The study of achievement goals is central to the achievement motivation literature. Achievement goals are defined as cognitive representations of competence-based end states that individuals strive to approach or avoid (Elliot, 1997). Several models of achievement goals (e.g., Dweck, 1986; Elliot, 1999; Nicholls, 1984) have been formulated over the years that vary regarding the focal components or number of achievement goals that individuals adopt in achievement situations. Despite some variation in conceptualization, theorists agree that the definition of competence (i.e., how doing well or poorly is defined) is an important dimension of achievement goals (Dweck & Leggett, 1988; Elliot & McGregor, 2001; Nicholls, 1984). The valence of competence (i.e., the approach-avoidance distinction) is also widely acknowledged as an important dimension of achievement goals (Elliot, 1997). Achievement goal frameworks, therefore, center on mastery goals (focusing on the development of competence and task- or self-based standards), performance goals (focusing on the demonstration of competence and other-based standards; Dweck, 1986; Nicholls, 1984), approach goals (approaching competence), and avoidance goals (avoiding incompetence; Elliot & Harackiewicz, 1996). In the present work, we focus on the 2 × 2 model of achievement goals in which the definition of competence is crossed with the valence of competence, resulting in four goals: Mastery-approach (striving to attain task- or self-based competence), performance-approach (striving to attain other-based competence), mastery-avoidance (striving to avoid task- or self-based incompetence), and performance-avoidance (striving to avoid other-based incompetence). This 2 × 2 model is commonly used in theoretical and empirical work in the achievement goal literature, as well as narrative and meta-analytic reviews (Butera et al., 2024).

Achievement goals are presumed to guide the way that people engage in achievement situations and how they cognitively, emotionally, and behaviorally respond to these situations (Ames, 1992; Dweck & Leggett, 1988). A great deal of research has shown that individuals’ achievement goals influence important outcomes such as performance, persistence, intrinsic motivation, and help-seeking behavior (for reviews, see Butera et al., 2024; Elliot & Hulleman, 2017; Senko, 2016). The influence of achievement goals has been extensively documented in a variety of different settings, especially school (e.g., Wirthwein et al., 2013), sports (Lochbaum & Gottardy, 2015), and work (Payne et al., 2007). In the present work we focus on the link between achievement goals and anxiety and depression symptoms and disorders.

Anxiety and depression are the most prevalent mental disorders and symptoms (APA, 2022). Theorists use the term “internalizing problems” for the cluster that includes clinical anxiety and depression or a combination of both symptoms or disorders (Yap et al., 2016). Anxiety affects a significant number of school-age children, with prevalence rates of an anxiety disorder of 7.1% in middle school and high school children (Ghandour et al., 2019), and prevalence rates of symptoms of 32% in college students (Sheldon et al., 2021). Depressive disorders have prevalence rates of approximately 3.5% in middle school and high school children (Ghandour et al., 2019) and the rate of depressive symptoms in college students reaches 25% (Sheldon et al., 2021). The prevalence rates of anxiety in the school-aged population have increased during the past few decades (Bitsko et al., 2018; Spoelma et al., 2023). Previous research has consistently shown that anxiety (e.g., generalized anxiety, panic disorder, separation anxiety, social anxiety; APA, 2013) is related to impairments across domains, such as emotional exhaustion (Koutsimani et al., 2019), and relational difficulties (Biswas et al., 2020). Similar results are also found for depression (e.g., disruptive mood dysregulation disorder, major depressive disorder; DSM-5; APA, 2013; Koutsimani et al., 2019; Marx et al., 2023). Previous narrative and meta-analytic reviews showed that in the educational context, anxiety and depression are related to a variety of negative outcomes, such as poor school attainment (see Riglin et al., 2014), poor school attendance (Finning et al., 2019), poor academic competence and greater school dropout (Brumariu et al., 2022), and problematic peer functioning (Christina et al., 2021), even at non-clinically diagnosed levels. Anxiety and depression show continuity in adulthood and, left untreated, could have severe consequences, even at subclinical levels, across one’s lifespan (e.g., poor quality of life and greater suicide risk; Marx et al., 2023).

Several narrative and meta-analytic reviews have tested how achievement goals are related to people’s emotional experiences, conceptualized as achievement emotions or affective states. In their meta-analysis, Payne et al. (2007) provided evidence of a negative relation between mastery-approach goals and state anxiety, as well as a positive relation between both performance-approach and performance-avoidance goals and state anxiety. Baranik et al.’s (2010) meta-analysis revealed positive relations of mastery-approach and performance-approach goals with positive affect, as well as positive relations of mastery-avoidance and performance-avoidance goals with negative affect. Huang (2011) replicated Baranik et al.’s findings, but also found a negative relation between performance-avoidance goals and positive affect, and a positive relation between performance-approach goals and negative affect.

Although these findings reveal that achievement goals are related to state anxiety and overall positive and negative affectivity, they do not address the links between these goals and clinical levels of anxiety and depression, or internalizing problems in general. Importantly, achievement emotions and affective states are conceptually distinct from clinical anxiety and depression, which are not based in a specific situation or event, but instead represent a constellation of symptoms that typically exert a significant impairment on one’s overall functioning (Luttenberger et al., 2018). Thus, the question of whether achievement goals are differentially related to internalizing problems has yet to be addressed meta-analytically. A growing number of individual studies have explored how achievement goals are related to internalizing problems (e.g., Madjar et al., 2021; Măirean & Diaconu-Gherasim, 2020; Sideridis, 2005), however, nothing is known about the cumulative nature and strength of these relations. The present meta-analysis advances the literature by synthesizing the existing empirical evidence on the relations between achievement goals and internalizing symptoms and disorders, and by investigating moderators of these links. In our work we focused on clinical anxiety and depression, and the combination of the two (i.e., global internalizing problems), as indicators of internalizing problems (Melton et al., 2016).

Basic Components and Models of Achievement Goals

As noted above, several theoretical models have guided research on achievement goals and we will test these different models in the present work. Initially, achievement goals models were dichotomous (Dweck, 1986; Nicholls, 1984), distinguishing between mastery goals (or task or learning goals) and performance goals (or ego or ability goals). Each goal was conceptualized as focusing on approaching success (Ames, 1992). Subsequently, the dichotomous model was extended by including the approach-avoidance distinction. Specifically, Elliot and Harackiewicz (1996) proposed the trichotomous model which bifurcates performance-based goals into performance-approach (focused on demonstrating other-based competence) and performance-avoidance (focused on avoiding the demonstration of other-based incompetence). In the 2 × 2 model, Elliot (1999) extended the trichotomous model by applying the approach-avoidance distinction to mastery-based goals; mastery-avoidance goals were conceptualized in terms of avoiding task-based or intrapersonal incompetence (see also Pintrich, 2000). Thus, the 2 × 2 model distinguishes among mastery-approach, mastery-avoidance, performance-approach, and performance-avoidance goals. This model implies that the task and self components of mastery-based goals combine together, when at times they may not. Therefore, in the 3 × 2 model (Elliot et al., 2011), mastery-based goals were split according to the standards used to evaluate competence, resulting in two approach goals (focused on attaining task-based competence and self-based competence) and two avoidance goals (focused on avoiding task-based incompetence and self-based incompetence). Other-based goals – other-approach and other-avoidance – are the same as the original performance-approach and performance-avoidance goals. The 3 × 2 model has not been tested with regard to internalizing problems, thus we focused on the 2 × 2 model in the present research.

