Search

The present study uses doubly latent models to estimate the effect of average mathematics achievement at the class level on students' subsequent mathematics achievement (the "Peer Spillover Effect") and mathematics self-concept (the "Big-Fish-Little-Pond-Effect; BFLPE"), controlling for individual differences in prior mathematics achievement. Our data, consisting of 13-year-old students from Canada, the USA, and New Zealand, come from a unique cross-national database with a longitudinal design at the student level: the Second International Mathematics Study (SIMS80). This historical survey was administered by IEA in the 1980s and highly influenced the development of educational policies in the following decades. We replicate a widely cited study based on SIMS80, interrogating the validity of its findings of a positive peer spillover effect. When we adjust for measurement error, using doubly latent models, we observe that originally positive peer spillover effects become less positive or disappear altogether. On the contrary, negative BFLPEs become more negative and remain statistically significant throughout. Our study is the only cross-national study to have evaluated both the BFLPE and the peer spillover effect with controls for a true measure of prior achievement -- and the only study to test the peer spillover effect cross-nationally using doubly latent models. Our findings question the empirical results of past and current research evaluating school- and class-level compositional effects based on sub-optimal models that fail to control for measurement error.

School victimization issues remain largely unresolved due to over-reliance on unidimensional conceptions of victimization and data from a few developed OECD countries. Thus, support for cross-national generalizability over multiple victimization components (relational, verbal, and physical) is weak. Our substantive-methodological synergy tests the cross-national generalizability of a three-component model (594,196 fifteen-year-olds; nationally-representative samples from 77 countries) compared to competing (unidimensional and two-component) victimization models. We demonstrate the superior explanatory power of the three-component model--goodness-of-fit, component differentiation, and discriminant validity of the three components concerning gender differences, paradoxical anti-bullying attitudes (the Pro-Bully Paradox) whereby victims are more supportive of bullies than of other victims, and multiple indicators of well-being. For example, gender differences varied significantly across the three components, and all 13 well-being indicators were more strongly related to verbal and particularly relational victimization than physical victimization. Collapsing the three components into one or two components undermined discriminant validity. Cross-nationally, systematic differences emerged across the three victimization components regarding country-level means, gender differences, national development, and cultural values. These findings across countries support a tripartite model in which the three components of victimization--relational, verbal, and physical--relate differently to key outcomes. Thus, these findings advance victimization theory and have implications for policy, practice, and intervention. We also discuss directions for further research: the need for simultaneous evaluation of multiple, parallel components of victimization and bullying, theoretical definitions of bullying and victimization and their implications for measurement, conceptual bases of global victimization indices, cyberbullying, anti-bullying policies, and capitalizing on anti-bullying attitudes.

Conventional wisdom suggests that parents' educational expectations (how far they expect their children to go) and aspirations (how far they want their children to go) positively impact academic outcomes and benefits from attending high-ability schools. However, here we juxtapose the following: largely positive effects of educational expectations (of parents, teachers, and students); small, mixed effects of parent aspirations; largely adverse effects of parental aspiration-expectation gaps; and negative effects of school-average achievement on expectations, aspirations, and subsequent outcomes. We used a large, nationally representative longitudinal sample (16,197 Year-10 students from 751 US high schools).Controlling background (achievement, SES, gender, age, ethnicity, academic track, and a composite risk factor), Year 10 educational expectations of teachers and parents had consistently positive effects on the following: student expectations in Years 10 and 12, Year 10 academic self-concept, final high-school grade-point-averages, and long-term outcomes at age 26 (educational attainment, educational and occupational expectations). Effects of parent aspirations on these outcomes were predominantly small and mixed in direction. However, the aspiration-expectation gap negatively predicted all these outcomes. Contrary to our proposed Goldilocks Effect (not too much, not too little, but just right), non-linear effects of expectations and aspirations were small and largely non-significant.Parent, teacher, student expectations, and parent aspirations were all negatively predicted by school-average achievement (a big-fish-little-pond effect). However, these adverse effects of school-average achievement were larger for parents and particularly teachers than students. Furthermore, these expectations and aspirations partly mediated the adverse impacts of school-average achievement on subsequent grade-point-average and age-26 outcomes.

