ARTICLE SUMMARY 6

Growth Goal Setting in High School: A Large-Scale Study of Perceived Instructional Support, Personal Background Attributes,

and Engagement Outcomes

Andrew J. Martin1, Emma C. Burns2, Rebecca J. Collie1, Keiko C. P. Bostwick1, Anaïd Flesken3, and Ian McCarthy3

1 School of Education, University of New South Wales 2 School of Education, Macquarie University

3 Centre for Education Statistics and Evaluation, New South Wales Department of Education, New South Wales, Australia

The present investigation examined the role of teachers’ instructional support (student reports of rele- vance, organization and clarity, feedback-feedforward) in predicting students’ growth goal setting and, in turn, the roles of instructional support and growth goal setting in predicting students’ academic engagement (perseverance, aspirations, school attendance, homework behavior). Also examined was the question of whether the relationship between students’ background attributes and engagement is moder- ated by their growth goal setting (e.g., whether growth goal setting attenuates negative effects of low socioeconomic status). The sample comprised N = 61,879 students in grades 7–10 from schools across New South Wales, Australia. The results of structural equation modeling showed that perceived instruc- tional relevance and feedback-feedforward from teachers positively predicted students’ growth goal set- ting; that growth goal setting predicted gains in students’ perseverance, aspirations, and homework behavior; and that growth goal setting significantly mediated the relationship between perceived instruc- tional support and engagement. Additionally, growth goal setting appeared to significantly bolster some outcomes for low achieving students and students from low socioeconomic backgrounds. These findings add to the growing body of literature about the positive role of growth goal setting in students’ out- comes and provide direction for educational practice.

Educational Impact and Implications Statement This study investigated growth goal setting among a large statewide sample of high school students. Growth goal setting refers to the pursuit of specific, challenging, and competitively self-referenced targets that match or exceed a previous best effort or performance. Findings demonstrated that per- ceived instructional support (student reports of teachers’ feedback-feedforward and instructional relevance) was associated with students’ growth goal setting and that students’ growth goal setting was associated with significant gains in their academic engagement (perseverance, aspirations, and homework behavior). Importantly also, growth goal setting reduced the potential negative effects of low socioeconomic status and low prior achievement on some engagement outcomes. These find- ings provide important direction for enhancing students’ growth goal setting and engagement. Accordingly, practical suggestions for enhancing instructional support and growth goal setting are provided.

Keywords: engagement, goal setting, growth goals, instruction, motivation

Supplemental materials: https://doi.org/10.1037/edu0000682.supp

Editor’s Note. Doug Lombardi served as action editor for this article.—PK

This article was published Online First June 28, 2021. Andrew J. Martin https://orcid.org/0000-0001-5504-392X

Emma C. Burns https://orcid.org/0000-0001-6323-1816

Rebecca J. Collie https://orcid.org/0000-0001-9944-2703

Keiko C. P. Bostwick https://orcid.org/0000-0003-0631-6738

Anaïd Flesken https://orcid.org/0000-0003-4860-6493

The authors thank Nicole Hare, Brianna McCourt, and Samuel Cox for feedback on the article. This study was funded by the New South Wales Department of Education (UNSWRG193170).

Correspondence concerning this article should be addressed to Andrew

J. Martin, School of Education, University of New South Wales, NSW

2052, Australia. Email: [email protected]

752

Journal of Educational Psychology

© 2021 American Psychological Association 2022, Vol. 114, No. 4, 752–771 ISSN: 0022-0663 https://doi.org/10.1037/edu0000682

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

Goal setting plays an important role in students’ academic development (Elliot, 2005; Linnenbrink-Garcia et al., 2008; Locke & Latham, 2002, 2013; Maehr & Zusho, 2009). The present study is focused on a recently proposed construct within the goal-setting domain: growth goal setting. Growth goal set- ting refers to the pursuit of specific, challenging, and competi- tively self-referenced targets that match or exceed a previous best effort or performance; they are typically operationalized as personal best (PB) goals and self-based goals (Bostwick et al., 2020; Burns et al., 2018, 2019; Elliot et al., 2011, 2015; Martin & Liem, 2010; Martin & Elliot, 2016a, 2016b; Travers et al., 2015; Yu & Martin, 2014). Examples include setting and striv- ing to meet personal challenges, striving to outperform one’s previous best efforts or performance, and striving for self- improvement. The primary aims of the current study were to extend previous work on growth goal setting in two ways. First, we investigated the role of teachers’ instructional support (stu- dent reports of relevance, organization and clarity, feedback- feedforward) in predicting students’ growth goal setting and, in turn, the roles of both perceived instructional support and growth goal setting in predicting students’ academic engage- ment (perseverance, aspirations, school attendance, homework behavior). Second, we examined the question of whether the associ- ations between students’ background attributes and engagement are moderated by students’ growth goal setting (e.g., whether growth goal setting attenuates negative effects of low socioeconomic sta- tus). Figure 1 demonstrates the key relationships to be investigated.

Theoretical Backdrop Guiding the Hypothesized Model

The present investigation was grounded in the triadic model of social–cognitive theory (SCT; Bandura, 1986). The triadic model explains how relationships among environmental, personal, and behavioral/outcome factors are implicated in human agency. In the school context, environmental dimensions include factors such as teacher support. Personal factors comprise various self-strategies such as goal setting. Behavioral/outcome factors refer to students’ choices (e.g., aspirations) and related outcomes (e.g., attendance) that may be conceptualized under the umbrella of academic engagement (Martin, 2012a). Recent research has applied the tria- dic model to investigate the role of growth goal setting in students’ academic lives (Burns et al., 2018). This work supported the tria- dic approach, finding that positive teacher-student relationships (environmental) predicted students’ growth goal setting (personal) that in turn both positively predicted global engagement (behav- ioral/outcome). The present study seeks to expand on this work.

According to SCT, goal setting is influenced by those who are sig- nificant to the individual, such as teachers, parents, and peers (Ban- dura, 1991; see also Wentzel, 1999). Indeed, Burns et al. (2018) found that students who experience more positive social-emotional support from teachers, parents, and peers are more likely to utilize growth goal setting. Our study expands on this research by examin- ing another form of environmental support—teachers’ instructional support—and its role in predicting growth goal setting. Personal fac- tors such as goals provide the foundational beliefs and skills that

Figure 1 Hypothesized Model of Outcomes, Moderators, and Antecedents Relevant to Growth Goal Setting

T1

T1

T1

T1

Organization

and Clarity

Relevance

Feedback-

feedforward Growth Goal

Setting

Perseverance

Aspirations

Attendance

Homework

Behavior

ENVIRONMENT (Instruction) PERSON (Student Strategy) OUTCOME (Engagement)

Year Level

Gender

SES

Prior Achievement

Interactions (e.g., Growth

Goals x Gender, etc.)

Direct Effects of

Instruction on

Engagement

Note. SES = socioeconomic status; T1 = prior engagement.

GROWTH GOAL SETTING 753

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

represent an essential translational step from environmental support to more active factors such as engagement (Bandura, 1991). Impor- tantly, although Bandura (1986, 1991) is clear that these processes are ultimately reciprocal over time, he also argues that rather than examining the reciprocal relationships simultaneously, researchers can interrogate the subprocesses (via process models) longitudinally to determine the initial impact of factors and their later resultant changes. In the case of this study, such an approach enables the assessment of the initial link between environmental factors (instruc- tional support) and personal factors (growth goals) and the resultant gains in behavioral/outcome factors (engagement). As per Figure 1, our study represents one integrative analytic longitudinal framework that explores the predictive relationships from environmental factors to personal factors to behavioral factors; hence, a model comprising direct effects of perceived instructional support on growth goal set- ting and gains in engagement, direct effects of growth goal setting on gains in engagement, and indirect effects of perceived instructional support on gains in engagement via growth goal setting. These hypothesized links are described in more detail below. For growth goal setting studies that have conducted other longitudinal analytic approaches, including in terms of reciprocity, we refer the reader to Martin’s (2015) and Martin and Liem’s (2010) cross-lag panel research. We also note Bandura’s (1997) suggestion that personal factors

are likely to dynamically interact, such that background factors par- ticular to a student and their motivation may interact to affect aca- demic outcomes. In the case of our study, this would entail testing the interaction between students’ background attributes (year level at school, gender, students’ socioeconomic status [SES], and prior achievement) and their growth goal setting. Given that growth goal setting is the modifiable factor in these interactions, we position it as the focal moderator of the effects from background attributes to aca- demic outcomes. For example, previous research has demonstrated that the benefits of growth goal setting vary across sociodemo- graphic student groups (Martin, 2012b). Martin (2012b) found that growth goal setting was especially beneficial for students with atten- tion-deficit/hyperactivity disorder (ADHD) relative to classmates without ADHD. He suggested this was because focusing on them- selves as the benchmark allowed students with ADHD to strive for accessible success (which is motivating) while also diverting their attention away from aversive and demotivating comparisons with peers who typically outperformed them. In line with this, it may be the case that growth goal setting similarly attenuates well-docu- mented disparities in engagement across other student groups: year level, gender, socioeconomic status, and prior achievement. Previous research has demonstrated that boys tend to report lower behavioral engagement relative to girls (Cook et al., 2007; Van de Gaer et al., 2009) and that behavioral engagement declines as students move through high school (Burns et al., 2019). Similarly, researchers have identified disparities in engagement between students from low and high socioeconomic and low and high prior achievement, resulting from inequities in access to educational support and resources (Cut- more et al., 2018; Sirin, 2005). It may be the case, as noted by Mar- tin (2012b), that a focus on personal growth and challenge is likely to reinvigorate engagement for all students and may be particularly beneficial for students who typically experience steeper drop offs in this regard: boys, older students, low socioeconomic background, low prior achievement. Along similar lines to research finding growth mindset and goal setting can attenuate adverse academic

effects attributable to gender-related and socioeconomic-related fac- tors (Claro et al., 2016; Schippers et al., 2015), growth goal setting may attenuate the associations between personal background factors and engagement to minimize the engagement disparities that exist across these groups.

Motivational Theorizing Around Growth

We locate growth goal setting under the broader umbrella of motivational theorizing around growth. Although not intended to be exhaustive, we briefly link to four theories—achievement goal theory, goal setting theory, self-determination theory, and self- concordance theory—as indicative of the conceptual space that is relevant to key elements of the specific growth goal setting con- struct we operationalize and investigate. When considered from an integrative perspective, these theories offer some convergence on the key facets and dynamics particular to growth goal setting, including the role of self-set goals in academic engagement, the drive for personal growth and improvement (achievement goal theory), the energizing function of challenge (goal setting theory), the need for autonomy (self-determination theory), and the voli- tional aspects of goals that are integrated with one’s self and val- ues (self-concordance theory).

Achievement goal theory considers goals in terms of the reasons why individuals pursue their targets (Elliot, 2005). At the most fundamental level, individuals can pursue goals for mastery and learning (mastery goals) or goals focused on outperforming others and demonstrating relative competence (performance goals). More recently, goal theory has expanded to include self-based goals where individuals strive for personal growth and self-improvement (Elliot et al., 2011, 2015). Whereas achievement goal theory tends to address the why of motivational striving, goal setting theory tends to be more focused on the what of motivational striving (Martin, 2011), with goal setting theorists articulating the nature of goals that are most effective. One aspect of which refers to the level of challenge; high but attainable levels of challenge are cen- tral to growth goal setting (Locke & Latham, 2002, 2013; Travers et al., 2015). Indeed, it is goal setting theory that most closely con- nects to our growth goal construct, and so we review goal setting theory in more detail further below.

Self-determination theory (Ryan & Deci, 2000) is another domi- nant theory in motivational psychology and represents important conceptual space for informing and considering growth goal set- ting. Self-determination theory identifies that individuals have a need for autonomy, competence, and relatedness; when these needs are met, individuals are more likely to adaptively develop and function. As noted by Collie et al. (2016), a core feature of growth goals is that they are determined by students, about them- selves, and for themselves, indicating that they may help satisfy students’ need for autonomy. This is further supported by recent theorizing seeking to integrate achievement goal theory and self- determination theory that has suggested that the self-based (growth) goals of achievement goal theory (Elliot et al., 2011) are quite aligned with the need for autonomy in self-determination theory (Vansteenkiste et al., 2014). The self-concordance model (Sheldon & Elliot, 1999) also attends to the self-determined nature of individuals’ goals and their growth elements. This model pro- poses that the consistency between goals and the individual’s interests and core values has significant implications for goal

754 MARTIN ET AL.

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

striving, goal attainment, and well being outcomes. Self-concord- ant goals are those that are integrated with the self. This has posi- tive effects on well being via the intrinsic and self-determined nature of self-concordant goals and the internal locus of control that these goals entail (Sheldon & Elliot, 1999). In contrast, exter- nally set or referenced goals tend not to align with individuals’ enduring interests and values and thus lack the volitional strength that self-concordant goals possess (Sheldon & Elliot, 1999). A growth goal is a goal target set by the student, as is the means by which they strive toward that target. To the extent that this is the case, there may be adaptive self-concordance in growth goal setting.