The specific conceptualizations of achievement goals vary across studies in the literature, depending on which component (standpoints or standards) of competence is targeted. The standpoints of competence perspective views competence in terms of developing it vs. demonstrating it, whereas the standards of competence perspective evaluates competence with regard to task/self-based vs. other-based standards (Korn & Elliot, 2016). For example, in some studies, mastery-based goals focus on the development of ability, whereas in other studies they focus on task- or self-based standards; likewise for performance-based goals, some studies focus on the demonstration of ability (or appearance goals), whereas other studies focus on other-based standards (or normative goals; see Korn et al., 2019 for an overview).

Furthermore, the terminology and specific content of the goal measures varies somewhat across studies (see Hulleman et al., 2010 for an overview of these types of variation). Specifically, mastery-approach goals are conceptualized and operationalized in terms of fulfillment of one’s potential (e.g., mastery-approach goals, Elliot & McGregor, 2001), development of ability/competence (e.g., development-approach goals; Elliot et al., 2011), interest and curiosity (e.g., task orientation, Nicholls, 1984), or doing better than one has done in the past (e.g., learning goals; Dweck, 1986). Mastery-avoidance goals are conceptualized and operationalized in terms of being unable to reach one’s potential (mastery-avoidance goals, Elliot & McGregor, 2001), avoiding the development of inability/incompetence (development-avoidance goals, Korn & Elliot, 2016), or avoiding task-based incompetence (task-avoidance goals, Elliot et al., 2011). Performance-approach goals are conceptualized and operationalized in terms of doing better than others (e.g., performance-approach goals Elliot, 1999), appearing competent to others (e.g., demonstration-approach goals, Korn & Elliot, 2016), demonstrating ability relative to others (e.g., ego orientation, Nicholls, 1984), or confirming one’s ability to an audience (e.g., ability goals, Grant & Dweck, 2003). Performance-avoidance goals are conceptualized and operationalized in terms of avoiding doing worse than others (e.g., performance-avoidance goals, Elliot & Church, 1997), avoiding the demonstration of incompetence relative to others (e.g., demonstration-avoidance goals, Korn & Elliot, 2016), or avoiding negative judgments from others (e.g., avoid orientation, VandeWalle, 1997). We attend to this variation in terminology in our meta-analysis.

Theory and Empirical Work on Achievement Goals and Internalizing Problems

Achievement goals are about competence, a basic human need that must be satisfied to sustain mental health and well-being (Elliot et al., 2002; Ryan & Deci, 2019). Achievement goals represent the cognitive-dynamic lens through which people frame and interpret contexts, events, and outcomes (Dweck, 1986; Nicholls, 1984). As such, it is sensible to posit that these forms of self-regulation are associated with mental health indicators such as anxiety and depression, and several researchers have proposed links accordingly.

The most influential model of achievement goals and internalizing problems is the goal-orientation model of depression vulnerability (Dykman, 1998). Essentially all theorizing in this area is directly or indirectly grounded in this model. The model postulates that achievement goals lead to specific cognitive sets and appraisal patterns that have important implications for people’s mental health. Individuals focused on mastery-based goals are oriented toward growth, learning, and improvement, perceive negative outcomes as opportunities for self-development, and their self-worth is not contingent on performance or social comparison; thus, they experience lower levels of depressive symptoms and are more resilient to failure. Individuals focused on performance-based goals, on the other hand, are orientated toward comparing their success/ability to others and their self-worth is contingent on demonstrating competence relative to others. They tend to evaluate challenging or difficult situations as a reflection of their personal traits (e.g., incompetence, unlikability) and as a test of their ability; thus, they report lower self-worth, are less resilient to failure, and are more vulnerable to depression.

Although the goal orientation model is primarily focused on depression, both Dykman (1998) and others (e.g., Sideridis, 2007) have argued that it is also applicable to anxiety, with comparable patterns expected for depression and anxiety. Furthermore, researchers have extended this analysis to include the approach-avoidance distinction, contending that mastery-approach goals are the most beneficial and performance-avoidance goals are the most detrimental to internalizing problems (Duchesne et al., 2014; Van Boekel & Martin, 2014; Wang et al., 2021). In addition, researchers have noted that anxiety and depression are accompanied by impaired cognitive functioning, lack of access to resources, and poor self-regulation which exert an influence on the type of achievement goals that individuals adopt (e.g., less mastery-approach and more performance-avoidance goals; Duchesne et al., 2014; Măirean & Diaconu-Gherasim, 2020). In other words, the relation between achievement goals and internalizing problems is posited to be reciprocal.

There is a growing body of research examining the relation between achievement goals and internalizing symptoms, and somewhat inconsistent findings have been reported. Some studies have found that mastery-approach goals are related to a low level of anxiety symptoms (e.g., Ariani, 2017; Sideridis, 2005; Wei, 2018) and depressive symptoms (e.g., Madjar et al., 2021; Măirean & Diaconu-Gherasim, 2020; Sideridis, 2005), while others have not found these relations (e.g., Duchesne et al., 2014). Performance-approach and performance-avoidance goals have been linked to higher levels of anxiety (Ariani, 2017; Madjar et al., 2021) and depressive symptoms (Duchesne et al., 2014; Măirean & Diaconu-Gherasim, 2020), but null effects have also been reported (Duchesne et al., 2014; Madjar et al., 2021; Sideridis, 2005). Very few studies have investigated the relation between mastery-avoidance goals and internalizing symptoms; mixed findings have been reported (e.g., Liu et al., 2019; Wang et al., 2021).

Moderators of the Relation Between Achievement Goals and Internalizing Problems

The present meta-analysis also addresses potential moderators of the relation between achievement goals and internalizing problems. Several of these moderator variables (e.g., achievement goal model, achievement goal terminology) have been used in prior achievement goal meta-analyses focused on other outcome measures (Huang, 2011; Hulleman et al., 2010; Senko & Dawson, 2017). We examined four broad moderation categories: Conceptualization of achievement goals, conceptualization of internalizing problems, sample characteristics, and methodology- and publication-based characteristics.