Purpose: Past research shows the Big-Fish-Little-Pond Effect (BFLPE; negative effect of school-average achievement on student-level academic self-concept) to generalize across countries. However, such evidence is largely limited to math and science. Given that reading self-concept is highly differentiated from math and science self-concepts and plays an important role in predicting educational outcomes, it is essential to examine the universality of the BFLPE and its underlying social-comparison process within this domain. Method: We assess the cross-national generalizability of the BFLPE for 15-year-olds' reading self-concept using Programme for International Students Assessment 2018 (533,165 students, 72 countries). To demonstrate that the BFLPE operates with a relative--rather than absolute--frame of reference for comparison, we juxtapose difficulty experienced with reading in general (self-concept perceived difficulty; relative frame of reference), with difficulty experienced with reading specifically during the PISA test (PISA test difficulty; absolute frame of reference). Results: Our findings show that the BFLPE for both the reading self-concept perceived competence and difficulty subscales was robust across countries. Further, the BFLPE was strong for self-concept subscales, but very weak for the PISA test difficulty scale. Conclusions: Our findings extend support for the generalizability of the BFLPE to reading self-concept and highlight the role of social comparison processes underlying this effect.

Children who are excluded from school may end up being excluded from other social institutions. Yet little research has considered whether low school belonging is a risk factor for not being in education, employment, or training after graduation. Using two longitudinal cohorts from Australia (N = 14,082; 51% Boys), we explored this relationship. Controlling for a range of individual and school-level covariates, we found that low school belonging at age 15 is a consistent and practically significant predictor of not in education, employment, or training (NEET) status at ages 16 to 20. We conclude that this relationship is unlikely to be the product of low school belonging lowering the chances of students graduating high school. Rather, low school belonging had a unique association with NEET beyond graduation. Given that NEET represents a range of vulnerabilities, educational policy and practice must find ways for schools to create opportunities for all students to feel included, valued, and accepted.

Improving educational outcomes for Indigenous Australian students is a key strategy to helping Indigenous people reach their full potential. This has resulted in well-intentioned efforts by Australian educators and governments to ensure Indigenous children have positive school experiences. However, Indigenous students still lag behind their non-Indigenous counterparts in educational outcomes. This is particularly so for Indigenous students living in rural and remote parts of Australia where educational opportunities are limited, especially in high school. One solution to this problem has been to enrol these students in boarding schools in urban and metropolitan centres. While research on the success of boarding schools for Indigenous students is scarce, what little that does exist is not encouraging. The focus of this research was to examine the effects of boarding for Indigenous (n = 11) and non-Indigenous students' (n = 158) wellbeing (N = 1423) in two large private boys' schools. Participating students aged 12-18 years old completed a survey measuring wellbeing constructs on two occasions, 12 months apart. Non-Indigenous boys were generally higher in wellbeing compared with Indigenous boys. There was also evidence of improved social wellbeing beyond that of non-Indigenous boarders over time. Overall, while evidence of merit was weak, boarding schools may benefit their Indigenous students' development in social wellbeing.

Academic self-concept and achievement have been found to be reciprocally related across time. However, existing research has focused on self-concept and achievement scores that have been averaged over long time-periods. For the first time, the present study examined intraindividual (within-person) relations between momentary (state) self-concept and lesson-specific perceived achievement (i.e., self-reported comprehension) in students' everyday school life in real time using intensive longitudinal data. We conducted an experience-sampling (e-diary) study with 372 German secondary school students in Grades 9 and 10 over a period of 3 weeks after each mathematics lesson. Multilevel confirmatory factor analyses confirmed a two-factor between-level and within-level structure of the state measures. We used dynamic structural equation modeling to specify a multilevel first-order vector autoregressive model to examine the dynamic relations between self-concept and perceived achievement. We found significant reciprocal effects between academic self-concept and perceived achievement on a lesson-to-lesson basis. Further, we found that these relations were independent of students' gender, reasoning ability, or mathematics grades. We discuss implications for methodology, theory, and practice in self-concept research and educational psychology more generally.