Growth Goal Setting: Key Elements and Processes

As noted, we suggest goal setting theory is very closely con- nected to our growth goal setting construct and we therefore give this perspective relatively more attention in this study. According to Locke and Latham (2002, 2013), the three key elements of effective goals are difficulty, specificity, and reference. Difficulty refers to the degree to which the target represents an optimal chal- lenge. Thus, for growth goal setting, difficulty is where the target matches or is higher than a previous best (Martin, 2006). Specific- ity is when the target is clearly articulated and well-defined (Locke & Latham, 2002). With growth goal setting, the effort or perform- ance needed to outperform a previous best is specifically known (as opposed to trying to beat others, whose performance is often unknown until the task is finished; Martin, 2006). Reference refers to whom or what one is comparing when setting and striving to- ward a goal. Researchers have identified personal, rather than external references to be the most adaptive for students (Hulleman et al., 2010). In the case of growth goal setting, the reference is oneself and one’s previous effort or performance (Martin & Liem, 2010). Theories of goal setting (e.g., Locke & Latham, 2002, 2013)

have shed light on how goals impact academic outcomes. Martin and Elliot (2016b) drew on this conceptualizing to explain how growth goal setting may function to enhance outcomes. First, growth goals make the target clear to students, so they know what to strive for to outdo a previous best. Second, growth goals direct students to goal-relevant tasks that help keep them focused and on-track. Third, growth goals are self-concordant (self-set and in line with one’s own benchmark and values; Koestner et al., 2002) and tend to evoke volitional resources that are energizing, espe- cially in the face of task-irrelevant temptations. Fourth, growth goals create a dissonance between current and desired states—a dissonance that students are motivated to resolve (Martin, 2011; Martin & Elliot, 2016b). In fact, this latter point is consistent with SCT theorizing that self-improvement occurs through a cycle of personal discrepancy production and reduction (Bandura, 1986). Burns et al. (2018) argued that taking these functions together,

growth goal setting provides a motivational foundation for pro- moting positive behaviors such as engagement (see also Martin, 2006; Martin & Elliot, 2016b; Martin & Liem, 2010). They described how growth goals can act as mental scaffolds that indi- viduals use to identify the task-relevant behaviors that are neces- sary for pursuing and attaining one’s goals (see also DeShon & Gillespie, 2005). To this extent, growth goal setting can be

considered an important translational step to active behaviors such as engagement (Burns et al., 2018; see also Dweck, 2017).

Perceived Instructional Support, Growth Goal Setting, and Engagement

Instructional support is a salient antecedent of goal setting under SCT’s triadic model (Bandura, 1986; Burns et al., 2018). It is also well established that instructional factors have a major impact on students’ outcomes, including engagement (Hattie, 2009; Hulle- man et al., 2017; Martin et al., 2020). We investigated students’ perceptions of three instructional support factors—instructional relevance, instructional organization and clarity, and feedback- feedforward.

Relevance, Organization and Clarity, and Feedback- Feedforward

In the context of our study, relevance refers to the personal alignment with and meaningfulness of content and tasks. This fol- lows previous research and theorizing identifying the importance of content and tasks that are accessible to the learner and that align with prior experience and knowledge (Martin, 2016; Martin & Evans, 2018, 2019; Sweller, 2012; Van de Pol et al., 2010), as well as instruction characterized by provision of meaningful and relevant content and tasks (Lei et al., 2017; Ryan & Deci, 2000). According to Assor et al. (2002), clarifying and enhancing the relevance of schoolwork helps students to understand the contribu- tion made by the lesson in realizing their personal goals, interests, and values. Research also suggests that when teachers emphasize the personal relevance of subject matter, students are likely to engage (i.e., persist and participate) because of more positive feel- ings toward their schoolwork (Assor et al., 2002). Indeed, in a relevance intervention experiment (encouraging students to make connections between learning in their science courses and their own lives), Hulleman and Harackiewicz (2009) found that increased personal relevance boosted academic outcomes in sci- ence. These results were replicated in a subsequent study, which also found that the frequency with which students made connec- tions between academic subject matter and their personal lives enhanced their engagement and confidence (Hulleman et al., 2017). Hulleman and Harackiewicz (2009) argued that facilitating the perception of personal relevance in a topic energizes students as they discover reasons for becoming more involved in learning (i.e., engage), and helps students identify with future academic pathways and careers in a particular subject area.

Organization and clarity in our study refer to the teacher’s orga- nization and clarity of the content and tasks, as well as manage- ment of lesson time to optimize learning. In practice, these typically comprise scaffolding and structured practice during learning (Martin, 2016; Martin & Evans, 2018; Mayer & Moreno, 2010; Sweller, 2012), as well as organization of content, time, and tasks that are appropriate to the learner’s level of knowledge and skill, along with appropriate opportunities for guided self-directed activity (Grolnick, 2003; Reeve, 2006; Van de Pol et al., 2010; Vansteenkiste et al., 2012). Particularly when students are novices, instructional organization and clarity enable them to better pro- cess, store, and develop requisite knowledge and skill (Mayer & Moreno, 2010). Meta-analysis has demonstrated that instructional

GROWTH GOAL SETTING 755

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

organization and clarity are significantly associated with enhanced student learning and affective engagement (Titsworth et al., 2015). Feedback-feedforward refers to corrective information and

improvement-oriented guidance to students during learning (fol- lowing calls to combine classic feedback [corrective information] with improvement-oriented guidance [feedforward]; Basso & Oli- vetti Belardinelli, 2006; Burns et al., 2019; Hattie & Timperley, 2007; Martin, 2016). Researchers have identified its key elements in terms of clear corrective feedback, explicit expectations, and improvement-oriented guidance (Grolnick, 2003; Martin, 2016; Reeve, 2006; Van de Pol et al., 2010; Vansteenkiste et al., 2012). Feedback-feedforward provides information to students that can be a basis for adapting their learning via improved self-regulation, better strategies for working on tasks, enhanced understanding, and specific and explicit information about what more needs to be learned or understood (Hattie & Timperley, 2007). Indeed, the most recent meta-analysis on this topic found a significant associa- tion between feedback-feedforward and student learning (Wis- niewski et al., 2019). These three dimensions of instructional support have been well

articulated in numerous theories of instruction that have also vari- ously articulated links with students’ goal setting and academic engagement (for example, self-determination theory and constitu- ent needs-supportive teaching, Reeve, 2006; Ryan & Deci, 2000; expectancy value theory, Wigfield & Eccles, 2000; load reduction instruction, Martin & Evans, 2018, 2019; CLASS framework, Hamre et al., 2007). An expansive summary is beyond the scope of the present study. However, as a point of illustration we present one conceptual framework that we suggest has direct relevance to our three instructional support factors—and to students’ growth goal setting and academic engagement: the Hidi and Renninger (2006) four-phase model of interest development.

Interest Development, Instructional Support, Growth Goal Setting, and Engagement

The Hidi and Renninger (2006) four-phase model emphasizes the importance of relevance, organization and clarity, and feed- back-feedforward for interest development. The first phase involves sparking situational interest through relevant and person- alized content and tasks. In phase two of their model, they suggest that situational interest is sustained through ongoing instructional attention to relevance. This second phase is also substantially externally supported, such as through instructional organization and clarity and individualized support (i.e., feedback-feedforward) if needed. The final phases of their model involve more self- directed activity by the student that relies on feedback-feedforward from the teacher (see also Lipstein & Renninger, 2006). Each of these phases highlights the positive benefits of instruc-

tional support for students’ interest development—but also for other adaptive educational outcomes. For example, Hidi and Renninger (2006) suggest that these instructional support approaches should also benefit students’ goal setting and academic engagement. Spe- cifically, the higher levels of interest developed through these instructional supports are theorized to comprise self-generated and self-selected pursuits, as well as iterative increases in self-set chal- lenges. From this we might infer that high levels of interest include growth goal setting because such goals are self-generated and self- selected and involve an iterative escalation of challenge (Martin,

2006). The various phases of interest development have also been associated with students’ academic engagement (Renninger & Hidi, 2015). According to Renninger and Hidi (2020), engagement at each phase of interest development is triggered by a process that propels information searches that deepens knowledge and perceived relevance. Thus, instructional approaches that support students’ in- terest should also support students’ growth goals and engagement.

Summary of Perceived Instructional Support

Our primary thesis is that major forms of instructional support (instructional relevance, organization and clarity, and feedback- feedforward, as reported by students) are positively associated with growth goal setting, which in turn is positively associated with engagement. This being the case, and also following the tria- dic model, it is plausible to examine the indirect association between instructional support and engagement via growth goal set- ting. Indeed, research has shown that these forms of instructional support are directly associated with student goal setting (e.g., Burns et al., 2019; Eberley et al., 2011) and engagement (e.g., Col- lie et al., 2019; Martin, et al., 2020; Reeve et al., 2004; Renninger & Hidi, 2015, 2020). We therefore explore the roles of instruc- tional relevance, organization and clarity, and feedback-feedfor- ward as (a) direct predictors of students’ growth goal setting, (b) direct predictors of student engagement, and (c) indirect predic- tors, via growth goal setting, of engagement. Figure 1 demon- strates the hypothesized links.

Growth Goal Setting and Academic Engagement

Academic engagement is not only a means to desirable ends (e.g., achievement), it is also recognized as a desirable end in itself. Thus, for example, research has not only demonstrated links between academic engagement and outcomes such as achieve- ment, school completion, postschool pathways, and so forth (Froi- land & Worrell, 2016; Reyes et al., 2012; Van Ryzin et al., 2009); research has also shown that engagement is valued by students, teachers, and parents/carers as an important outcome in its own right (Bempechat & Shernoff, 2012; Martin & Bolliger, 2018). Accordingly, academic engagement is identified as the “outcome” factor in our investigation. Indeed, although achievement was out- side the scope of this study, we do note prior research showing that growth goals are directly associated with achievement out- comes (e.g., Martin & Elliot, 2016b; Martin & Liem, 2010) and also indirectly associated with achievement via academic engage- ment (e.g., Martin et al., 2016); thus, our study of engagement as the focal outcome of instructional support and growth goal setting may be considered an important first step for future research inves- tigating achievement.

Consistent with the triadic model that has well-established cog- nitive–behavioral foundations (Bandura, 1986), we adopted a cog- nitive–behavioral approach to engagement. Cognitive engagement is defined as students’ mental investment and striving in their learning (Fredricks et al., 2004). Our cognitive engagement meas- ures comprise perseverance and aspirations. Perseverance is a form of investment and striving in line with the Fredricks et al. (2004) framework and is deemed a critical attribute for navigating academic challenge as well as for completing large, extended, and/ or multipart academic tasks (Martin, 2007, 2009). Academic aspi- rations are associated with continued academic investment and

756 MARTIN ET AL.

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

school completion (Burns, 2020; Burns et al., 2021; Gutman & Schoon, 2018). Behavioral engagement refers to students’ actions and active

participation in academic activities (Fredricks et al., 2004). Our behavioral engagement measures comprised homework behavior and school attendance. Doing homework is important for develop- ing important academic self-regulatory skills (e.g., managing time, strategy development, delaying gratification, etc.; Ramdass & Zimmerman, 2011) and is also associated with enhanced learning, particularly among high school students (Fan et al., 2017). With regard to school attendance, it has been found that absentees receive fewer hours of instruction and are slower to develop core skills that underpin their learning (Willms, 2003); thus, school attendance is a vital behavioral engagement outcome. Taken to- gether, our study sought to examine the extent to which growth goal setting may provide a motivational foundation for enhancing homework completion and attendance (behavioral engagement), as well as perseverance and aspirations (cognitive engagement) (Burns et al., 2018). Research into growth goal setting supports our inclusion of per-

severance, aspirations, homework behavior, and attendance as aca- demic engagement outcomes. In a study of growth goal setting, Martin and Liem (2010) found significant associations with per- sistence, but their work did not control for background attributes known to be associated with persistence and growth goals (e.g., gender, age, prior achievement; Burns et al., 2017, 2018; Martin, 2007, 2009); our study does. For aspirations, Burns et al. (2018) found that growth goal setting was associated with positive aca- demic aspirations. However, their measure of academic aspirations was part of a higher-order engagement factor, whereas our study modeled aspirations as a distinct factor to examine its unique asso- ciation with growth goals. Similarly, Bardach et al. (2020) found that growth goals were associated with lower dropout intentions, but their study was of university students, not high school stu- dents. In relation to homework, Martin (2012b) found a significant positive association with growth goals. However, his measure of homework was a single-item indicator only capturing completion; our study included a multi item measure capturing homework more comprehensively (enjoyment, learning, completion). Finally, although Burns et al. (2018) found that growth goal setting was associated with class participation, this is different from other be- havioral engagement factors such as attendance (which our study includes), and that have been identified as critical to narrowing achievement and school completion gaps between students and particular student groups (e.g., low-SES students, etc.; Lamb et al., 2015).

Accounting for Personal Background Attributes and Moderation

The triadic model (Bandura, 1986) emphasizes the importance of accounting for a range of personal factors, beyond motivation, that may impact students’ agency, such as their background attrib- utes. Moreover, Bandura (1997) argued that these various personal factors are likely to dynamically interact, such that background attributes particular to a student and their motivation may interact to affect academic outcomes—for example, it may be that growth goal setting impacts engagement differently for low vs high SES students. Thus, it is critical to account for these background factors

in investigations of growth goal setting and engagement. These background attributes include: students’ year level at school (younger students are typically more engaged and more likely to pursue growth goals: Anderman & Maehr, 1994; Martin et al., 2015), gender (girls are typically higher in engagement: Burns et al., 2018; Burns et al., 2019; Martin, 2004), SES (higher SES stu- dents are typically more engaged: Burns et al., 2018), and prior levels of achievement (higher prior achievement is typically asso- ciated with engagement and growth goal setting: Burns et al., 2017, 2018; Collie et al., 2016; Liem et al., 2012; see Figure 1).