Conceptualization of Achievement Goals

Achievement goal models vary not only in the number of goals but also in the way goals are conceptualized, with some models emphasizing the focus of competence alone (i.e., the dichotomous model) and others placing equal emphasis on the focus and the valence of competence (e.g., the 2 × 2 model; Elliot & Hulleman, 2017). Further, the terminology used to refer to the different goal categories varies across studies (e.g., development-approach, learning goals, task orientation for mastery-approach goals; see Hulleman et al., 2010 for an overview of terminology). Thus, in our meta-analysis we tested whether the achievement goal model (dichotomous, trichotomous, 2 × 2) and achievement goal terminology (i.e., the labels used for the different goals) moderated the relation between achievement goals and internalizing problems. Several different scales are used in the literature to assess achievement goals (e.g., the Patterns of Adaptive Learning Scale; PALS; Midgley et al., 1993; the Achievement Goal Questionnaire [Revised]; AGQ(-R); Elliot & Church, 1997) (see Hulleman et al., 2010 for an overview of scales). The different scales may emphasize different conceptual aspects of the goals (e.g., the PALS measures both normative and appearance aspects, whereas the AGQ-R measures only normative aspects), therefore we tested whether achievement goal scale (PALS, AGQ/AGQ-R, TEOSQ, other scales) moderated the relation between achievement goals and internalizing problems. We also tested whether achievement goal setting (general, specific) served as a moderator. Finally, we also assessed whether the specific type of setting (e.g., academic, sports) moderated the achievement goal-internalizing problem relation.

Conceptualization of Internalizing Problems

Both anxiety and depressive disorders vary in their manifestations and consequences (e.g., avoidance behavior and excessive fear in anxiety disorders; persistent sad mood and change in sleep and appetite in depressive disorders; APA, 2013). Thus, we evaluated form of anxiety and depression as a moderator of the relation between achievement goals and anxiety and depression. Furthermore, clinical anxiety and depression involve significant functional impairment (Craske et al., 2017), so we also evaluated whether the observed findings were more robust at the diagnostic level.

Sample Characteristics

We evaluated whether the participants’ education level (middle school, high school, college), cultural context where the studies were conducted (Western, Eastern), and participants’ gender and age served as moderators.

Methodology- and Publication-based

We evaluated several methodology- and publication-based characteristics as moderators: Measurement approach (self-report, other report), type of design utilized (cross-sectional, longitudinal), direction of the relation between achievement goals and internalizing problems, year of publication, and publication type (peer reviewed, non-peer reviewed).

The Present Research

The present meta-analysis seeks to synthesize the results of studies that have focused on the associations between achievement goals and internalizing symptoms or disorders, and investigates moderators of these relations. We aimed to evaluate the strength of the relations between the four achievement goals of the 2 × 2 model (i.e., mastery-approach, mastery-avoidance, performance-approach, and performance-avoidance) and internalizing symptoms and disorders (i.e., anxiety, depression, and a combination of both). We predicted, based on the aforementioned theorizing, that high levels of mastery-approach goals would be related to lower levels of internalizing symptoms and disorders, and high levels of performance-approach and performance-avoidance goals would be related to higher levels of internalizing symptoms and disorders. We did not make predictions for mastery-avoidance goals, given that they represent a hybrid of adaptive (mastery) and maladaptive (avoidance) components (Elliot & McGregor, 2001). By synthesizing data across studies that varied in several different ways, this meta-analysis promises to yield a more clear and thorough understanding of how achievement goals and internalizing problems are related than any individual study can provide. The relations are important for psychological functioning in general, including affect, cognition, and behavior among children, adolescents, and young adults in educational settings.

We examined four sets of moderators regarding the relations between achievement goals and internalizing problems: Conceptualization of achievement goals, conceptualization of internalizing problems, sample characteristics, and methodology- and publication-based characteristics. We had no a priori hypotheses for these moderators; these analyses are exploratory in nature. Evaluation of these moderators is important in order to understand when achievement goals are associated with anxiety and depression, thus furthering the precision and depth of our knowledge regarding these relations.

Method

Our systematic review was conducted according to the PRISMA 2020 (Preferred Reporting Items for Systematic reviews and Meta-Analyses; Page et al., 2021) guidelines. The review and meta-analysis protocol was pre-registered in PROSPERO International Prospective Register of Systematic Reviews (protocol number CRD42022298463).

Literature Search

The initial literature search was conducted in July 2021 in six electronic databases: Web of Science, PsycINFO, PubMed, ERIC, Academic Search Premier (EBSCO), and ProQuest. The final literature search for the work reported herein was conducted in December 2022. The same search strategy was applied in all databases, using the following combination of keywords: ("achievement goal" OR "goal orientation" OR "mastery goal" OR "mastery approach goal" OR "mastery orientation" OR mastery-approach OR "task goal" OR "task orientation" OR "learning goal" OR "development-approach goal" OR "task-approach goal" OR "self-approach goal" OR "mastery avoidance goal" OR mastery-avoid OR "development-avoidance goal" OR "task-avoidance goal" OR "self-avoidance goal" OR "performance goal" OR "performance approach goal" OR "ego goal" OR "ego orientation" OR "ability goal" OR "prove goal orientation" OR "performance-prove orientation" OR "self-enhancing goal orientation" OR "demonstration-approach goal" OR "other-approach goal" OR "performance avoidance goal" OR performance-avoidance OR "avoid goal orientation" OR "self-defeating ego orientation" OR "self-defeating orientation" OR "demonstration-avoidance goal" OR "other-avoidance goal") AND (internalizing OR "internalizing symptoms" OR "internalizing problems" OR "internalizing disorders" OR depression OR depressive OR depressed OR "depressive disorders" OR "depressive problems" OR "depressive symptoms" OR "depression symptoms" OR "mood disorders" OR sadness OR anxiety OR "generalized anxiety" OR "social anxiety" OR "anxiety symptoms" OR "anxiety problems" OR "anxiety disorders" OR anxious OR worry OR worries OR fear OR phobia OR phobic OR panic). The search was restricted to title and abstract fields and to the publication period 1980 (when the achievement goal approach emerged) to December, 2022 (when the final search was conducted). No restrictions were applied for publication type. All types of empirical research reports were eligible, including peer reviewed journal articles, conference papers, book chapters, and dissertation theses. Records and studies published in English, French, Spanish, and German were eligible for inclusion, due to the authors’ language competencies. The reference lists of the studies eligible for inclusion were searched for possible additional relevant studies. Unpublished studies were sought by contacting the authors with two or more published studies eligible for inclusion.