We juxtapose (positive and negative) compositional effects of school-average achievement and school-average socioeconomic status (SES) on students' academic self-concept (ASC), final high-school grade-point-average (GPA), and long-term outcomes at age 26 (educational attainment and educational and occupational expectations). We used doubly-latent multilevel compositional models with a large, nationally representative longitudinal sample (16,197 Year-10 students from 751 US high schools), controlling background variables (gender, age, ethnicity, academic track, and a composite risk factor). At the individual-student level, the effects of achievement, SES, ASC, and GPA on long-term outcomes were consistently positive. However, mostly consistent with a priori theoretical predictions, (1) the compositional effects of school-average achievement on ASC, GPA, and educational and occupational expectations were significantly negative (although non-significant for final attainment); (2) the compositional effects of school-average SES on ASC, educational attainment, and educational and occupational expectations were significantly positive (but nonsignificant for GPA); and (3) the compositional effects on long-term outcomes were partly mediated by ASC and particularly by GPA. These findings demonstrate that the positive effects of school-average SES are distinguishable from the adverse effects of school-average achievement. We discuss how these findings extend Göllner et al.'s (Psychological Science 29:1785-1796, 2018) highly controversial conclusion regarding the benefits of schools with high school-average SES but low school-average achievement. We also relate our research to Luthar et al.'s (American Psychologist 75:983-995, 2020) findings of adverse mental health problems associated with attending high-achieving schools. Our results have important implications not only for theory and methodology but also for parents' selection of schools for their children and policy regarding the structure of schools (a substantive-methodological synergy).

The proliferation of information and divergent viewpoints in the 21st century requires an educated citizenry with the ability to critically evaluate information and make informed decisions. To meet this demand, adaptive epistemic understandings and beliefs about the nature of knowledge are needed, such as believing that scientific knowledge is evolving (development of knowledge) and needs to be justified through experimentation (justification of knowledge). Our study is the first to use nationally representative samples from 72 countries/regions (PISA 2015 database; N = 514,119 students) to examine how scientific epistemic beliefs about development and justification of knowledge in science are associated with students' science motivation, achievement, and career aspirations in the STEM fields, as well as the cross-national generalizability of these relations. Results showed that: (a) students who had more adaptive beliefs about knowledge being changeable and stemming from experimentation were likely to have high science self-efficacy, utility value, and particularly high intrinsic value; (b) epistemic beliefs were more strongly linked to science achievement than were motivational constructs; (c) the positive relation between epistemic beliefs and STEM-related career aspirations was largely explained by motivation and achievement; (d) the pattern of results generalized well across societies. Our findings suggest that epistemic beliefs are substantially positively associated with adolescents' science learning, implying that developing effective interventions that focus on development and justification of knowledge would be fruitful for promoting science educational outcomes.

Using the dual process theoretical framework (Brandtstädter & Rothermund, 2002), we examined whether the tendency to pursue goals tenaciously (TEN), in conjunction with the tendency to flexibly adjust one's preferences (FLEX), would be beneficial or detrimental for high school students' self-reported life satisfaction and achievement on tests of academic competencies. We used data from a sample of German high school students (N = 10,957) from the National Educational Panel Study (NEPS), and a combination of machine-learning-based psychometrics and latent structural equation modeling. The results showed that the interaction of TEN and FLEX was negatively associated with both outcomes of life satisfaction and test performance. Having high levels of both TEN and FLEX appear to have no benefit for students' life satisfaction and seem to be counterproductive for their achievement on competencies tests. The findings are consistent with the theory that having high levels of both TEN and FLEX would represent a regulatory dilemma in which the two tendencies work against each other. The results suggest the need for targeted interventions that build on students' personal strengths in tenacity or flexibility to minimize their regulatory dilemma of tenaciously pursuing goals while flexibly adapting them at the same time.

Principal strain and burnout is a major issue in desperate need of further investigation and solutions. Deepening our understanding of emotional exhaustion, the central dimension of burnout, would greatly further this pursuit. Using a large, longitudinal, representative sample of Australian school principals, the present study decomposed emotional exhaustion into occasion-specific state, enduring autoregressive, and stable trait components using the STARTS (stable trait, autoregressive trait, and state) model. The results showed evidence for variance in all three components, indicating that principals' emotional exhaustion is approximately evenly split between the enduring autoregressive component and stable trait component, with slightly less variance being observed for the occasion-specific state. Heterogeneity in this profile was mainly associated with individual characteristics of the principals themselves (i.e., experience and gender) rather than characteristics of the job (school sector and level). The results revealed that less experienced and male principals have more malleable (enduring autoregressive and state-like) emotional exhaustion, while more experienced and female principals have more trait-like emotional exhaustion. This emphasizes a likely development of emotional exhaustion from acute to chronic under persistent exposure to burnout-inducing situations, with additional evidence for a possible dispositional tendency toward emotional exhaustion. Thus, measures to tackle emotional exhaustion need to be based on the type of emotional exhaustion the principal is experiencing and ideally include elements that target both the situational/contextual and the individual factors that cause emotional exhaustion in school principals.