The inclusion of these factors allowed us to control for their known influence on engagement; this helped us to identify unique effects of growth goal setting on engagement purged of variance attributable to these personal background attributes. Moreover, their inclusion also allowed us to investigate the extent to which growth goal setting may moderate the effects of these background attributes on engagement (e.g., whether growth goal setting may have particular benefit for low SES students; discussed above).

Aims of the Present Study

A growing body of work has identified the value of growth goal setting in fostering positive academic outcomes. However, much of this research has tended to focus on global engagement (not multidimensional engagement) and has not considered the extent to which growth goal setting may moderate the effects of students’ background on their engagement. Furthermore, minimal research has considered the role of perceived instructional support in pro- moting growth goal setting, nor has it examined the impact of instructional support and growth goal setting on engagement. The present study addressed these gaps by investigating the role of teachers’ instructional support in predicting students’ growth goal setting and, in turn, the roles of both instructional support and growth goal setting in predicting students’ academic engagement outcomes (thus, direct and indirect associations). We also exam- ined the question of whether the effects of students’ background attributes on engagement are moderated by their growth goal set- ting. Figure 1 shows the model we tested.

We hypothesized that instructional support (perceived rele- vance, organization and clarity, feedback-feedforward) would pos- itively predict growth goal setting, growth goal setting would positively predict academic engagement (perseverance, aspira- tions, school attendance, homework behavior), and the direct and indirect (via growth goal setting) associations between instruc- tional support and engagement would also be positive (Burns et al., 2019; Collie et al., 2019; Martin et al., 2020; Reeve et al., 2004; Renninger & Hidi, 2015; 2020). Prior research has sug- gested that growth goal setting may moderate the effects of student background factors on engagement (e.g., Martin, 2012b). To the extent that this may apply to other background attributes, we hypothesized that growth goal setting may attenuate the effects of year level, gender, socioeconomic status, and prior achievement to minimize the engagement disparities across these groups. However, it should be noted that there is not enough research to warrant spe- cific hypotheses about how growth goal setting will moderate rela- tionships for each specific engagement construct (perseverance, aspirations, school attendance, homework behavior). Thus, we posi- tioned moderation effects as open empirical questions.

GROWTH GOAL SETTING 757

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

Method

Participants, Sampling, and Procedure

The present investigation examined a sample of N = 61,879 high school students from 290 government schools across New South Wales (NSW), Australia’s most populous state. The data were collected as part of the NSW Department of Education’s an- nual “Tell Them from Me”1 (TTFM) student survey, which has been conducted since 2013. The survey was originally developed in Canada (Willms, 2014) but has been adapted to the Australian context. Evidence of validity for the scale scores has been demon- strated, and the scale is used in several different countries, includ- ing Australia and the United States (The Learning Bar, 2019). The survey captured information about students’ motivation, engage- ment, and wellbeing in school, as well as their perceptions of teachers’ instructional support. The NSW Center for Education Statistics and Evaluation

(CESE), located within the NSW Department of Education, was responsible for the recruitment of schools and data management. The survey itself was conducted by the survey company The Learning Bar. The survey was offered free of charge to all NSW government schools to capture student voice and to provide schools with data to drive student and school improvement. Schools could request participation via an online expression of in- terest portal. The online portal for requesting participation was open from Term 4 in the previous year up until the start of the sur- vey in Term 1 of the survey year—thus, for example, from Term 4 of 2017 until Term 1 of 2018. The survey ran on an opt-out con- sent basis, with schools having to ensure that parents were given the opportunity to decline consent for their child’s participation. To facilitate this process, CESE provided schools with opt-out consent forms and communication templates in 22 community lan- guages. The present study received ethical approval from the first author’s institutional review board. In 2018 and 2019, the survey years used in this study, 65.5% of high schools participated in the survey, with an average student response rate of 70.8%. For the present study, data were collected from students in

Term 1 of 2018 and again in Term 1 of 2019 (i.e., one academic year apart). The same students were sampled at both time points (some students moved schools during Term 1 and could not be longitudinally matched to a school, some students had invalid link- ing data, some students had repeated a year; these students were excluded from the sample and subsequent analysis). The students sampled were in grades 7 (30.7%), 8 (25.7%), 9 (24.5%), and 10 (19.1%) in 2018 and grades 8–11 in 2019. Approximately half the sample was female (50.2%). SES was assessed using an index of students’ social and economic resources at home (comprising measures of parent education, parent occupation, and educational resources; statewide M = 0, SD = 1; scores, 0 reflect below aver- age state-wide SES). The average SES score for the sample was M = .07 (SD = 1.00), indicating that the students included in the analysis were around the state-wide SES average and signaling our sample’s representativeness for SES. Prior achievement was assessed via students’ scores on the National Assessment Program —Literacy and Numeracy (NAPLAN). NAPLAN is a national standardized assessment aimed at ascertaining the extent to which students are meeting national literacy and numeracy benchmarks.

NAPLAN generates raw scores (scaled from 100 to 800) and band scores (scaled from 1 to 10) for students’ literacy and numeracy performance. The average raw score achievement attained by this study’s students was almost identical to that attained in the state as a whole (within 6.5 units from each other), further signaling the representativeness of our sample in terms of prior achievement. The majority of schools were located in major urban centers of NSW (78% of schools in the sample; the NSW average is 66%), with the remainder in regional and remote areas of NSW. Schools were coeducational (85% of schools in the sample; the NSW aver- age is 89%), single-sex boys’ school (6%; the NSW average is 5%), or single-sex girls’ schools (9%; the NSW average is 6%).

Materials

The measures in this study comprised growth goal setting, per- ceived instructional support, academic engagement, and student background attributes. Analyses utilized Term 1 2019 data on all measures, as well as prior data (Term 1, 2018) on the engagement (outcome) measures. Descriptive statistics and reliability measures for the substantive factors are reported in Table 1.

Growth Goal Setting

Growth goal setting was assessed via four items from the origi- nal Multidimensional Personal Best Goal Scale (Martin, 2006; referred to as the “Growth Orientation Scale” in the TTFM publi- cations). This was a slightly broader measure than the later (final) Personal Best Scale by Martin and Liem (2010). Our study’s mea- sure comprised two items from the four-item self-improvement subscale and two items from the four-item challenge subscale of the original Multidimensional Personal Best Scale (Martin, 2006). These items are purposefully worded so that the self is the referent (i.e., the student strives to outperform or improve on themselves). All four growth goal setting items were domain general (i.e., rated by students across all subjects, not in a specific subject) and are as follows: “I set challenges for myself in my schoolwork”; “I like to work toward challenging goals in my schoolwork”; “When I do my schoolwork, I try to do the best that I've ever done”; “When I do my schoolwork, I try to improve on how I've done before.” They were assessed on a scale of 0 (strongly disagree) to 4 (strongly agree). The growth goal setting factor demonstrated reli- ability (x2019 = .92).

We point out that this particular goal factor is different from performance goals (under achievement goal theory; Elliot, 2005) where others are the target referent (i.e., the student strives to out- perform others or demonstrate relative ability). These growth goals are also distinct from mastery goals where the task/learning is the central referent (i.e., the student strives to master or learn the task). Indeed, Yu and Martin (2014) found that self-based growth goals (via personal best goals) and mastery goals both uniquely predicted academic outcomes in a large sample of Chinese stu- dents (with growth goals more strongly predictive of engagement than mastery goals), concluding: “mastery and PB goals explain unique variance in distinct academic outcomes such that mastery goals appear more salient in mapping onto motivation factors

1 “Tell Them From Me” is a registered trademark belonging to The

Learning Bar.

758 MARTIN ET AL.

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

while PB goals appear more salient in mapping onto engagement” (p. 635). Subsequently, Martin and Elliot (2016a) showed that self-based growth goals (assessed via personal best goals), mastery goals, and performance goals each uniquely predicted academic outcomes. With special reference to mastery goals, they concluded “there is a potentially positive and complementary (not opposing or mutually exclusive) role for both personal best and mastery goals and that together, mastery and personal best goals capture a greater totality of variance in students’ motivation and engage- ment” (p. 1296). Most recently, Bostwick et al. (2020) demon- strated that self-based goals such as the growth goals we administer in this study are a distinct and separable factor from task-based goals (such as mastery goals).

Perceived Instructional Support

Perceived instructional support was assessed via three separate factors: instructional relevance, instructional organization and clarity, and feedback-feedforward. To reduce respondent burden, students were randomly assigned to one domain (i.e., mathematics, science, or English) for all measures of perceived instructional support (thus, e.g., one student was randomly assigned to report on only their perceived instructional support from their science teacher, whereas another student was randomly assigned to report on only their mathematics teacher). Critically, the items were worded the same across subject domains (i.e., parallel items). Stu- dents’ responses to these items, regardless of their subject domain, were used as indicators of the factor in modeling (viz., relevance item #1, relevance item #2, etc. as indicators of a relevance factor). For example, if Student A was assigned the mathematics instruc- tion items, Student B was assigned the science instruction items, and Student C was assigned the English instruction items, then A’s relevance item #1 in mathematics, B’s parallel relevance item #1 in science, and C’s parallel relevance item #1 in English would all be used as the item 1 indicator for relevance (see Table S7 in the online supplemental materials for visual example). This approach was considered viable given previous research demon- strating the positive associations between students’ domain-gen- eral and domain-specific perceptions (Bong, 2001; Green et al., 2007) and given previous suggestions that there may be trait-like motivations that traverse school subjects even in the context of do- main-specificity (e.g., Trautwein, Lüdtke, Kastens, et al., 2006; Trautwein, Lüdtke, Schnyder, et al., 2006). It is also important to

note that because students in high school typically have different teachers for different school subjects, asking students to report on one particular teacher enabled students to focus more specifically on their experiences in one class, rather than their overall experien- ces in many different classrooms that might vary widely. This not- withstanding, domain-specific models were run as well (e.g., where only mathematics instructional support items were included) and derived the same substantive findings (reported in the online supplemental materials).

Relevance was assessed via the TTFM Teaching Relevance Scale (three items with parallel wording across domains). This included items about the meaningfulness, usefulness, and purpose- fulness of the teaching and content (e.g., “[In the past two weeks] We explored ideas and topics that are meaningful”). All items were assessed on a scale of 0 (strongly disagree) to 4 (strongly agree). The three items were used as latent indicators of the rele- vance factor, which demonstrated reliability (x2019 = .84).

Organization and clarity was assessed via the TTFM Effective Learning Time Scale (six items with parallel wording across domains; e.g., “Our [math/science/English] teacher is good at explaining difficult ideas”). This included items about manage- ment of lesson time to optimize learning and organization and clarity of content and tasks. All items were assessed on a scale of 0 (strongly disagree) to 4 (strongly agree). The six items were used as latent indicators of the organization and clarity factor, which demonstrated reliability (x2019 = .94).

Feedback-feedforward was assessed via the TTFM Explicit Teach- ing Practice and Feedback Scale (six items with parallel wordings across domains; e.g., “The feedback from assessments and quizzes helps me learn”). To represent the breadth of the feedback-feedfor- ward construct, these items captured reciprocity in information from and to students, as well as teacher explanations, corrective informa- tion, and improvement-oriented information (thus connecting to explicit instructional elements; Martin, 2016). Four items were rated from 0 = never or hardly ever to 3 = in all lessons; and two items were rated from 0 = strongly disagree to 4 = strongly agree). These six items were used as latent indicators of the feedback-feedforward factor, which demonstrated reliability (x2019 = .88).

Academic Engagement

Academic engagement included perseverance and aspirations (as cognitive engagement indicators) and attendance and homework

Table 1 Descriptive and Factor Analytic Statistics

Variable M SD Mean factor loading (range)

Organization and clarity 2.62 0.83 .85 (.76–.91) Feedback-feedforward 2.11 0.61 .74 (.72–.79) Relevance 2.27 0.84 .80 (.73–.86) Growth goal setting 2.55 0.88 .87 (.85–.88)

2018 2019 2018 2019 2018 2019

Perseverance 2.50 2.40 1.00 1.05 .80 (.75–.84) .82 (.76–.85) Aspirations 3.27 3.21 1.15 1.26 1.00 (1.00) 1.00 (1.00) Attendance 2.93 2.90 0.11 0.16 .75 (.74–.75) .79 (.79–.79) Homework behavior 1.63 1.49 0.59 0.61 .71 (.61–.77) .72 (.62–.81)

Note. We have adjusted the instructional support factor names from the original terminology used in their source survey—see the Method section for details. Aspirations is a single-item indicator where the loading is fixed to unit value and the residual is fixed to zero.