Study Selection

After removing duplicates, all of the records identified during the search stage were screened based on title and abstract by applying the eligibility criteria for inclusion and exclusion described below. Further, all retrieved full-text studies were assessed for eligibility and selected based on the inclusion and exclusion criteria. The same three coders were involved in the screening and selection of studies for inclusion in the systematic review. The coders overlapped with each other on a random sample of 50% of the studies and coded the studies independently. Kappa agreement between coders ranged from 0.76 to 0.91 for the abstract screening and 0.89 to 0.93 for the study selection. Disagreements between coders were resolved by discussion and consensus. Before the formal screening, the selection procedure was piloted on a random sample of studies. The flow diagram of the study selection process (following PRISMA 2020) is depicted in Fig. 1.

Fig. 1
figure 1

The PRISMA 2020 flow diagram

Full size image

The original quantitative studies that met the eligibility criteria were selected for inclusion. Inclusion criteria: (1) studies must include ratings of achievement goals (mastery-approach, mastery-avoidance, performance-approach, performance-avoidance, or their variants) and internalizing symptoms and disorders (depression, anxiety, or the combination of the two); if a study included an intervention or quasi-experiment, the achievement goals and internalizing measures had to have been collected prior to the intervention or quasi-experiment; (2) the study was written in English, French, Spanish, or German; (3) statistically relevant information was available for the relations between achievement goals and internalizing symptoms or disorders (e.g., correlation, sample size), allowing for the computation of effect size statistics. For the studies with insufficient reported data, the study’s investigators were contacted and requested to provide additional data (e.g., correlation coefficients).

Exclusion criteria: (1) theoretical papers, systematic reviews or meta-analyses, and qualitative studies; (2) studies that measured achievement goals at the group level (or achievement goal structures), and studies that induced achievement goals situationally; (3) studies measuring situational anxiety or anxiety related to specific settings such as educational or sport settings (e.g., test anxiety, sport anxiety, academic anxiety, learning/classroom anxiety, fear of failure, state anxiety, task anxiety, fear of failure, achievement-related emotions); (4) studies for which no full texts were accessible or sent by the authors upon request; (5) studies in which no statistical values for the relations between achievement goals and internalizing problems were reported or sent by the authors upon request. Reported data had to be independent of other studies included in the meta-analysis.

As depicted in Fig. 1, the database search resulted in 1992 records. After duplicates were removed (n = 955), the remaining 1037 unique records were screened based on the abstract, and 777 records were excluded. 260 eligible records were sought for retrieval and the 232 full-text reports that were available were assessed for inclusion, based on the eligibility above criteria. Further, 196 reports were excluded for the following reasons: Did not asses achievement goals or internalizing problems (n = 93), were not empirical, quantitative studies (n = 4), examined situational or context-specific anxiety or emotions (n = 79), assessed achievement goals or goal structures at the group level (n = 2), statistical data for calculating the effect size was unavailable (n = 14), and were published in other languages (n = 4). An additional 21 reports were identified from other sources (i.e., websites, references of included studies) and assessed for eligibility. Overall, 44 reports meeting the eligibility criteria were included in the meta-analysis (see Fig. 1 and Section 7 of the Supplementary Material). The reports include a total of 22,387 participants, in 47 samples of children or adults, between 11 and 60 years old. A summary description of the included studies is presented in Table S1, Supplementary Material.

Data Extraction and Coding

The following data were extracted and coded from each study: Conceptualization of achievement goals (achievement goal model, achievement goal terminology, achievement goal scale, achievement goal setting, type of informant); conceptualization of internalizing problems (indicators of internalizing problems, type of internalizing problems, form of anxiety, form of depression, type of informant), sample characteristics (education level, cultural context, age, gender); methodological and publication characteristics (type of design, direction of relation, type of publication, year of publication). The data extracted and coded are presented in Table 1. The correlation coefficients between each achievement goal (mastery-approach, mastery-avoidance, performance-approach, performance-avoidance) and each indicator of internalizing symptoms and disorders (anxiety, depression, their combination) were also extracted. The same three coders from the screening stage extracted and coded the data from the included studies. The coders overlapped on a random sample of 25% of the studies that where coded independently by two coders. Kappa agreement between coders was higher than 94% for all of the categories. Disagreements between the coders were resolved by discussion and consensus coding.

Table 1 Coding of data extracted from the studies
Full size table

Effect-Size Calculation

Correlation coefficients (i.e., Pearson’s r) between each type of achievement goal (mastery-approach, mastery-avoidance, performance-approach, and performance-avoidance) and each indicator of internalizing problems (anxiety, depression, their combination) were extracted for effect sizes. When other effect size indicators (e.g., F tests) were reported, we converted them to correlation coefficients.

To ensure the independence of effect sizes, from each study a single effect size derived from a particular sample was included. If a study reported effect size information for different samples (e.g., students), these were considered independent, and the effect sizes from each sample were included. When a study reported more than one effect size from one sample for a particular analysis (e.g., the correlation between performance-avoidance goals and anxiety), several decisions were made to avoid dependency: 1) if a study reported both cross-sectional and longitudinal associations between two indicators, we only included the longitudinal effect size to take advantage of longitudinal research; 2) if a study reported cross-sectional associations from multiple time points between two indicators without reporting longitudinal associations, the coefficients were aggregated into a single effect size; 3) if a study reported the effect sizes for multiple measures of the same indicator (e.g., more than one measure of anxiety) from the same sample and time point, these were aggregated into a single effect size; 4) if a study reported several longitudinal effect sizes between two indicators from multiple time points, these were aggregated into a single effect size.

To test our hypotheses, a random effects meta-analysis model was conducted (Hedges & Olkin, 1985). We computed the correlation coefficient for the relation between each type of achievement goal and each indicator of internalizing problems, along with a 95% confidence interval (CI). Egger’s intercept test was used as a publication bias assessment at the global level, testing the funnel plot’s symmetry. Also, a funnel plot analysis was performed for testing publication bias at the moderator level.

