., & Gaudreau, P. (2008). From optimism and pessimism vational contexts. In K. R. Wentzel & A. Wigfield (Eds.), Hand-
to coping: The mediating role of academic motivation. International book of motivation at school (pp. 301–322). New York, NY:
Journal of Stress Management, 15(3), 269–288. doi:10.1037/ Routledge.
a0012941 Wigfield, A., Eccles, J. S., Fredricks, J. A., Simpkins, S., Roeser, R., &
Upadyaya, K., & Salmela-Aro, K. (2013). Development of school Schiefele, U. (2015). Development of achievement motivation and
engagement in association with academic success and well-being engagement. In R. M. Lerner (Series Ed.) M. Lamb (Volume Ed.),
in varying social contexts. European Psychologist, 18 (3), 136–147. Handbook of child psychology and developmental science. Vol. 3.
doi:10.1027/1016-9040/a000143 Socioemotional processes (7th ed., pp. 657–700). New York, NY:
Van Ryzin, M. J. (2011). Protective factors at school: Reciprocal effects John Wiley.
among adolescents’ perceptions of the school environment, engage- Wolchik, S. A. & I. N. Sandler (Eds.). (1997). Handbook of chil-
ment in learning, and hope. Journal of Youth and Adolescence, 40, dren’s coping: Linking theory and intervention. New York,
1568–1580. doi:10.1007/s10964-011-9637-7 NY: Plenum.