GROWTH GOAL SETTING 759

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

behavior (as behavioral engagement indicators). All engagement fac- tors were domain general and assessed once in Term 1 2019 (time 2 [T2]—with all the other measures in the study) and once in Term 1 2018 (time 1 [T1]). Perseverance was assessed via four items (e.g., “I finish whatever I begin”), rated on a scale of 0 (strongly agree) to 4 (strongly disagree), and demonstrated reliability at T1 (x2018 = .88) and T2 (x2019 = .89). Aspirations was assessed with a single item (“I plan to finish year 12”), rated on a scale of 0 (strongly disagree) to 4 (strongly agree). Attendance was assessed via two items (e.g., “[In the past four weeks] I have missed a day at school without permis- sion”) which were measured on a scale from 0 (never) to 3 (almost every day); these items were then reverse coded (viz., 0!3; 1!2; 2!1; 3!0). The attendance factor demonstrated reliability at T1 (x2018 = .72) and T2 (x2019 = .77). Homework behavior was assessed via three items (e.g., “I enjoy doing my homework and studying”) and measured on a scale of 0 (always) to 3 (never) and then reverse coded. The homework behavior factor demonstrated reliability at T1 (x2018 = .76) and T2 (x2019 = .77).

Student Personal Background Factors

The student personal background factors included in the present investigation were year level, gender, socioeconomic status, and prior achievement. Year level was assessed as a continuous vari- able (year 7 to year 10). Gender was assessed as a dichotomous variable (0 = male; 1 = female). Socioeconomic status was assessed as a continuous measure of an index of students’ social and economic resources (described above). Prior achievement was assessed as a continuous measure via students’ band scores (scaled from 1–10) on NAPLAN (described above). In this study, the liter- acy and numeracy band scores were standardized by grade, and these standardized scores were then used to generate a mean prior achievement score (x = .93).

Data Analysis

Data analysis proceeded through four stages: confirmatory fac- tor analysis (CFA), measurement invariance testing, structural equation modeling (SEM), and indirect effects testing. All analy- ses were conducted in Mplus v8.40 (Muthén & Muthén, 2017). Maximum likelihood with robustness to non-normality (MLR) was included as the estimator and full-information maximum like- lihood (FIML) was included to handle missing data. The Complex and Cluster commands in Mplus were used to cluster students by school to account for the hierarchical nature of the data. Namely, standard errors were adjusted for the nesting of students within schools. Student weights (based on 2018 data) were also included to avoid inflated standard errors due to larger cluster sizes. Importantly, to guard against giving undue weight to small

effect sizes that are nonetheless statistically significant (given the large sample size), we used Keith’s (2006) guidelines for educa- tional research to help determine whether a finding was interpreta- ble. According to Keith (2006), effect sizes of .05 # b , .10 are considered small, .10 # b , .25 are considered moderate, and b $ .25 are considered large in educational research. For our study, only the findings that are significant at p , .001 and effect sizes of at least b$ .05 are interpreted and subsequently discussed in this article.

Preliminary Analysis of Multilevel Properties and Measurement Invariance

Goal-setting theory and social–cognitive theory position goal setting as an intrapsychic individualistic phenomenon that is influ- enced by personal perceptions—thus, inherently a student-level process. This has been supported in empirical work showing that students’ own motivation and engagement are significantly con- nected, but climate perceptions are unrelated to students’ own motivation and engagement (Ruzek & Schenke, 2019); and also work showing substantial within-group student heterogeneity (Schenke et al., 2017) and that most variance in motivation and engagement appears to be at the student-(not school-) level (Marsh et al., 2008; Martin et al., 2011). However, for completeness the viability of multilevel modeling (i.e., both student- and school- level analysis to account for the school context) was assessed given the hierarchical nature of the data. Variance at the school- level was determined via the intraclass correlation estimates (ICC) and the estimates of the variances of slopes (r2) between growth goal setting and the engagement factors at the school-level. The recommended level of between-level variance is ICC $ 10% (Byrne, 2012). All but two variables demonstrated # 3% variance at the school-level (aspirations and homework behavior each dem- onstrated 8% variance at the school-level). Analysis of slope var- iance (r2) was estimated for the following parameters to assess the potential for school-level variance in associations: (a) growth goal setting ! perseverance, (b) growth goal setting ! aspira- tions, (c) growth goal setting ! attendance, and (d) growth goal setting ! homework behavior. All r2s , 1%, indicating a lack of variance at the school level. Taken together, there is little support for multilevel modeling for the hypothesized model and so we pro- ceeded with single-level modeling—that is, including school-level effects (student-level data aggregated to school-level to account substantively for context, rather than just statistically for cluster- ing) was not warranted. In addition, none of the items had “school” as the referent and so modeling at the school-level was not conceptually appropriate. Moreover, the nature of sampling was such that students were not clustered within classrooms, meaning class-level analyses were also not possible. The results of all these preliminary multilevel analyses are reported in Table S1 in the online supplemental materials.

In addition to testing the viability of multilevel analysis, tests of multigroup and longitudinal measurement invariance were con- ducted. Multigroup invariance tests were run to ensure that all of the substantive measures (i.e., antecedents, mediators, and out- comes) included in this investigation were invariant across key subgroups: year level, gender, socioeconomic status, and prior achievement (the latter two via mean split). This was done through a series of six gradually constrained single-level, multigroup CFAs (Byrne, 2012; Meredith, 1993). The configural (baseline) model allowed all parameters to be freely estimated across groups. The subsequent models constrained factor loadings (factor loading invariance); factor loadings and intercepts (intercept invariance); factor loadings, intercepts, and residuals (factor variance invari- ance); factor loadings, intercepts, and covariances (factor covari- ance invariance); and factor loadings, intercepts, covariances, and residuals (full factor invariance; Byrne, 2012; Vandenberg & Lance, 2000). Intercept invariance is considered the minimum cri- terion by which to conclude group invariance and to examine a

760 MARTIN ET AL.

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

single sample (Van de Schoot et al., 2012). It should be noted that because no pairs of items were flagged as issues in the baseline models, it was not necessary to constrain any error covariances (Byrne, 2012). Tests of longitudinal measurement invariance were conducted

to ensure that factors were measured consistently across time. Because only the engagement factors were measured across time, only these factors were included in these tests. Importantly, aspira- tions were not included in these tests because aspirations were measured via a single-item indicator; thus, tests of longitudinal invariance focused on perseverance, attendance, and homework behavior. Longitudinal invariance was conducted via a series of four gradually constrained single-level CFAs. The configural model (Table S2b in the online supplemental materials) included the following constraints: factor loading of the first item of each variable was set to 1, intercept of the first item of each variable was set to 0. Factor means, variances, and covariances were freely estimated. To demonstrate factor loading invariance, the factor loadings of like items were constrained to be the same across time. To demonstrate intercept invariance, the intercepts of like items were constrained to be the same across time. To demonstrate fac- tor variance invariance, the residuals of like items were con- strained to be the same across time (Sterba, 2017; Vandenberg & Lance, 2000; Widaman et al., 2010). Changes in the comparative fit index (CFI) and the root mean

square error of approximation (RMSEA) were monitored to determine invariance for both multigroup and longitudinal tests; DCFI # .010 and DRMSEA # .015 were considered the cut-offs for invariance (Chen, 2007; Cheung & Rensvold, 2002; Vanden- berg & Lance, 2000). For completeness, the chi-square statistics are also reported; results of chi-square difference tests were not reported because difference testing would need to be conducted with the loglikelihood estimates and scaling correction factors, rather than the chi-square values, because the present investiga- tion used MLR as the estimator (Muthén & Muthén, 2017). These tests have demonstrated oversensitivity and bias within large sample sizes (Widaman et al., 2010), such as the sample size in the present investigation. For multigroup measurement invariance, findings indicated that

all factors attained the minimum criterion of intercept invariance. More fine-grained follow-up analyses revealed factor variance invariance for year level and SES; intercept invariance was achieved for prior achievement; and full factor invariance was achieved for gender. For longitudinal invariance, strict invariance was demonstrated. Taken together, the measurement properties for the factors in this study can be concluded as invariant for all key subgroups and across time. The fit statistics of all models tested are presented in Tables S2a and S2b in the online supplemental materials.

Confirmatory Factor Analysis and Structural Equation Modeling

CFA was used to test the bivariate latent correlations among the substantive factors and student background factors and SEM was used to test the predictive parameters of the hypothesized model in Figure 1 (Kline, 2016). Both the CFA and SEM included all sub- stantive and student background factors within a single analytic model. CFI and RMSEA were the indices used to assess model fit.

Adequate model fit is indicated by CFI values greater than .90 and RMSEA values lower than .08. Excellent model fit is indicated by CFI values greater than .95 and RMSEA values lower than .05 (Hu & Bentler, 1999).

For the SEM, longitudinal (auto-regressive) paths were included for the engagement outcome variables (e.g., T1 2018 perseverance ! T2 2019 perseverance). By including these autoregressive pa- rameters, it is possible to assess the extent to which, for example, growth goal setting predicted unique variance in engagement above and beyond prior variance in engagement. Because of this, significant unique effects of the antecedent variables can be inter- preted as gains (or declines) in engagement over time. We also point out that for this analytical design there are advantages in modeling students’ perceptions of instructional support, growth goals, and engagement in the same year (while controlling for prior variance in engagement). If we had used prior instructional support (in 2018) to predict 2019 goals, we would potentially introduce undesirable bias because we would be using ratings about previous teachers. We suggest that our analytical design brings rigor with the longitudinal modeling of outcomes but retains important contemporaneous contextual and instructional information.

The role of growth goal setting in moderating the relationship between student background factors and engagement was also examined. A total of 16 interactions were examined. Latent inter- actions were tested in the SEM via the XWITH command with Type = Random in Mplus. It is important to note that standardized b-values were used to identify the presence of interaction effects and unstandardized b-values were used to plot the precise nature of the interactions.

Indirect Effects

Indirect effects testing was conducted to examine the extent to which growth goal setting mediated the link between the perceived instructional support factors and the engagement factors (e.g., instructional relevance ! growth goal setting ! homework behavior). A total of 12 indirect paths could potentially be assessed. To meet the criteria for testing, the relationship between the antecedent and mediator needs to be significant, as well as between the mediator and outcome. These indirect effects were tested using nonparametric bootstrapping (1,000 draws; Shrout & Bolger, 2002).

Results

Confirmatory Factor Analysis

The CFA demonstrated adequate fit: v2(737) = 72,998.74, p , .001; CFI = .938; RMSEA = .040. All factor loadings are pre- sented in Table 1, and bivariate latent correlations for the hypothe- sized associations presented in Figure 1 are presented in Table 2 (see Tables S8 in the online supplemental materials for a full cor- relation matrix). Given the number of correlations that were exam- ined, the correlations presented here in-text are those significant (p , .001) and specific to the study’s substantive concerns (viz., growth goal setting, perceived instructional support, and engage- ment). It is important to note that all auto-regressive correlations for the engagement outcome factors were significant at p, .001.

GROWTH GOAL SETTING 761

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

Growth goal setting was correlated with all outcomes (at T2): perseverance (r = .77), aspirations (r = .44), attendance (r = .22), and homework behavior (r = .67)—such that growth goal setting was associated with higher scores on these factors. Perceived instructional support was positively associated with growth goal setting, such that organization and clarity (r = .48), feedback-feed- forward (r = .52), and relevance (r = .54) were related to higher rates of growth goal setting in the sample. There were also significant links between perceived instruc-

tional support and student engagement. Organization and clarity was correlated with perseverance (r = .42), aspirations (r = .32), attendance (r = .18), and homework behavior (r = .48). Feedback- feedforward was correlated with perseverance (r = .47), aspira- tions (r = .33), attendance (r = .17), and homework behavior (r = .52). Relevance was correlated with perseverance (r = .45), aspira- tions (r = .29), attendance (r = .10), and homework behavior (r = .58).

Structural Equation Modeling

We then examined the extent to which (a) perceived instruc- tional support predicted growth goal setting, (b) growth goal set- ting predicted engagement, (c) growth goal setting mediated the link between perceived instructional support and student engage- ment, and (d) growth goal setting moderated the relationship between student background factors and engagement (see Figure 1). The SEM (without interactions) demonstrated adequate fit: v2(753) = 77,928.30, p , .001; CFI = .935; RMSEA = .041. Importantly, the inclusion of 16 interactions in the one model sig- nificantly improved model fit, as per the results of a chi-square dif- ference test: v2(16) = 653.85, p , .001. All significant and nonsignificant beta estimates are presented in Table 3. Here we describe all substantive beta estimates that attain the level of inter- pretability (at least a b $ .05 [as per Keith, 2006] and p , .001). All interpretable main effects are presented in Figure 2.

Main Effects

Regarding students’ perceptions of instructional support, both feedback-feedforward (b = .31; large effect size, as per Keith, 2006) and relevance (b = .38; large effect size) predicted higher levels of growth goal setting. Growth goal setting predicted gains in perseverance (b = .64; large effect size), aspirations (b = .46; large effect size), and homework behavior (b = .38; large effect size). Growth goal setting did not have a significant main effect on gains in attendance. There were also direct effects of perceived instructional support on engagement. Perceptions of organization and clarity predicted increases in attendance (b = .10; medium effect size). Perceived feedback-feedforward predicted gains in perseverance (b = .07; small effect size) and homework behavior (b = .10; medium effect size). Perceived relevance predicted a decrease in attendance (b = �.09; small effect size) and an increase in homework behavior (b = .25; large effect size).