Moderator Analysis

Heterogeneity of effect size was computed using Q statistics (QB; Borenstein, 2009) to test whether the relations between the achievement goals and internalizing problems were moderated by the following categorical moderators: 1) conceptualization of achievement goals: Achievement goal model, achievement goal terminology, achievement goal scale, achievement goal setting, and achievement goal informant; 2) conceptualization of internalizing problems: Type of internalizing problems, form of anxiety, form of depression, and internalizing problems informant; 3) sample characteristics: Education level, and cultural context; 4) methodology- and publication-based characteristics of the studies: Type of study design, direction of the relations in longitudinal studies, year of publication, and type of publication. Categorical moderators were evaluated using subgroup analyses. We assessed the significant differences between the categories based on the Q Between test for subgroup differences for a random effect when a moderator had more than two categories. If the moderator had more than two categories, we also conducted a follow-up analysis and compared pairs of categories. For each pair, we assessed the significant differences between categories based on the Q Between test for subgroup differences, as we did in the case of moderators with only two categories. Consistent with previous meta-analytic work (e.g., Brumariu et al., 2022), we included a potential moderator only if there were four or more studies available per level. The following moderators were not included in subgroup analyses due to an insufficient number of reports (less than four): Type of informant (all studies used self-reports of achievement goals and internalizing problems), type of internalizing problems (no study assessed disorders), form of depression, and direction of the relation between achievement goals and indicators of internalizing problems. Global internalizing problems (a combination of both depression and anxiety) were examined in a single study, so we were unable to conduct a separate meta-analytic evaluation of the relations between achievement goals and this variable. Not all categories were available for each categorical moderator (e.g., the TEASQ category for the relation between mastery-approach goals and anxiety), thus Tables 2 and 3 present the results of the categorical moderators when each category is at least k = 4. For achievement goal scale the “other scales” category was created, and this category included other validated scales: The Goal Orientation Inventory (Dykman, 1998), the Goal Inventory (Roedel et al., 1994), the Goal Orientations and Motivational Beliefs scale (Niemivirta, 2002), and one author-created scale (Stornes & Bru, 2011). The results for follow-up analyses, when a moderator had more than two categories, are presented in the text. Mastery-avoidance goals are not included in any moderator analyses due to an insufficient number of studies. Meta-regression analyses were conducted to evaluate the role of continuous moderators (participant age, percentage of males, and publication year) in the relation between each type of achievement goal and internalizing problems. We present results for these moderators in the text.

Table 2 Results of the meta-analyses between achievement goals and anxiety and their categorical moderators
Full size table
Table 3 Results of the meta-analyses between achievement goals and depression and their categorical moderators
Full size table

The meta-analyses were conducted using R (Version 4.3.2; R Core Team, 2023) and the R-packages dmetar (Version 0.1.0; Harrer et al., 2019), esc (Version 0.5.1; Lüdecke, 2019), MAd (Version 0.8.3; Hoyt, 2014), maditr (Version 0.8.4; Demin, 2024), and meta (Balduzzi et al., 2019; Version 6.5.0; Harrer et al., 2019).

Results

Associations Between Achievement Goals and Anxiety

As seen in Table 2, mastery-approach goals were negatively related to anxiety, whereas performance-approach goals were positively related. Performance-avoidance goals were also positively related to anxiety, whereas the relation between mastery-avoidance goals and anxiety was not significant (see Forest plots in Section 2 of Supplementary Material). The significant within-group heterogeneity estimates (QW values) suggested that there is heterogeneity of the effect sizes.

Moderators of the Relation Between Achievement Goals and Anxiety (see Table 2)

Mastery-Approach Goals

Three of eight categorical variables tested were significant moderators of the negative relation between mastery-approach goals and anxiety. Achievement goal terminology was significant; the negative relation was stronger for development-approach goals than mastery goals/orientation (QB = 7.79, p = 0.005), or mastery-approach goals (QB = 8.07, p = 0.004), and did not differ from all other terms (all QB > 0.05). Achievement goal setting was significant; the negative relation was stronger in studies in a general setting than in specific settings. Education level was significant; the negative relation was stronger in studies with college samples than in studies with middle school samples (QB = 6.89, p = 0.014) or high school samples (QB = 6.17, p = 0.035). Achievement goal model, achievement goal scale, form of anxiety, type of design and type of publication were not significant. Age (B =  − 0.004), percentage of males (B =  − 0.002), and year of publication (B = 0.006) were not significant.

Performance-Approach Goals

Three of seven categorical variables tested were significant categorical moderators of the positive relation between performance-approach goals and anxiety. Achievement goal terminology was significant; the positive relation was stronger for demonstration-approach goals than for performance-approach goals. Achievement goal setting was significant; the positive relation was stronger in a general setting than in specific settings. Education level was significant; the positive relation was stronger for college students than for high school (QB = 6.17, p = 0.013) or middle school (QB = 6.89, p = 0.009) students. Achievement goal model, achievement goal scale, form of anxiety, and type of publication were not significant moderators. The meta-regression indicated that the percentage of males was a significant moderator; the relation between performance-approach goals and anxiety decreased as the percentage of males increased (B =  − 0.003, p = 0.03). Age (B = 0.009) and year of publication (B =  − 0.003) were not significant.

Performance-Avoidance Goals

Two of six categorical variables tested were significant categorical moderators of the positive relation between performance-avoidance goals and anxiety. Achievement goal model was significant; the positive relation was stronger in studies using the trichotomous model than in studies using the 2 × 2 model. Achievement goal scale was significant; the positive relation was stronger in studies that used other scales than in studies that used the AGQ/AGQ-R (QB = 5.94, p = 0.015). Form of anxiety, education level, type of design and type of publication were not significant. The meta-regression for age was significant; the relation between the performance-avoidance goals and anxiety increased as participants’ age increased (B = 0.008, p = 0.002). Percentage of males (B =  − 0.0003) and year of publication (B =  − 0.005) were not significant.

Associations Between the Achievement Goals and Depression

Mastery-approach goals were negatively related, whereas performance-avoidance goals were positively related to depression (see Table 3). Performance-approach goals and mastery-avoidance goals were not significantly related to depression, and mastery-avoidance goals was not significantly related to depression (see Forest plots in Section 3 of Supplementary Material). The significant within-group heterogeneity estimates (QW values) suggested that there is heterogeneity of the effect sizes.

Moderators of the Relations Between the Achievement Goals and Depression (see Table 3)

Mastery-Approach Goals

None of categorical variables tested (achievement goal model, achievement goal terminology, achievement goal scale, achievement goal setting, education level, cultural context, type of publication) were significant moderators of the negative relation between mastery-approach goals and depression. The meta-regression for year of publication was significant (B =  − 0.007, p = 0.014); the relation between the mastery-approach goals and depression decreased for more recent publications. Age (B =  − 0.002) and percentage of males (B =  − 0.001) were not significant.

Performance-Approach Goals

Three of seven categorical variables tested were significant moderators of the relation between performance-approach goals and depression. Achievement goal model was significant; the relation was positive and significant only in studies that used the dichotomous model. Achievement goal scale was significant, but no statistical differences were observed by follow-up analyses (all QB > 0.05), however, the relation between performance-approach and depression was negative in studies that used the PALS and positive in studies that used all other scales. Cultural context was significant; the relation was positive and significant only in studies conducted in Western countries. Achievement goal terminology, achievement goal setting, education level and type of publication were not significant. The meta-regressions indicated that year of publication was significant (B =  − 0.011, p = 0.042); the relation between the performance-approach goals and depression decreased for more recent publications. Age (B = 0.01) and percentage of males (B =  − 0.002) were not significant.

Performance-Avoidance Goals

None of five categorical variables (achievement goal model, achievement goal scale, education level, type of publication, cultural context) tested were significant moderators of the positive relation between performance-avoidance goals and depression. Meta-regressions indicated that age (B = 0.0001), percentage of males (B =  − 0.001), and year of publication (B = 0.003) were not signification moderators.