How do peer groups influence academic self-concept formation? We evaluate developmental issues in the big-fish-little-pond effect (BFLPE; negative effects of class-average achievement on math self-concept [MSC]) and its generalizability to peer-group-average achievement (1,017 primary-school students tested in Years 4 and 6, 46 classes, 130 peer groups). The effects of peer-group-average and class-average achievement on MSC were both negative when we considered these two contextual effects separately. However, the effect of peer-group-average became nonsignificant in models with both contextual effects; the negative effect of class-average achievement was relatively unaffected. Results for peer-group-average achievement contradict predictions based on local dominance theory (that the BFLPE should be more negative for peer-group-average achievement than the more local frame of reference, a contrast effect) and social comparison choice studies (that peer-group-average achievement effects should be positive, an assimilation effect). Unsurprisingly, we found BFLPEs based on class-average achievement and gender differences favoring boys in both Years 4 and 6. However, consistent with theories of the cognitive development of social comparison and gender socialization/intensification processes, we also found negative effects of class-average and gender differences favoring boys on change in MSC (MSC in Year 6 controlling for MSC in Year 4) over this critical late-childhood period. Our results support the robustness of the BFLPE based on class-average achievement and developmental processes underpinning it, but do not support the posited effects of peer-group-average achievement.

Equally able students have lower academic self-concept in high achieving schools or classes, a phenomenon known as the big fish little pond effect (BFLPE). The class (more so than the school) has been shown to be the pivotal frame-of-reference for academic self-concept formation--a local dominance effect. However, many school systems worldwide employ forms of course-by-course tracking, thus exposing students to multiple class environments. Due to the high correlation between multiple student environments, the frame-of-reference used for academic self-concept formation in course-by-course tracked systems is unclear to date. We addressed this unresolved issue by using data from a comprehensive survey that measured the entire population of Austrian eighth-grade students in the domain of mathematics in 2012. General secondary school students were in the core subjects (i.e., mathematics, German, and English) grouped according to ability, whereas regular class composition was the same in all other subjects. Using cross-classified multilevel models, we regressed math self-concept on average math achievement of students' school, math class, and regular class. Consistent with the local dominance effect, we found the BFLPE on the school level to be weak after controlling for the class levels. We found a stronger BFLPE on the regular class level and the strongest BFLPE on the math class level. Our study demonstrates the importance of multiple class environments as frames-of-reference for academic self-concept formation.

The big-fish-little-pond effect (BFLPE), the negative effect of school-/class-average achievement on academic self-concept, is one of educational psychology's most universal findings. However, critiques of this research have proposed moderators based on achievement motivation theories. Nevertheless, because these motivational theories are not sufficiently well-developed to provide unambiguous predictions concerning moderation of the BFLPE and underlying social comparison processes, we developed a Theory-Integrating Approach; bringing together a panel of experts, independently making theoretical predictions, revising the predictions over several rounds based on independent feedback from the other experts, and a summary of results. We pit a priori hypotheses derived from achievement motivation theories against the more parsimonious a priori prediction that there is no moderation based on previous BFLPE empirical research and Darwinian-economic theory (N = 1,925 Hong Kong students, 47 classes, M age = 12 years). Consistent with both BFLPE research and Darwinian perspectives, but in contrast to achievement motivation theory predictions, the highly significant BFLPE was not moderated by any of the following: prior achievement, expectancy-value theory variables, achievement goals, implicit theories of ability, self-regulated learning strategies, and social interdependence theory measures. Although we cannot "prove" that there are no student-level moderators of the BFLPE, our synthesis of social comparison posited in the BFLPE theory and an evolutionary perspective support BFLPE's generalizability. We propose further integration of our Theory-Integrating Approach with traditional Delphi methods, combining quantitative and qualitative approaches to develop a priori theoretical predictions and identify limitations in existing theory as an alternative form of systematic review.

Paradoxically, school leaders as a group report high levels of burnout but also high job satisfaction and passion for their work. School principals are passionate about their job, but this passion can be a double-edged sword leading to good (job satisfaction) and bad (burnout) outcomes. We extend the dualistic model of passion (DMP) in a study of changes over a 1-year period (3,996 school leaders) in three components of work passion (general, harmonious, and obsessive), burnout, and job satisfaction. We begin by placing the DMP into a broader framework of passion research and how passion is related to other well-being constructs. We then develop the rationale for integrating within the same study the obsessive passion and harmonious passion constructs used in DMP studies and general passion used in most other passion studies, but also the inappropriateness of the exclusion strategy based on general passion in DMP research. We found that Time1 harmonious passion was related to job satisfaction (positively) and burnout (negatively), while the pattern of relations was the opposite of obsessive passion. However, over time, harmonious passion and obsessive passion were moderated by general passion. The protective effect of harmonious passion on burnout was smaller when general passion was high, while undesirable obsessive passion effects were only evident when general passion was high. School principals' work passion may help them avoid burnout and maintain high job satisfaction, but only when manifested harmoniously. We describe how future studies of longitudinal passion can be integrated into the broader job demands-resources model.