Moderation Effects

In addition to testing these main effects, we also examined the extent to which growth goal setting moderated the effects of back- ground factors (year level, gender, SES, prior achievement) on engagement. Of the 16 interactions tested, two interactions attained the level of interpretability (i.e., p , .001 and at least b $ .05; see the Method section): Growth Goal Setting 3 Prior Achievement on aspirations (b = �.07; small effect size) and Growth Goal Setting 3 SES on attendance (b = �.10; medium effect size). The nature of these interactions is presented in Figure 3 and Figure 4, respectively. The results of simple slopes analysis are also included in the figures.

In terms of aspirations (see Figure 3), for students low in growth goal setting, there was a stronger positive relationship between prior achievement and academic aspirations than for students high in growth goal setting—suggesting that growth goals narrow the gap in aspirations between low and high prior achieving students

Table 2 Standardized Results of the CFA

Variable

Organization and clarity (2019)

Feedback- feedforward

(2019) Relevance (2019)

Growth goal setting (2019)

Perseverance (2019)

Aspirations (2019)

Attendance (2019)

Homework behavior (2019)

2018 (prior) engagement .55*** .50*** .33*** .59*** Student background Year level (2019) �.02** .01 �.06*** �.02*** .01 .04*** �.04*** .03*** Gender (2019) .01 .01 �.03*** .05*** �.02* .14*** .05*** .07*** SES (2019) .16*** .19*** .16*** .26*** .30*** .26*** .19*** .30*** Prior achv. (2018) .11*** .13*** .03*** .12*** .18*** .32*** .21*** .21***

Perceived instructional support Organization and clarity (2019) —

Feedback-feedforward (2019) .79*** —

Relevance (2019) .68*** .62*** —

Central factor Growth goal setting (2019) .48*** .52*** .54*** —

Outcomes Perseverance (2019) .42*** .47*** .45*** .77*** —

Aspirations (2019) .32*** .33*** .29*** .44*** .42*** —

Attendance (2019) .18*** .17*** .10*** .22*** .21*** .26*** —

Homework behavior (2019) .48*** .52*** .58*** .67*** .66*** .46*** .19*** —

Note. SES = socioeconomic status; Prior achv. = prior achievement; CFA = confirmatory factor analysis. We have adjusted the instructional support fac- tor names from the original terminology used in their source survey—see the Method section for details. * p , .05. ** p , .01. *** p , .001.

762 MARTIN ET AL.

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

smaller. In terms of attendance (see Figure 4), for students low in growth goal setting, there was a nonsignificant relationship between SES and attendance (though marginally nonsignificant, p = .054), whereas there was a negative relationship between SES and attendance for students high in growth goal setting (the abso- lute difference between the two slopes yielded the overall signifi- cant moderation effect in Table 3)—suggesting a reversal of the negative effects of low SES for high growth goal students. Taken together, these results provide some support for growth goal set- ting moderating the negative effects of low prior achievement (on aspirations) and the negative effects of low-SES background (on attendance).

Indirect Effects Testing

Indirect effects were examined to determine the extent to which growth goal setting may mediate the association between students’ perceptions of instructional support and their engage- ment. Of the 12 possible indirect effects, nine were tested because the nonsignificant association between growth goal set- ting and attendance ruled out the remaining three. Of the nine indirect effects tested, six attained the level of interpretability (at least b $ .05 and p , .001). The interpretable effects are described here, and all results are presented in Table 4. Students’ report of their teacher’s feedback-feedforward was significantly positively related (via growth goal setting) to gains in persever- ance (b = .20; medium effect size), aspirations (b = .08; small effect size), and homework behavior (b = .11; medium effect size). Perceived relevance was significantly positively associated with (via growth goal setting) gains in perseverance (b = .24; large effect size), aspirations (b = .10; medium effect size), and homework behavior (b = .13; medium effect size). Perceptions of instructional organization and clarity did not have significant indirect links with any engagement outcomes via growth goal setting. Taken together, these results suggest that students’

perceptions of feedback-feedforward and relevance were not only directly linked to engagement, they were also indirectly associated with engagement via students’ growth goal setting. That is, for example, if students perceive lessons to be relevant, they are more likely to set growth goals, which in turn is posi- tively associated with their engagement.

Discussion

The present investigation contributes to existing knowledge about growth goal setting in four ways. First, it identified the role of perceived instructional support in predicting growth goal setting and academic engagement, with most of the significant paths yielding medium and large effect sizes. Second, it revealed signifi- cant yields of growth goal setting for improvements on several dimensions of academic engagement, with all significant paths having large effect sizes. Third, the investigation showed that the link between students’ perceptions of instructional support and engagement was significantly partially mediated by growth goal setting, with most significant paths having medium effect sizes. Fourth, the study identified how growth goal setting may adap- tively moderate the effects of some student background attributes on engagement, yielding small and medium effect sizes.

Perceived Instructional Support, Growth Goal Setting, and Engagement

Students’ goal setting is impacted, in part, by their perceptions of the support they receive from teachers (Eberley et al., 2011). In our study, two of the three elements of instructional support examined (relevance and feedback-feedforward) were found to significantly predict growth goal setting. In addition, student reports of organiza- tion and clarity significantly predicted gains in attendance; feed- back-feedforward significantly predicted gains in perseverance and homework behavior, and relevance significantly predicted gains in

Table 3 Standardized Results of SEM

Variable Growth goal setting (2019)

Perseverance (2019)

Aspirations (2019)

Attendance (2019)

Homework behavior (2019)

2018 (prior) engagement .28***† .35***† .28***† .31***†

Student background Year level (2019) .04*** .06***† �.02*** .10***†

Gender (2019) �.06***† .07***† .02* .02** SES (2019) .06***† .16***† .10***† .07***†

Prior achv. (2019) .03*** .05*** .07***† .07***†

Perceived instructional support Organization and clarity (2019) �.03* .01 .05*** .10***† �.03*** Feedback-feedforward (2019) .31***† .07***† .04*** .02 .10***†

Relevance (2019) .38***† .01 .03*** �.09***† .25***†

Central Factor Growth goal setting (2019) .64***† .46***† �.03 .38***†

Interactions Growth Goal Setting 3 Year Level .01 �.02*** �.01 �.01* Growth Goal Setting 3 Gender .01* �.03*** .05*** .02** Growth Goal Setting 3 SES .01** �.05*** �.10***† .03*** Growth Goal Setting 3 Prior Achv. �.02*** �.07***† �.02 �.02***

Note. SES = socioeconomic status; Prior achv. = prior achievement; SEM = structural equation modeling. We have adjusted the instructional support fac- tor names from the original terminology used in their source survey—see the Method section for details. Please note that the covariates and perceived instructional support factors were correlated within the model. * p , .05. ** p , .01. *** p , .001. ***†Regression estimates that reach the level of interpretability of b $ .05 and p , .001.

GROWTH GOAL SETTING 763

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

homework behavior but also declines in attendance. Predominantly, then, students who felt they received adequate instructional support from teachers were more likely to set goals targeting personal growth and improvement. In turn, these perceptions of instructional support were linked to students’ engagement directly and via their growth goal setting. Indeed, we identified significant indirect paths between perceived instructional support and engagement via growth goal setting (see Table 4). With regard to relevance, we earlier argued how models of in-

terest development emphasize the importance of aligning content and tasks with students’ personal interests and experience base. When students perceive relevance in academic content and activ- ity, they are likely to establish and pursue self-set goals (Koestner, 2008; see also Aelterman et al., 2019). Thus, perceived instruc- tional support characterized by relevance will underlie goals that are integrated with the self—a hallmark of growth goals (Martin & Elliot, 2016a). Our findings supported this thesis: perceived instructional relevance significantly predicted growth goal setting. Interestingly, perceived relevance was differentially related to

homework behavior and attendance. Relevance was associated with gains in adaptive homework behavior and this is consistent with long-standing advice to educators to ensure relevance in order for students to engage with homework in a quality manner (e.g., Wilson & Rhodes, 2010; see also Hulleman & Harackiewicz, 2009; Hulleman et al., 2017). However, after controlling for the

impact of organization and clarity and feedback-feedforward, rele- vance was associated with declines in school attendance. Accord- ing to Hidi and Renninger (2006), relevant content and tasks are effective for initial and situational interest. This may be assistive for behaviors such as homework but perhaps not sufficient for motivating more substantial behaviors such as school attendance. It is when students are also provided with appropriate guidance and ongoing support that they are more likely to develop a deeper and more enduring drive to engage (Hidi & Renninger, 2006). We would therefore suggest that when seeking to enhance students’ attendance at school, relevance alone may not be sufficient; teach- ers should especially attend to organization and clarity, and feed- back-feedforward.

As we noted in the Introduction, researchers have looked to combine classic feedback (i.e., corrective information; Shute, 2008) with improvement-oriented guidance (feedforward)—lead- ing to greater consideration of a construct referred to as feedback- feedforward (e.g., Basso & Olivetti Belardinelli, 2006; Burns et al., 2019; Hattie & Timperley, 2007; Martin, 2016). We contended that to establish the validity of this feedback-feedforward con- struct, research should demonstrate that it is associated with improvement-oriented strategies and outcomes. Indeed, this was what our study found: Student reports of feedback-feedforward were associated with growth goal setting (an improvement-ori- ented strategy) and gains in perseverance and homework behavior.

Figure 2 Significant Standardized Substantive Results From Structural Equation Model

T1

T1

T1

Relevance

Feedback-

feedforward

Growth Goal

Setting

Perseverance

Aspirations

Attendance

Homework

Behavior

T1

.31

.38

.28

.35

.28

.31

.64

.46

.38

.10

.10

.07

-.09

.25

ENVIRONMENT (Instruction) PERSON (Student Strategy) OUTCOME (Engagement)

Organization

and Clarity

Note. All results presented in figure attain the level of interpretability (b $ .05 and p , .001). See Table 3 for all significant and nonsignificant main, covariate, and moderating effects. We have adjusted the instructional support factor names from the original terminology used in their source survey—see the Method section for details. T1 = 2018 (prior) engagement.

764 MARTIN ET AL.

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

Growth Goal Setting and Gains in Engagement

Growth goal setting predicted significant gains across one aca- demic year in students’ perseverance, aspirations, and homework behavior. This suggests that students who set growth goals are more likely to persevere, aspire to complete year 12, and demon- strate more adaptive homework behaviors. Thus, with regard to this study’s engagement factors, it appears that personally rele- vant goals that push students to achieve new standards of perso- nal success motivate them to think and behave in ways that are

constructive to personal goal completion. These findings are fur- ther “proof of concept” for growth goal setting: growth goals are fundamentally about growth, and our findings demonstrated growth in engagement was significantly associated with stu- dents’ use of growth goal setting. Importantly, these findings occurred in a large, representative sample of Australian high school students (N = 61,879).

DeShon and Gillespie (2005) proposed that goals create a men- tal framework that influences individuals’ choices and behaviors. Harnessing this conception of goals, Burns et al. (2018) suggested

Figure 3 Plot of the Moderating Effects of Growth Goal Setting on the Relationship Between Prior Achievement (P.Achv) and Aspirations

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

Low P.Achv. High P.Achv.

As pi

ra �o

ns

Low Growth Goal Se�ng High Growth Goal Se�ng

Low GG CI High GG CI

Note. The results presented here are from the unstandardized output. �1 SD (low growth goal setting): b = .27, p , .001; þ1 SD (high growth goal setting): b = .10, p , .001. GG CI = growth goal setting confidence interval (95%).

Figure 4 Plot of the Moderating Effects of Growth Goal Setting on the Relationship Between Socioeconomic Status (SES) and Attendance

Note. The results presented here are from the unstandardized output. �1 SD (low growth goal setting): b = .01, p = .054; þ1 SD (high growth goal setting): b = �.07, p , .001. GG CI = growth goal setting confidence interval (95%).

GROWTH GOAL SETTING 765

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

that growth goals reflect a personally referenced improvement-ori- ented mental framework that leads to adaptive choices and behav- iors. Moreover, Burns et al. (2018) contended that because growth goals are inherently self-referenced and self-generated, they pre- dict activities and processes that require ongoing personal invest- ment, such as engagement. Our findings supported these ideas through the gains in aspirations, perseverance, and homework behavior that occurred as a function of students’ growth goal setting. It was interesting that growth goal setting did not significantly

predict attendance. This may reflect the wording of our growth goal setting measure that focused on schoolwork specifically, whereas school attendance is a broader engagement construct that encompasses not just schoolwork but also major activities such as one’s social and physical (e.g., sport) life. However, it may also reflect the reality that school attendance is an issue for some stu- dents more than others. It may be that growth in attendance is con- strained by factors beyond and separate from students’ growth goals (e.g., transport to school, physical health, etc.). It was thus noteworthy that we found a significant interaction (moderation) between growth goal setting and socioeconomic status for attend- ance, discussed below.

The Moderating Role of Growth Goal Setting

The study also examined the interaction between growth goal setting and student background attributes. It was found that growth goal setting moderated the association between prior achievement and students’ aspirations, and between SES and attendance. Thus, in line with SCT (Bandura, 1997), our findings demonstrated that personal factors do interact to impact human agency; specifically, background factors particular to a student and their growth goal setting interacted in their association with academic engagement. Regarding the moderating associations with aspirations, students

who reported low growth goal setting and who were lower in prior achievement reported the lowest intentions to complete year 12; however, the aspirations gap between students with low and high prior achievement decreased if students with low prior achieve- ment adopted growth goal setting. This suggests that although growth goal setting improved all students’ aspirations to complete year 12 (as seen in the significant main effect of growth goal set- ting on aspirations; Table 3 and Figure 2), it seemed to have an even greater bolstering effect for lower achieving students—

apparently helping to reduce the aspirations gap. This finding is in line with previous research (Bardach et al., 2020) which showed that growth goals helped to attenuate the relationship between con- textual problems (e.g., lack of support) and dropout intentions among university students. Thus, students who pursue growth goals may be relatively more protected from the potential negative impact stemming from a variety of sources (e.g., low prior achievement, contextual issues).