Publication Bias

Egger’s test for studies evaluating anxiety indicated there was no indication of publication bias for mastery-approach goals, t(32) =  − 1.00, p = 0.32, and that the probability of publication bias was significant for performance-approach goals, t(33) = 3.44, p = 0.002, and performance-avoidance goals, t(22) = 2.11, p = 0.04. A trim-and-fill analysis suggested a significant left asymmetry for performance-approach goals, and that an overestimation of the global effect size was plausible (11 powerful studies and one low-powered study should be included to compensate the possible overestimation of the effect; see Funnel plots in Section 4 of Supplementary Material). Similar bias was observed for performance-avoidance goals, suggesting that the effect could be overestimated (2 low-powered studies and 4 medium-powered studies should be included to compensate the possible overestimation of the effect). For depression, there was no indication of publication bias in the studies assessing mastery-approach goals, t(34) =  − 1.73, performance-approach goals, t(32) =  − 1.21, and performance-avoidance goals, t(22) =  − 0.18, all ps > 0.05 (see Funnel plots in Section 5 of Supplementary Material).

Discussion

In the present meta-analytic work, we evaluated the strength of the relations of the four goals of the 2 × 2 achievement goal model with anxiety and depression. We found significant effect sizes linking mastery-approach goals and performance-avoidance goals to both anxiety and depression, and performance-approach goals to anxiety but not depression; no significant relations were found for mastery-avoidance goals. We also found significant moderation of these relations, indicating variation as a function of conceptualization of achievement goals, sample characteristics, and methodology- and publication-based characteristics.

Direct Relations Between Achievement Goals and Internalizing Problems

Our findings indicate that achievement goals are related to anxiety and depression. The effect sizes are small to medium in magnitude, with the significant relations ranging from r = 0.05 (for performance-avoidance goals and depression) to r = 0.25 (for performance-avoidance goals and anxiety). We found that individuals with higher levels of mastery-approach goals experience lower levels of both anxiety and depression. Individuals with higher levels of performance-approach goals, on the other hand, experience higher levels of anxiety, and exhibit a trend toward higher levels of depression. Those with higher levels of performance-avoidance goals experience higher levels of both anxiety and depression. Mastery-avoidance goals appear to be unrelated to anxiety and depression, although a positive trend is evident for these goals and depression.

These results are in line with and extend the goal-orientation model of depression (Dykman, 1998). Individuals who pursue mastery-approach goals are less vulnerable to anxiety and depression, likely because they are focused on growth and self-improvement, and appraise difficult situations/tasks as challenging (Dykman, 1998). Individuals who pursue performance-based goals experience higher levels of anxiety and depression, most likely because they are focused on comparing their success/ability to others (Dweck, 1986), and interpret difficult situations/tasks as a test of their ability (Dykman, 1998). Of note, achievement goals are associated with anxiety as well as depression (Sideridis, 2007), and the connection between performance-based goals and internalizing problems is, descriptively, particularly prominent for the avoidance manifestation of such goals (i.e., performance-avoidance). It is also important to note that our findings are mute regarding the causal direction of the relations; it is possible that the relations are bi-direction and feed into one another (Duchesne et al., 2014; Măirean & Diaconu-Gherasim, 2020).

Although our meta-analysis did not test processes that might explain the observed findings, several cognitive processes seem likely candidates. By encouraging growth and learning, mastery-approach goals might reduce dysfunctional attitudes and negative attributions (e.g., attributing failure to temporary, changeable factors), thus reducing vulnerability to anxiety and depression (Steare et al., 2024). Mastery-approach goals might also promote beliefs that ability can be developed through practice (growth mindset) that further encourage adaptive ways of coping with stress and failure that are related to low levels of anxiety and depression (Dykman, 1998; Yeager & Dweck, 2023). By promoting external standards based on social comparison, performance-based goals might lead to more negative attributions (e.g., attributing difficulty to internal, stable factors) and interpreting failure as a threat to one’s self-worth, facilitating stress and rumination, and elevated levels of anxiety and depression (Steare et al., 2024). Performance-based goals might also promote beliefs that ability cannot be developed through practice (fixed mindset), which is related to increased risk of anxiety and depression (Yeager & Dweck, 2023). Finally, to the degree that anxiety and depression themselves reduce mastery-approach and promote performance-based goals pursuit, they may do so by impairing cognitive functioning, coping, and self-regulation processes (Duchesne et al., 2014).

Mastery-avoidance goals are not significantly related to anxiety and depression which at first glance may suggest that these goals are not relevant to this type of psychopathology. However, these goals are focused on not losing one’s skills and abilities, and thus may be particularly prevalent among and applicable to older people (Elliot & McGregor, 2001). The vast majority of studies included in this meta-analysis were conducted on samples of middle school, high school, and college students, and the observed relations may be stronger in samples where age is more evenly distributed. It is also important to take care in interpreting the mastery-avoidance goal results given the small number of existing studies on anxiety (k = 4) and depression (k = 4). Although the results did not reach significance for mastery-avoidance goals and depression, the effect size was similar to those for mastery-approach and performance-avoidance goals and depression; thus, more research is needed to more clearly determine the precise nature of the mastery-avoidance goal-depression relation.

Considering the findings for the four achievement goals together, the pattern of results between achievement goals and internalizing problems is similar to the pattern found for other achievement-relevant processes and outcomes in the literature. Meta-analyses on variables such as achievement, intrinsic motivation, help seeking, etc. have consistently revealed that mastery-approach goals have the most adaptive pattern of relations, performance-avoidance goals have the least adaptive pattern, and the pattern for performance-approach and mastery-avoidance goals lies in between (e.g., Baranik et al., 2010; Hulleman et al., 2010; Wirthwein et al., 2013; see also Butera et al., 2024 for a narrative review). Perhaps most relevant to the current work, prior meta-analytic work has indicated that mastery-approach goals are negatively, and performance-approach, performance-avoidance, and mastery-avoidance goals are positively, related to negative achievement emotions (Huang, 2011). Our findings are consistent with this pattern for all but mastery-avoidance goals (which produce null results herein); this is interesting as it suggests that the mastery-avoidance goal pursuit may be less pernicious, with regard to emotional experience, than the two performance-based goals. Critically, our findings extend the prior work on negative affect by linking achievement goals to clinical anxiety and depression marked by persistent and intense emotional experience. Thus, the present work expands the achievement goal nomological network to include broad outcomes beyond the achievement domain and relevant to overall psychological functioning and mental health.

From a theoretical standpoint, our work may be seen as contributing to an understanding of the links between approach and avoidance motivation on one hand and anxiety and depression on the other. Gray’s Reinforcement Sensitivity Theory (Fowles, 1994; Gray, 1982) focuses on two basic, biologically-based motivational systems, the behavioral inhibition system (BIS; avoidance motivation) and the behavioral activation system (BAS; approach motivation). In this theory, the BIS is sensitive to stimuli representing nonreward, punishment, and novelty and involves moving away from (avoiding) or inhibiting undesirable affective states; high BIS sensitivity is positively associated with both anxiety and depression. The BAS is sensitive to stimuli representing reward and escape from punishment and involves moving toward (approaching) or maintaining desirable affective states; high BAS sensitivity is negatively association with depression, but not anxiety (see Katz et al., 2020 for further elaboration on these constructs and findings). Our findings at the goal level indicate that it is one type of avoidance motivation that is positively associated with anxiety, namely performance-avoidance goals, and that it is one type of approach motivation that is negatively associated with depression, namely mastery-approach goals. In addition, our findings indicate that one type of approach goals, performance-approach, is positively associated with anxiety. Thus, our findings show that the motivation-internalizing symptoms relations become more nuanced and specific as people regulate their basic energization tendencies with more concrete directional aims (Elliot & Thrash, 2002; see also Dickson & MacLeod, 2004 for related work on personal goals and internalizing symptoms).

Moderators of the Relations Between Achievement Goals and Internalizing Problems

Several of the tested moderator variable candidates were significant. Below we highlight what we perceive to be the most informative moderator variable findings.

Achievement goal model was a robust moderator across types of goals and internalizing problems. Specifically, the findings were significantly stronger for the dichotomous model than for the trichotomous and 2 × 2 models for the positive relation between performance-approach goals and depression (and they were descriptively stronger for the negative relation between mastery-approach goals and anxiety, and the positive relation performance-approach goals and anxiety). Further, the findings were significantly stronger for the trichotomous model than for the 2 × 2 model for the positive relations between performance-avoidance goals and both anxiety and depression (and they were descriptively stronger for the negative relations between mastery-approach goals and both anxiety and depression, and the positive relation between performance-approach goals and anxiety). Operationally, the achievement goals in the trichotomous model contain content other than goal standards per se, including a preference for challenge (mastery-approach), a desire to impress important others (performance-approach), and worries and fears (performance-avoidance; see Elliot & Church, 1997; Middleton & Midgley, 1997; Skaalvik, 1997; Vandewalle, 1997). These added components essentially create goal complexes that encompass both the goal standard and the reasons for pursuing that standard, and these reasons likely add to the predictive power of the standard (Senko & Tropiano, 2016; Sommet & Elliot, 2017). In essence (that is, heuristically, not technically), one can compare the trichotomous model effect sizes to those for the 2 × 2 model to get a rough estimate of the predictive utility gained by adding reason-based content to the goal standard content (for more on achievement goal complexes, see Sommet et al., 2021; Liem & Senko, 2022).

Another robust moderator variable was achievement goal terminology. We found that the goals using “development” and “demonstration” terminology showed the strongest relations when there were sufficient sample sizes for this moderator to be tested. Specifically, the negative relation between development-approach goals and anxiety (and, descriptively, depression), as well as the positive relation between demonstration-approach goals and anxiety were stronger than the relations observed for other achievement goal terms. It is interesting to note that the development and demonstration labels aren’t just distinct terms, but they also represent somewhat distinct content. That is, mastery-based goals include both task-based and self-based competence standards, whereas development-based goals focus on self-based standards only (Elliot et al., 2011; Korn & Elliot, 2016). The stronger associations of development- and demonstration-based goals are mirrored in the significant moderation for achievement goal scale. Studies using the PALS (Midgley et al., 1993) and the AGQ/AGQ-R (Elliot & Church, 1997) showed weaker associations than studies using other scales (e.g., Dykman, 1998). For example, the PALS measured both the normative and appearance aspect of performance-approach goals, the AGQ/AGQ-R measured only normative goals, whereas other scales measured performance goals as demonstration/appearance goals. Our results thus suggest a stronger negative impact of the development or demonstration/appearance component on anxiety and depression. As such, our results suggest that it is appetitive temporal striving – trying to increase one’s competence (development-approach) – that is most strongly linked to lower anxiety and depression. Likewise, performance-approach goals include other-based standards, whereas demonstration-based goals focus more on showing one’s ability to others. As such, our results suggest that a focus on appearance and demonstration may create a propensity for external motivators and a potential reliance on others for validation, which is likely to perpetuate anxious/depressed thoughts (see Hulleman et al., 2010; Senko & Dawson, 2017 for related work).

A third robust moderator variable was achievement goal setting. We found that the goals focusing on competence in general relative to those focusing on specific domains showed the strongest relations when there were sufficient sample sizes for this moderator to be tested. That is, domain-general mastery-approach goals had a stronger negative relation with both anxiety and depression than domain-specific mastery-approach goals. Likewise, domain-general performance-approach goals had a stronger positive relation with anxiety than domain-specific performance-approach goals, and domain-general mastery-approach goals had a stronger negative relation with anxiety (and, descriptively, depression) than domain-specific mastery-approach goals. This pattern of findings likely reflects the correspondence principle, which states that the relationship between two variables will be strongest when they are matched in level of generality-specificity (Ajzen & Fishbein, 1977). The anxiety and depression variables that were the focus of the present research represent broad, domain-general indicators of mental health, so it is sensible that they would be more strongly related to broad, domain-general indicators of achievement goals than narrow, domain-specific indicators. Achievement goal researchers (and researchers across domains and disciplines) would do well to attend to this often-overlooked principle in their work (for an empirical demonstration of the importance of correspondence in work on other achievement motivation constructs, see Chan et al., 2023).

A final point that we would like to highlight concerns the robustness of the performance-avoidance goal findings across moderators. The analyses did show that the performance-avoidance goals links to both anxiety and depression were moderated by a number of difference variables. However, this moderation almost exclusively revealed differences in the relative strength of significant findings, rather than revealing a significant finding under one condition but not another. For example, for each tested moderator of the positive relation between performance-avoidance goals and anxiety, the relation was significant and positive, only varying in magnitude; in fact, all but two of the eleven observed effect sizes for this relation dropped below the r = 0.30 mark. This robustness across moderators was unique to performance-avoidance goals, testifying to how this type of self-regulation represents a particularly pervasive and pernicious mental health vulnerability (Elliot & Hulleman, 2017).