For the moderating association with attendance, we found an attenuation in the gap between students from low- and high-SES backgrounds as a function of growth goal setting. In these data, for students with high growth goals, there was a negative relationship between SES and attendance, whereas this relationship was (mar- ginally) nonsignificant for students with low growth goals. This suggests that growth goal setting may minimize differences in school attendance between students from low- and high-SES back- grounds. As we continue to strive to do better on narrowing school completion and attendance gaps between students and particular student groups (e.g., low-SES etc.; Lamb et al., 2015), the findings here provide guidance on how we might do so from a motivation perspective. Specifically, for students with low prior achievement and low-SES backgrounds, growth goal setting was associated with gains in aspirations to complete school and in school attend- ance. When students attend school and continue through to year 12, they receive more hours of instruction and are significantly more likely to develop the core skills and knowledge that underpin their learning through school—and beyond (Willms, 2003). Addi- tionally, year 12 completion is associated with a variety of positive life outcomes, such as higher employment rates and life satisfac- tion (Bridgeland et al., 2006).

Implications for Theory

A major objective of the current study was to examine how recent theoretical developments in goal setting (viz., growth goal setting) function in the context of SCT and triadic theorizing (Ban- dura, 1986, 1991). Consistent with Burns et al. (2018), who grounded their research of growth goal setting in SCT, our find- ings supported the viability of including growth goal setting as a personal (self-) strategy within the SCT and the triadic model. Our data also confirmed Burns et al.’s observation that growth goal set- ting may represent an effective means of operationalizing one of the key mechanisms proposed under SCT. Specifically, SCT

Table 4 Standardized Results of Indirect Effects Testing (Nonparametric Bootstrapping)

Indirect path Indirect effect SE Lower 95% CI Upper 95% CI

Organization and clarity ! Growth goal setting ! Perseverance (2019) �.02** .007 �.03 �.01 Feedback-feedforward ! Growth goal setting ! Perseverance (2019) .20***† .007 .19 .21 Relevance ! Growth goal setting ! Perseverance (2019) .24***† .006 .23 .25 Organization and clarity ! Growth goal setting ! Aspirations (2019) �.01** .003 �.01 �.01 Feedback-feedforward ! Growth goal setting ! Aspirations (2019) .08***† .003 .07 .08 Relevance ! Growth goal setting ! Aspirations (2019) .10***† .003 .09 .10 Organization and clarity ! Growth goal setting ! Homework behavior (2019) �.01** .004 �.02 �.01 Feedback-feedforward ! Growth goal setting ! Homework behavior (2019) .11***† .004 .10 .11 Relevance ! Growth Soal setting ! Homework behavior (2019) .13***† .004 .13 .14

Note. The indirect paths involving Organization and clarity ! Growth goal setting are to be interpreted cautiously because the direct path between these variables, although significant at p , .05, did not meet the initial level of interpretability (see Figure 2 and Table 3). ** p , .01. ***†Regression estimates that reach the level of interpretability of b $ .05 and p , .001.

766 MARTIN ET AL.

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

contends that self-improvement occurs through a personal discrep- ancy production and reduction process (Bandura, 1991). Here, individuals set and pursue personally challenging goals to reach their desired future states (Bandura, 1991). When they reach their desired future state, the process begins again. Growth goal setting scales (e.g., Martin, 2006) comprise items that inherently access this function and thus may represent an important development for operationalizing key functions within SCT (Burns et al., 2018). The present study’s demonstration of engagement gains through growth goal setting provided further support for this.

Implications for Practice

One key finding was that students’ growth goal setting was associated with academic engagement gains. Thus, promoting growth goal setting within the classroom is likely to have positive engagement yields. Martin (2006) has articulated various practical directions for promoting growth goal setting. For example, stu- dents can be taught how to identify clear, realistic, and personally challenging goals (Martin, 2006). It is also important that students are taught how to strive toward their growth goal, such as by map- ping out the steps involved in working toward their growth goal and monitoring their progress toward that goal (Martin, 2011; see also Locke & Latham, 2013). Indeed, experimental research has demonstrated the effectiveness of growth goal setting intervention on students’ academic outcomes (Ginns et al., 2018; Martin et al., 2014; Martin & Elliot, 2016b). In addition to targeting growth goal setting directly, our findings

suggest that teachers’ instructional support may assist students’ growth goal setting and their academic engagement (directly and indirectly via growth goal setting). Based on our findings, this is particularly the case for instructional support characterized by feedback-feedforward and relevance. Here we suggest some indic- ative ways that teachers may embed more feedback-feedforward and relevance into their instructional approaches. Importantly, though, in identifying these instructional strategies we make the point that our study employed student reports of instructional sup- port. Thus, it is perceived instructional support that our study addresses, which has two implications for interpretation of find- ings and recommendations for practice. First, given perceptions of instructional support are clearly linked to growth goal setting and engagement, we might encourage students to be more aware of the instructional support they are receiving (thus, enhance their per- ception of instructional support). Second, notwithstanding this, given perceptions (not objectively assessed practice) were our focus we must be appropriately circumspect in the practical advice we do provide, as follows. Reflective diary-keeping may be one means by which students

can become more aware of the instructional support they receive. Here, students would be asked to reflect on what aspects of instructional support may have helped them set and strive for goals and/or assisted them to engage in their schoolwork and with school (e.g., do homework, persist, etc.). Research has shown this to be an effective means of enhancing growth goal setting and striving (Travers et al., 2015). This may instill in students an appreciation for the different ways that teachers are trying to sup- port their learning. With regard to feedback-feedforward, there has been substantial

guidance about feedback and the importance of it being timely,

concrete, accurate, and so forth (Centre for Education Statistics & Evaluation, 2020; Hattie & Timperley, 2007; Shute, 2008) and so we endorse that line of advice for practitioners. Additionally, we recommend the importance of that feedback being accompanied by improvement-oriented guidance (i.e., feedforward) to the stu- dent. This, we contend, is a key means for growth—a claim we suggest is borne out by the positive association between feedback- feedforward and growth goal setting in the present study (and the association between feedback-feedforward and gains in persever- ance and homework behavior).

In promoting relevance, researchers have emphasized the im- portance of content and tasks that are personally meaningful, sig- nificant, useful, and interesting (Hidi & Renninger, 2006; Lei et al., 2017; Van de Pol et al., 2010). These approaches to enhancing relevance enable content and tasks to be more integrated with the self which optimizes adaptive goal pursuit (Sheldon & Elliot, 1999). Importantly, however, as per interest development theoriz- ing (Hidi & Renninger, 2006; and our finding that relevance was associated with lower attendance), it is critical that relevance is accompanied by quality organization and clarity and feedback- feedforward; for some engagement outcomes, relevance alone may not be sufficient.

Limitations and Future Directions

Alongside the various contributions to research, theory, and practice, the present study comprised some limitations important to consider when interpreting findings and that provide some direction for future research. First, although we had objective (prior) achievement as a covariate, our outcome variables were self-reported. This was a valid means for measuring personal phe- nomena (Brener et al., 2003), but there are known limitations (e.g., misinterpreting items, under- or overreporting, etc.; Karabe- nick et al., 2007). Future research might look to include data from other sources, such as teacher observations and parent reports of students’ engagement. Relatedly, our data were based on student reports of instructional support (hence, perceived instructional support). Collecting instructional data from the teachers them- selves or from observations will augment future research in this space. Moreover, when exploring these additional data, it may be illuminating to test alternative models. Ours was a “fully-forward” model that was not assessed against models that had paths removed or against models that more explicitly and fully test re- ciprocal effects.

Second, although our growth goal setting and engagement measures were domain-general, we constructed domain-general instructional support measures through collapsing (domain-spe- cific) items from English, mathematics, and science subjects. There is a need for domain-specific data to be collected on growth goal setting and engagement to enable us to investigate the links between domain-specific instructional support and domain-specific goals and engagement. Third, we did not have between-class data and our variance components tests revealed negligible between- school variance. Thus, our analyses were conducted at the student- level (though, we did include cluster weighting and we did adjust standard errors to account for the nesting of students within schools). Future investigations would do well to design research appropriate for multilevel modeling that accounts for student- and class-level relationships.

GROWTH GOAL SETTING 767

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

Fourth, we focused on processes and factors consistent with SCT and the triadic model (Bandura, 1986, 1991), and we acknowledge other factors that are potentially implicated. For example, integrating measures of perceived goal structures may be helpful (Meece et al., 2006). Also, we focused on goal-setting theory (Locke & Latham, 2013) but not so much on other goal frameworks such as achievement goal theory (Elliot, 2005), which would involve additional growth-oriented measures under the pro- posed 3 3 2 goal framework (Elliot et al., 2011). Finally, although correlational data such as those in our study can provide tentative insights into possible causal ordering, experimental research is best placed to establish causality. Encouragingly, intervention research into growth goal setting has demonstrated promise (e.g., Ginns et al., 2018; Martin & Elliot, 2016b; Martin et al., 2014). It would now also be interesting to conduct experimental research that explores promotion of our instructional support factors and effects on growth goal setting.

Conclusion

The present study demonstrates that their growth goal setting is associated with significant gains in their academic engagement. There is also evidence that growth goal setting buffers the poten- tial negative effects of low SES and low prior achievement on some engagement outcomes. The findings further revealed a sig- nificant role of perceived instructional support in predicting stu- dents’ growth goal setting—and thus, a mediating role of growth goal setting between instructional support and academic engage- ment. Taken together, then, this study sheds further light on the role of growth goal setting in the context of major theoretical frameworks informing human agency. It also provides specific direction about the instructional support factors that educators can address when seeking to optimize students’ growth goal setting and engagement at school.

References

Aelterman, N., Vansteenkiste, M., Soenens, B., Fontaine, J., Haerens, L., Delrue, J., & Reeve, J. (2019). Towards a fine-grained understanding of the components of need-supportive and need-thwarting teaching: The merits of a gradual approach. Journal of Educational Psychology, 111(3), 497–521. https://doi.org/10.1037/edu0000293

Anderman, E. M., & Maehr, M. L. (1994). Motivation and schooling in the middle grades. Review of Educational Research, 64(2), 287–309. https:// doi.org/10.3102/00346543064002287

Assor, A., Kaplan, H., & Roth, G. (2002). Choice is good, but relevance is excellent: Autonomy-enhancing and suppressing teacher behaviours pre- dicting students’ engagement in schoolwork. The British Journal of Educational Psychology, 72(Pt. 2), 261–278. https://doi.org/10.1348/ 000709902158883

Bandura, A. (1986). Social foundations of thought and action. Prentice Hall.

Bandura, A. (1991). Social cognitive theory of self-regulation. Organiza- tional Behavior and Human Decision Processes, 50(2), 248–287. https://doi.org/10.1016/0749-5978(91)90022-L

Bandura, A. (1997). Self-efficacy: The exercise of control. Freeman. Bardach, L., Yanagida, T., Klassen, R. M., & Lüftenegger, M. (2020). Norma- tive and appearance performance-approach goal structures: Two-level factor structure and external linkages. Journal of Experimental Education. Advance online publication. https://doi.org/10.1080/00220973.2020.1729081

Basso, D., & Olivetti Belardinelli, M. (2006). The role of the feedforward paradigm in cognitive psychology. Cognitive Processing, 7(2), 73–88. https://doi.org/10.1007/s10339-006-0034-1

Bempechat, J., & Shernoff, D. J. (2012). Parental influences on achieve- ment motivation and student engagement. In S. Christenson, A. Reschly & C. Wylie (Eds.), Handbook of research on student engagement (pp. 315–342). Springer. https://doi.org/10.1007/978-1-4614-2018-7_15

Bong, M. (2001). Between- and within-domain relations of academic moti- vation among middle and high school students: Self-efficacy, task value, and achievement goals. Journal of Educational Psychology, 93(1), 23–34. https://doi.org/10.1037/0022-0663.93.1.23

Bostwick, K., Collie, R. J., Martin, A. J., & Durksen, T. (2020). Teacher, classroom, and student growth orientation in mathematics: A multilevel examination of growth goals, growth mindset, engagement, and achieve- ment. Teaching and Teacher Education, 94, 103100. https://doi.org/10 .1016/j.tate.2020.103100

Brener, N. D., Billy, J. O., & Grady, W. R. (2003). Assessment of factors affecting the validity of self-reported health-risk behavior among adoles- cents: Evidence from the scientific literature. The Journal of Adolescent Health, 33(6), 436–457. https://doi.org/10.1016/S1054-139X(03)00052-1

Bridgeland, J. M., Dilulio, J. J., & Morison, K. (2006). The silent epi- demic: Perspectives of high school dropouts. Civic Enterprises in asso- ciation with Peter D. Hart Research Associates for the Bill & Melinda Gates Foundation. http://eric.ed.gov/?id=ED513444

Burns, E. C. (2020). Factors that support high school completion: A longi- tudinal examination of quality teacher-student relationships and inten- tions to graduate. Journal of Adolescence, 84, 180–189. https://doi.org/ 10.1016/j.adolescence.2020.09.005

Burns, E. C., Bostwick, K. C. P., Collie, R. J., & Martin, A. J. (2019). Girls’ disengagement: Latent growth modelling of trajectories and the role of teacher and peer support. Journal of Youth and Adolescence, 48(5), 979–995. https://doi.org/10.1007/s10964-019-00986-4

Burns, E. C., Martin, A. J., & Collie, R. J. (2017). Understanding the role of adaptability and personal best (PB) goals in students’ academic out- comes: A social cognitive perspective. British Journal of Educational Psychology Monograph Series II, 12, 111–131.

Burns, E. C., Martin, A. J., & Collie, R. J. (2018). Adaptability, personal best (PB) goal setting, and gains in students’ academic outcomes: A lon- gitudinal examination from a social cognitive perspective. Contempo- rary Educational Psychology, 53, 57–72. https://doi.org/10.1016/j .cedpsych.2018.02.001

Burns, E. C., Martin, A. J., & Collie, R. J. (2019). Understanding the role of personal best (PB) goal setting in students’ declining engagement: A latent growth model. Journal of Educational Psychology, 111(4), 557–572. https://doi.org/10.1037/edu0000291

Burns, E. C., Martin, A. J., & Collie, R. J. (2021). A future time perspec- tive understanding of high school students’ engagement and disengage- ment: A longitudinal investigation. Journal of School Psychology, 84(4), 109–123. https://doi.org/10.1016/j.jsp.2020.12.003

Burns, E. C., Martin, A. J., & Evans, P. A. (2019). The role of teacher feedback–feedforward and personal best goal setting in students’ mathe- matics achievement: A goal setting theory perspective. Educational Psy- chology. Advance online publication. https://doi.org/10.1080/01443410 .2019.1662889

Byrne, B. M. (2012). Structural equation modeling with Mplus: Basic con- cepts, applications, and programming. Routledge. https://doi.org/10 .4324/9780203807644

Centre for Education Statistics and Evaluation. (2020). What works best: 2020 update. NSW Department of Education.

Chen, F. F. (2007). Sensitivity of goodness of fit indices to lack of mea- surement invariance. Structural Equation Modeling, 14(3), 464–504. https://doi.org/10.1080/10705510701301834

768 MARTIN ET AL.

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Model- ing, 9(2), 233–255. https://doi.org/10.1207/S15328007SEM0902_5

Claro, S., Paunesku, D., & Dweck, C. S. (2016). Growth mindset tempers the effects of poverty on academic achievement. Proceedings of the National Academy of Sciences of the United States of America, 113(31), 8664–8668. https://doi.org/10.1073/pnas.1608207113

Collie, R. J., Granziera, H., & Martin, A. J. (2019). Teachers’ motivational approach: Links with students’ basic psychological need frustration, maladaptive engagement, and academic outcomes. Teaching and Teacher Education, 86, 102872. https://doi.org/10.1016/j.tate.2019.07 .002

Collie, R. J., Martin, A. J., Papworth, B., & Ginns, P. (2016). Students’ interpersonal relationships, personal best (PB) goals, and academic engagement. Learning and Individual Differences, 45, 65–76. https://doi .org/10.1016/j.lindif.2015.12.002

Cook, T. D., Deng, Y., & Morgano, E. (2007). Friendship influences dur- ing early adolescence: The special role of friends’ grade point average. Journal of Research on Adolescence, 17(2), 325–356. https://doi.org/10 .1111/j.1532-7795.2007.00525.x

Cutmore, M., MacLeod, S., Donlevy, V., Spence, C., Martin, A. J., & Collie, R. J. (2018). Against the odds—Academically resilient students with a migrant background and how they succeed: Final report. Euro- pean Commission.

DeShon, R. P., & Gillespie, J. Z. (2005). A motivated action theory account of goal orientation. Journal of Applied Psychology, 90(6), 1096–1127. https://doi.org/10.1037/0021-9010.90.6.1096

Dweck, C. (2017). Mindset: Changing the way you think to fulfil your potential. Hachette.

Eberley, M. B., Holley, E. C., Johnson, M. D., & Mitchell, T. R. (2011). Beyond internal and external: A dyadic theory of relational attributions. Academy of Management Review, 36(4), 731–753. https://doi.org/10 .5465/amr.2009.0371

Elliot, A. J. (2005). A conceptual history of the achievement goal con- struct. In A. J. Elliot & C. S. Dweck (Eds.), Handbook of competence and motivation (pp. 52–72). Guilford Press.

Elliot, A. J., Murayama, K., & Pekrun, R. (2011). A 3 3 2 achievement goal model. Journal of Educational Psychology, 103(3), 632–648. https://doi.org/10.1037/a0023952

Elliot, A., Murayama, K., Kobeisy, A., & Lichtenfeld, S. (2015). Potential- based achievement goals. The British Journal of Educational Psychol- ogy, 85(2), 192–206. https://doi.org/10.1111/bjep.12051

Fan, H., Xu, J., Cai, Z., He, J., & Fan, X. (2017). Homework and students’ achievement in math and science: A 30-year meta-analysis, 1986-2015. Educational Research Review, 20, 35–54. https://doi.org/10.1016/j .edurev.2016.11.003

Fredricks, J. A., Blumenfeld, P. C., & Paris, A. H. (2004). School engage- ment: Potential of the concept, state of the evidence. Review of Educational Research, 74(1), 59–109. https://doi.org/10.3102/00346543074001059

Froiland, J. M., & Worrell, F. C. (2016). Intrinsic motivation, learning goals, engagement, and achievement in a diverse high school. Psychol- ogy in the Schools, 53(3), 321–336. https://doi.org/10.1002/pits.21901

Ginns, P., Martin, A. J., Durksen, T., Burns, E. C., & Pope, A. (2018). Per- sonal Best (PB) goal-setting enhances arithmetical problem-solving. Australian Educational Researcher, 45(4), 533–551. https://doi.org/10 .1007/s13384-018-0268-9

Green, J., Martin, A. J., & Marsh, H. W. (2007). Motivation and engage- ment in English, mathematics and science high school subjects: Towards an understanding of multidimensional domain specificity. Learning and Individual Differences, 17(3), 269–279. https://doi.org/10.1016/j.lindif .2006.12.003

Grolnick, W. S. (2003). The psychology of parental control: How well-meant parenting backfires. Erlbaum. https://doi.org/10.4324/9781410606303

Gutman, L. M., & Schoon, I. (2018). Emotional engagement, educational aspirations, and their association during secondary school. Journal of Adolescence, 67, 109–119. https://doi.org/10.1016/j.adolescence.2018 .05.014

Hamre, B. K., Pianta, R. C., Mashburn, A. J., & Downer, J. T. (2007). Building a science of classrooms: Application of the CLASS framework in over 4,000 U.S. early childhood and elementary classrooms. Report for Foundation for Childhood Development https://www.fcd-us.org/ building-a-science-of-classrooms-application-of-the-class-framework-in -over-4000-u-s-early-childhood-and-elementary-classrooms/

Hattie, J. (2009). Visible learning. Routledge. https://doi.org/10.4324/ 9780203887332

Hattie, J., & Timperley, H. (2007). The power of feedback. Review of Educa- tional Research, 77(1), 81–112. https://doi.org/10.3102/003465430298487

Hidi, S., & Renninger, K. A. (2006). The four-phase model of interest de- velopment. Educational Psychologist, 41(2), 111–127. https://doi.org/10 .1207/s15326985ep4102_4

Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covari- ance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55. https://doi.org/10.1080/ 10705519909540118

Hulleman, C. S., & Harackiewicz, J. M. (2009). Promoting interest and performance in high school science classes. Science, 326(5958), 1410–1412. https://doi.org/10.1126/science.1177067

Hulleman, C. S., Kosovich, J. J., Barron, K. E., & Daniel, D. B. (2017). Making connections: Replicating and extending the utility value inter- vention in the classroom. Journal of Educational Psychology, 109(3), 387–404. https://doi.org/10.1037/edu0000146

Hulleman, C. S., Schrager, S. M., Bodmann, S. M., & Harackiewicz, J. M. (2010). A meta-analytic review of achievement goal measures: Different labels for the same constructs or different constructs with similar labels? Psychological Bulletin, 136(3), 422–449. https://doi.org/10.1037/a0018947

Karabenick, S. A., Woolley, M. E., Friedel, J. M., Ammon, B. V., Blazevski, J., Bonney, C. R., Groot, E. D. E., Gilbert, M. C., Musu, L., Kempler, T. M., & Kelly, K. L. (2007). Cognitive processing of self-report items in educational research: Do they think what we mean? Educational Psycholo- gist, 42(3), 139–151. https://doi.org/10.1080/00461520701416231

Keith, T. Z. (2006).Multiple regression and beyond. Pearson Education. Kline, R. B. (2016). Principles and practice of structural equation model-

ing (4th ed.). Guilford Press. Koestner, R. (2008). Reaching one’s personal goals: A motivational per-

spective focused on autonomy. Canadian Psychology, 49(1), 60–67. https://doi.org/10.1037/0708-5591.49.1.60

Koestner, R., Lekes, N., Powers, T. A., & Chicoine, E. (2002). Attaining personal goals: Self-concordance plus implementation intentions equals success. Journal of Personality and Social Psychology, 83(1), 231–244. https://doi.org/10.1037/0022-3514.83.1.231

Lamb, S., Jackson, J., Walstab, A., & Huo, S. (2015). Educational oppor- tunity in Australia 2015: Who succeeds and who misses out? Centre for International Research on Education Systems, Victoria University, for the Mitchell Institute.

Lei, H., Cui, Y., & Chiu, M. M. (2017). The relationship between teacher support and students’ academic emotions: A meta-analysis. Frontiers in Psychology, 8, 2288. https://doi.org/10.3389/fpsyg.2017.02288

Liem, G. A., Ginns, P., Martin, A. J., Stone, B., & Herrett, M. (2012). Per- sonal best goals and academic and social functioning: A longitudinal perspective. Learning and Instruction, 22(3), 222–230. https://doi.org/ 10.1016/j.learninstruc.2011.11.003

Linnenbrink-Garcia, L., Tyson, D. F., & Patall, E. A. (2008). When are achievement goal orientations beneficial for academic achievement? A closer look at moderating factors. Revue Internationale de Psychologie Sociale, 21, 19–70.

Lipstein, R., & Renninger, K. A. (2006). “Putting things into words”: 12- 15-year-old students’ interest for writing. In P. Boscolo & S. Hidi

GROWTH GOAL SETTING 769

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

(Eds.), Motivation and writing: Research and school practice. Kluwer Academic/Plenum Press. https://doi.org/10.1163/9781849508216_008

Locke, E. A., & Latham, G. P. (2002). Building a practically useful theory of goal setting and task motivation. A 35-year odyssey. American Psy- chologist, 57(9), 705–717. https://doi.org/10.1037/0003-066X.57.9.705

Locke, E. A., & Latham, G. P. (Eds.). (2013). Goal setting theory: The cur- rent state. New developments in goal setting and task performance (pp. 623–630). Routledge. https://doi.org/10.4324/9780203082744

Maehr, M. L., & Zusho, A. (2009). Achievement goal theory: The past, present, and future. In K. R. Wentzel & A. Wigfield (Eds.), Handbook of motivation at school (pp. 77–104). Routledge.

Marsh, H. W., Martin, A. J., & Cheng, J. H. (2008). A multilevel perspec- tive on gender in classroom motivation and climate: Potential benefits of male teachers for boys? Journal of Educational Psychology, 100(1), 78–95. https://doi.org/10.1037/0022-0663.100.1.78

Martin, A. J. (2004). School motivation of boys and girls: Differences of degree, differences of kind, or both? Australian Journal of Psychology, 56(3), 133–146. https://doi.org/10.1080/00049530412331283363

Martin, A. J. (2006). Personal bests (PBs): A proposed multidimensional model and empirical analysis. The British Journal of Educational Psy- chology, 76(Pt. 4), 803–825. https://doi.org/10.1348/000709905X55389

Martin, A. J. (2007). Examining a multidimensional model of student moti- vation and engagement using a construct validation approach. The Brit- ish Journal of Educational Psychology, 77(Pt. 2), 413–440. https://doi .org/10.1348/000709906X118036

Martin, A. J. (2009). Motivation and engagement across the academic life- span: A developmental construct validity study of elementary school, high school, and university/college students. Educational and Psychological Measurement, 69(5), 794–824. https://doi.org/10.1177/0013164409332214

Martin, A. J. (2011). Personal best (PB) approaches to academic develop- ment: Implications for motivation and assessment. Educational Practice and Theory, 33(1), 93–99. https://doi.org/10.7459/ept/33.1.06

Martin, A. J. (2012a). Motivation and engagement: Conceptual, opera- tional and empirical clarity. Section Commentary. In S. Christenson, A. Reschly, & C. Wylie (Eds.), Handbook of research on student engage- ment. Springer. https://doi.org/10.1007/978-1-4614-2018-7_14

Martin, A. J. (2012b). The role of Personal Best (PB) goals in the achieve- ment and behavioral engagement of students with ADHD and students without ADHD. Contemporary Educational Psychology, 37(2), 91–105. https://doi.org/10.1016/j.cedpsych.2012.01.002

Martin, A. J. (2015). Implicit theories about intelligence and growth (per- sonal best) goals: Exploring reciprocal relationships. The British Journal of Educational Psychology, 85(2), 207–223. https://doi.org/10.1111/bjep .12038

Martin, A. J. (2016). Using Load Reduction Instruction (LRI) to boost motivation and engagement. British Psychological Society.