It is important to also note that some other moderators – achievement goal scale, participants’ education level, cultural context, and type of design – were also relevant, even if only for a small number of relations. For example, with regard to cultural context, we found evidence supporting a universalist perspective and evidence supporting cultural differences (see Zusho & Clayton, 2011). Findings indicated that mastery-approach goals were positively and performance-avoidance goals negatively related to depression across cultures (a universalist finding), and findings also indicated that performance-approach goals were positively related to depression for Western Cultures but unrelated for Eastern cultures (a cultural difference). There were not enough existing studies to test for mastery-avoidance goal differences. Several things are noteworthy here. First, prior work showing cultural differences regarding achievement goals has tended to find differences for performance-avoidance goals. These goals fit the stronger collectivistic emphasis on avoiding negative outcomes in Eastern, relative to Western cultures, which accounts for why performance-avoidance goals are sometimes not detrimental and can even be beneficial in such contexts (Elliot et al., 2001). Importantly, this has been found for performance-based outcomes but not experience-based outcomes such as intrinsic motivation (see Hulleman et al., 2010). It may be that performance-avoidance goals afford performance benefits in Eastern contexts, but that the stress of regulating according to a negative normative possibility still exacts a toll on experience and well-being (Roskes et al., 2014). Second, our finding that performance-approach goals are detrimental for depression in Western but not Eastern cultures may also be a function of individualistic and collectivistic emphases. In Western cultures, individualistic values emphasize personal achievement and success relative to others, perhaps amplifying the impact of performance-approach goals on mental health. In Eastern cultures, on the other hand, collectivistic values emphasize in-group (e.g. family) achievement and success, and thus performance-approach goals may have fewer implications for how the self is construed (King et al., 2017), mitigating the impact of these goals on depression. Third, the fact that there were not enough existing studies to test for cultural differences for the mastery-avoidance goal to depression link nor for cultural differences for any achievement goal and anxiety highlights the clear need for more research in this important area.

In sum, the moderator variable analyses yielded several informative findings that provide a more precise and rich empirical picture than that gleaned from the omnibus relations alone. Nevertheless, two cautions are in order. First, some of the moderator tests should be interpreted with caution given modest numbers of available studies (e.g., mastery-approach goal, task orientation goal, and learning goal orientation variants of mastery-approach goals); furthermore, in some instances there were not enough samples to test for moderation (e.g., ego goal, ability goal, and prove goal variants of performance-approach goals). Thus, some effect sizes may be unstable (note that we highlighted the particularly robust findings above), some differences in associations between moderator levels are practically negligible albeit significantly different (e.g., stronger correlation between performance-avoidance goals and depression measured with other scales than with the AGQ/AGQ-R), and some important moderator information may be missing altogether (e.g., performance-avoidance terminology, form of depression). Second, there is conceptual overlap in some of the moderators tested, especially regarding achievement goal conceptualization (e.g., achievement goal model – dichotomous, trichotomous, 2 × 2 – clearly has some overlap with achievement goal terminology – mastery goal/orientation, mastery-approach goal, etc.). As such, the number of significant moderators may be somewhat misleading, as some of the significant findings may emerge from nonindependent tests. Regardless, it is clear from the present work that there is considerable complexity and nuance underlying the direct relations between achievement goals and internalizing problems documented herein, performance-avoidance goals being the exceptional case.

Limitations, Future Directions, and Implications

Limitations of the present work should be noted; these limitations point to additional avenues for future research. First, we evaluated each achievement goal separately. However, achievement goals are not mutually exclusive and people commonly simultaneously adopt multiple goals (Barron & Harackiewicz, 2001; Pintrich, 2000). Future studies would do well to evaluate how multiple goals (e.g., high mastery-approach and performance-approach goals) are related to anxiety and depression. Second, all studies in our meta-analytic work used self-reported scales of achievement goals, anxiety, and depression; as such, the results might be affected by common method variance. Future studies would do well to use other informants (e.g., teachers’ reports of students’ achievement goals, clinical interviews to assess depression and anxiety) in order to more definitively document the focal relations. Third, there were not sufficient longitudinal data incorporating multiple measurements of both achievement goals and internalizing problems for us to test direction of causality. Future research would do well to attend to this important issue, particularly given the differing theoretical emphases on this matter. Fourth, although our research documented direct relations between achievement goals and internalizing problems, it did not document the mechanisms responsible for these relations. Future research is needed to test purported mediation of the observed links, such as cognitive appraisals (Dykman, 1998), perceived stress (Wang et al., 2021), and rumination (Van Boekel & Martin, 2014). Finally, the present work revealed minimal or no existing research on the following: Mastery-avoidance goals, different types of anxiety (e.g., generalized anxiety, agoraphobia) and depression (e.g., disruptive mood dysregulation disorder), and older adults. These issues are in clear need of future research attention.

The findings from the present research join the growing corpus of findings indicating that mastery-approach goals are beneficial for psychological functioning, whereas performance-avoidance goals are detrimental. The results reveal a new and promising perspective for prevention efforts, with achievement goals as sensible entry points to prevent poor mental health. Achievement goals are modifiable through interventions targeting school environment by, for example, emphasizing students’ personal growth and learning (evoking mastery-approach goals) (see Elliot & Hulleman, 2017 for a review). School-based interventions are needed to address the structural aspects of the educational system by emphasizing the development of abilities, understanding of material, and promoting mistakes as opportunities for growth (Liu et al., 2024; Steare et al., 2024). Accordingly, the take-home message from this research is similar to that of other achievement goal research: Teachers, coaches, employers, and parents would do well to structure their instructions, incentives, and feedback to those under their charge in ways that facilitate and support the pursuit of mastery-approach goals, and discourage and disrupt the pursuit of performance-avoidance goals (Bardach et al., 2020; Korn et al., 2019; Senko, 2016). The data are not yet clear enough for emphatic statements about performance-approach or mastery-avoidance goals. Of course, these recommendations must be made while acknowledging the relatively small number of studies conducted and the relatively modest overall effect sizes.

Concluding Thoughts

A clear and unequivocal conclusion that may be drawn from the present meta-analytic work is that achievement goals and internalizing problems are systematically related to each other. This conclusion is of conceptual and applied importance. Conceptually, it means that competence pursuits are relevant to mental health in general, they are not just relevant to competence-specific (e.g., achievement, intrinsic motivation) or domain-specific (e.g., school, work) outcomes, or affective states (e.g., achievement-relevant emotions). Competence is a basic human need, so it makes sense that competence-based pursuits would be linked to broad health and well-being indices (Elliot et al., 2002; Ryan & Deci, 2019). Further, researchers might consider achievement goals as an explanation for why depression and anxiety are related to individuals’ achievement and adjustment in various settings, including educational, work, and sport settings. In terms of application, it means that greater attention needs to be allocated – in the classroom, the workplace, and the ballfield – to the achievement goal-mental health nexus. Attending to the whole person, not just the individual’s short-term achievements, will likely pay dividends for both long-term accomplishment and overall flourishing and functioning. Accordingly, we believe there is strong reason to sound the call for increased research attention to the relation between achievement goal pursuit and internalizing problems.