Martin, A. J., & Elliot, A. J. (2016a). The role of personal best (PB) and di- chotomous achievement goals in students’ academic motivation and engagement: A longitudinal investigation. Educational Psychology, 36(7), 1285–1302. https://doi.org/10.1080/01443410.2015.1093606

Martin, A. J., & Elliot, A. J. (2016b). The role of personal best (PB) goal setting in students’ academic achievement gains. Learning and Individ- ual Differences, 45, 222–227. https://doi.org/10.1016/j.lindif.2015.12 .014

Martin, A. J., & Evans, P. (2018). Load Reduction Instruction: Exploring a framework that assesses explicit instruction through to independent learning. Teaching and Teacher Education, 73, 203–214. https://doi.org/ 10.1016/j.tate.2018.03.018

Martin, A. J., & Evans, P. (2019). Load reduction instruction: Sequencing explicit instruction and guided discovery to enhance students’ motiva- tion, engagement, learning, and achievement. In S. Tindall-Ford, S. Agostinho, & J. Sweller (Eds.), Advances in Cognitive Load Theory: Rethinking teaching. Routledge.

Martin, A. J., & Liem, G. A. (2010). Academic Personal Bests (PBs), engagement, and achievement: A cross-lagged panel analysis. Learning and Individual Differences, 20(3), 265–270. https://doi.org/10.1016/j .lindif.2010.01.001

Martin, A. J., Bobis, J., Anderson, J., Way, J., & Vellar, R. (2011). Patterns of multilevel variance in psycho-educational phenomena: Comparing motivation, engagement, climate, teaching, and achievement factors Zeitschrift für Pädagogische Psychologie/German Journal of Educa- tional Psychology, 25(1), 49–61. https://doi.org/10.1024/1010-0652/ a000029

Martin, A. J., Collie, R. J., Mok, M. M., & McInerney, D. M. (2016). Per- sonal best (PB) goal structure, individual PB goals, engagement, and achievement: A study of Chinese- and English-speaking background students in Australian schools. The British Journal of Educational Psy- chology, 86(1), 75–91. https://doi.org/10.1111/bjep.12092

Martin, A. J., Durksen, T. L., Williamson, D., Kiss, J., & Ginns, P. (2014). Personal best (PB) goal setting and students’ motivation in science: A study of science valuing and aspirations. Australian Educational and Developmental Psychologist, 31(2), 85–96. https://doi.org/10.1017/edp .2014.19

Martin, A. J., Ginns, P., Burns, E., Kennett, R., & Pearson, J. (2020). Load reduction instruction in science and students’ science engagement and science achievement. Journal of Educational Psychology. Advance online publication. https://doi.org/10.1037/edu0000552

Martin, A. J., Way, J., Bobis, J., & Anderson, J. (2015). Exploring the ups and downs of mathematics engagement in the middle years of school. The Journal of Early Adolescence, 35(2), 199–244. https://doi.org/10 .1177/0272431614529365

Martin, F., & Bolliger, D. U. (2018). Engagement matters: Student per- ceptions on the importance of engagement strategies in the online learning environment. Online Learning, 22(1), 205–222. https://doi .org/10.24059/olj.v22i1.1092

Mayer, R. E., & Moreno, R. (2010). Techniques that reduce extraneous cognitive load and manage intrinsic cognitive load during multimedia learning. In J. L. Plass, R. Moreno, & R. Brunken (Eds.), Cognitive load theory (pp. 131–152). Cambridge University Press. https://doi.org/10 .1017/CBO9780511844744.009

Meece, J. L., Anderman, E. M., & Anderman, L. H. (2006). Classroom goal structure, student motivation, and academic achievement. Annual Review of Psychology, 57, 487–503. https://doi.org/10.1146/annurev .psych.56.091103.070258

Meredith, W. (1993). Measurement invariance, factor analysis and facto- rial invariance. Psychometrika, 58(4), 525–543. https://doi.org/10.1007/ BF02294825

Muthén, L. K., & Muthén, B. O. (2017).Mplus user’s guide (8th ed.). Ramdass, D., & Zimmerman, B. J. (2011). Developing self-regulation

skills: The important role of homework. Journal of Advanced Academ- ics, 22(2), 194–218. https://doi.org/10.1177/1932202X1102200202

Reeve, J. (2006). Teachers as facilitators: What autonomy-supportive teachers do and why their students benefit. The Elementary School Jour- nal, 106(3), 225–236. https://doi.org/10.1086/501484

Reeve, J., Jang, H., Carrell, D., Jeon, S., & Barch, J. (2004). Enhancing students’ engagement by increasing teachers’ autonomy support. Moti- vation and Emotion, 28(2), 147–169. https://doi.org/10.1023/B:MOEM .0000032312.95499.6f

Renninger, K. A., & Hidi, S. (2015). The power of interest for motivation and engagement. Routledge. https://doi.org/10.4324/9781315771045

Renninger, K. A., & Hidi, S. E. (2020). To level the playing field, develop interest. Policy Insights from the Behavioral and Brain Sciences, 7(1), 10–18. https://doi.org/10.1177/2372732219864705

Reyes, M. R., Brackett, M. A., Rivers, S. E., White, M., & Salovey, P. (2012). Classroom emotional climate, student engagement, and aca- demic achievement. Journal of Educational Psychology, 104(3), 700–712. https://doi.org/10.1037/a0027268

770 MARTIN ET AL.

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.

Ruzek, E. A., & Schenke, K. (2019). The tenuous link between classroom perceptions and motivation: A within-person longitudinal study. Journal of Educational Psychology, 111(5), 903–917. https://doi.org/10.1037/ edu0000323

Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilita- tion of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68–78. https://doi.org/10.1037/0003-066X.55.1.68

Schenke, K., Ruzek, E., Lam, A. C., Karabenick, S. A., & Eccles, J. S. (2017). Heterogeneity of student perceptions of the classroom climate: A latent profile approach. Learning Environments Research, 20(3), 289–306. https://doi.org/10.1007/s10984-017-9235-z

Schippers, M., Scheepers, A., & Peterson, J. (2015). A scalable goal-set- ting intervention closes both the gender and ethnic minority achievement gap. Palgrave Communications, 1(1), 15014 https://doi.org/10.1057/ palcomms.2015.14

Sheldon, K. M., & Elliot, A. J. (1999). Goal striving, need satisfaction, and longitudinal well-being: The self-concordance model. Journal of Per- sonality and Social Psychology, 76(3), 482–497. https://doi.org/10 .1037/0022-3514.76.3.482

Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonex- perimental studies: New procedures and recommendations. Psychologi- cal Methods, 7(4), 422–445. https://doi.org/10.1037/1082-989X.7.4.422

Shute, V. J. (2008). Focus on formative feedback. Review of Educational Research, 78(1), 153–189. https://doi.org/10.3102/0034654307313795

Sirin, S. R. (2005). Socioeconomic status and academic achievement: A meta-analytic review of research. Review of Educational Research, 75(3), 417–453. https://doi.org/10.3102/00346543075003417

Sterba, S. K. (2017). Pattern mixture models for quantifying missing data uncertainty in longitudinal invariance testing. Structural Equation Mod- eling, 24(2), 283–300. https://doi.org/10.1080/10705511.2016.1250635

Sweller, J. (2012). Human cognitive architecture: Why some instructional procedures work and others do not. In T. Urdan, K. R. Harris, S. Graham (Ed.), APA educational psychology handbook (pp. 295–325). American Psychological Association. https://doi.org/10.1037/13273-011

The Learning Bar. (2019). Tell Them From Me Survey. https:// thelearningbar.com/about-us/?lang=aue

Titsworth, S., Mazer, J. P., Goodboy, A. K., Bolkan, S., & Myers, S. A. (2015). Two meta-analyses exploring the relationship between teacher clarity and student learning. Communication Education, 64(4), 385–418. https://doi.org/10.1080/03634523.2015.1041998

Trautwein, U., Lüdtke, O., Kastens, C., & Köller, O. (2006). Effort on homework in grades 5-9: Development, motivational antecedents, and the association with effort on classwork. Child Development, 77(4), 1094–1111. https://doi.org/10.1111/j.1467-8624.2006.00921.x

Trautwein, U., Lüdtke, O., Schnyder, I., & Niggli, A. (2006). Predicting homework effort: Support for a domain-specific, multilevel homework model. Journal of Educational Psychology, 98(2), 438–456. https://doi .org/10.1037/0022-0663.98.2.438

Travers, C. J., Morisano, D., & Locke, E. A. (2015). Self-reflection, growth goals, and academic outcomes: A qualitative study. The British Journal of Educational Psychology, 85(2), 224–241. https://doi.org/10 .1111/bjep.12059

Van de Gaer, E., Pustjens, H., Van Damme, J., & De Munter, A. (2009). School engagement and language achievement: A longitudinal study of gender differences across secondary school. Merrill-Palmer Quarterly, 55, 373–405. https://doi.org/10.1353/mpq.0.0034

Van de Pol, J., Volman, M., & Beishuizen, J. (2010). Scaffolding in teacher-student interaction: A decade of research. Educational Psychol- ogy Review, 22(3), 271–296. https://doi.org/10.1007/s10648-010-9127-6

Van de Schoot, R., Lugtig, P., & Hox, J. (2012). A checklist for testing measurement invariance. European Journal of Developmental Psychol- ogy, 9(4), 486–492. https://doi.org/10.1080/17405629.2012.686740

Van Ryzin, M. J., Gravely, A. A., & Roseth, C. J. (2009). Autonomy, belongingness, and engagement in school as contributors to adolescent psychological well-being. Journal of Youth and Adolescence, 38(1), 1–12. https://doi.org/10.1007/s10964-007-9257-4

Vandenberg, R. J., & Lance, C. E. (2000). A review and synthesis of the measurement invariance literature: Suggestions, practices, and recom- mendations for organizational research. Organizational Research Meth- ods, 3(1), 4–70. https://doi.org/10.1177/109442810031002

Vansteenkiste, M., Lens, W., Elliot, A. J., Soenens, B., & Mouratidis, A. (2014). Moving the achievement goal approach one step forward: To- ward a systematic examination of the autonomous and controlled rea- sons underlying achievement goals. Educational Psychologist, 49(3), 153–174. https://doi.org/10.1080/00461520.2014.928598

Vansteenkiste, M., Sierens, E., Goossens, L., Soenens, B., Dochy, F., Mouratidis, A., Aelterman, N., Haerens, L., & Beyers, W. (2012). Iden- tifying configurations of perceived teacher autonomy support and struc- ture: Associations with self-regulated learning, motivation and problem behavior. Learning and Instruction, 22(6), 431–439. https://doi.org/10 .1016/j.learninstruc.2012.04.002

Wentzel, K. R. (1999). Social-motivational processes and interpersonal relationships: Implications for understanding motivation at school. Jour- nal of Educational Psychology, 91(1), 76–97. https://doi.org/10.1037/ 0022-0663.91.1.76

Widaman, K. F., Ferrer, E., & Conger, R. D. (2010). Factorial invariance within longitudinal structural equation models: Measuring the same con- struct across time. Child Development Perspectives, 4(1), 10–18. https:// doi.org/10.1111/j.1750-8606.2009.00110.x

Wigfield, A., & Eccles, J. S. (2000). Expectancy–value theory of achieve- ment motivation. Contemporary Educational Psychology, 25(1), 68–81. https://doi.org/10.1006/ceps.1999.1015

Willms, J. D. (2003). Student engagement at school: A sense of belonging and participation: Results from PISA 2000. OECD Publishing.

Willms, J. D. (2014). Student engagement in New South Wales secondary schools: Findings from the Tell Them From Me Pilot. The Learning Bar.

Wilson, J., & Rhodes, J. (2010). Student perspectives on homework. Edu- cation, 131(2), 351–358.

Wisniewski, B., Zierer, K., & Hattie, J. (2019). The power of feedback revisited: A meta-analysis of educational feedback research. Frontiers in Psychology, 10, 3087. https://doi.org/10.3389/fpsyg.2019.03087

Yu, K., & Martin, A. J. (2014). Personal best (PB) and ‘classic’ achieve- ment goals in the Chinese context: Their role in predicting academic motivation, engagement, and buoyancy. Educational Psychology. An International Journal of Experimental Educational Psychology, 34, 635–658. https://doi.org/10.1080/01443410.2014.895297

Received August 14, 2020 Revision received March 12, 2021

Accepted March 15, 2021 n

GROWTH GOAL SETTING 771

T hi s do cu m en ti s co py ri gh te d by

th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n or

on e of

its al lie d pu bl is he rs .

C on te nt

m ay

be sh ar ed

at no

co st ,b ut

an y re qu es ts to

re us e th is co nt en ti n pa rt or

w ho le m us tg

o th ro ug h th e A m er ic an

Ps yc ho lo gi ca lA

ss oc ia tio

n.