Understanding how children's broader context influences their development is critical if we are to develop policies that help them flourish. Combining sociological, economic, and psychological literature, we argue that ability stratification--the degree to which children of similar levels of ability are schooled together--influences a child's academic self-concept. This is because countries with more ability stratification should have larger Big-Fish-Little-Pond Effects (the negative effect of school average achievement on academic self-concept). We used four cycles of the Trends in International Math and Science Study to test the hypothesis that more country-level ability stratification is associated with larger country-level Big-Fish-Little-Pond Effects for math self-concept. Findings strongly support this hypothesis. Our findings have implications for school system design and policy.

Drawing on expectancy-value theory, the present study examined the unique contributions of the four major value beliefs and self-concept on achievement, self-reported effort, and teacher-rated behavioral engagement in mathematics. In particular, we examined the multiplicative effects of self-concept and task values on educational outcomes using the latent moderated structural equation approach. Participants were 1,868 German ninth-grade students. The data analyses relied on a higher-order structure of value beliefs, which is suited to parsing the differential patterns of predictive relations for different value beliefs. The findings revealed that (a) self-concept was more predictive of achievement, whereas value beliefs were more predictive of self-rated effort; (b) self-concept and value beliefs emerged as equally important predictors of teacher-reported engagement; (c) among the four value beliefs, achievement was more associated with low cost, whereas effort was more associated with attainment value; and (d) latent interactions between self-concept and value beliefs predicted the three outcomes synergistically.

Job satisfaction is important for employee well-being and retention, which are both crucial for high-strained occupations such as teachers and school principals. We investigated the structure and relationship of teachers' job satisfaction (N = 142,280) and principals' job satisfaction (N = 8,869). Results of structural equation modeling using multilevel bifactor models provide strong evidence for the validity of a model including both specific (satisfaction with the profession vs. the working environment) and general job satisfaction factors. This differential conception of job satisfaction was further tested by including the student variables disciplinary climate and student achievement from linked TALIS 2013-PISA 2012 data. Only teacher job satisfaction factors (general and working environment) were related to the disciplinary climate perceived by students, whereas the working environment job satisfaction factor was related to student achievement for both teachers and principals. Controlling for socioeconomic status reduced these estimates, but did not change the pattern of results.

We evaluated STEM (science, technology, engineering, and mathematics) coursework selection by women and men (representative longitudinal sample, 10,370 Australians) in senior high school and university, controlling achievement and expectancy-value variables. A near-zero total effect of gender on high school STEM enrollment reflected pathways favoring boys through achievement and expectancy-value variables, but a counteracting direct effect of gender favoring girls. In contrast, subsequent university STEM enrollment favored boys. In both high school and university, enrollments favored girls in life sciences and boys in physical sciences, but at university there was a leaky pipeline in which girls who qualified to pursue physical sciences opted for non-STEM subjects. Qualitative analysis not only supported quantitative results but also highlighted alternative mechanisms of STEM engagement/disengagement, and mostly supported gender similarities rather than differences.

Drawing on Eccles' expectancy-value model of achievement-related choices, we examined how work values predict individual and gender differences in sciences, technology, engineering, and math (STEM) participations in early adulthood (ages of 25/27, 6 or 8 years after postsecondary school), controlling for subjective task values attached to academic subjects in late adolescence (11th grade, age 18). The study examined 1,259 Finnish participants using a person-oriented approach. Results showed that: (a) we could identify four profile groups based on five core work values (society, family, monetary, career prospects, and working with people); (b) work-value profiles predicted young adults actual STEM participation in two fields: math-intensive and life science occupations above and beyond academic task values (e.g., math/science) and background information; (c) work-value profiles also differentiate between those who entered support- vs. professional-level STEM jobs; and (d) gender differences in work value profiles partially explained the differential representation of women across STEM sub-disciplines and the overall underrepresentation of women in STEM fields.

This study extends the classic constructive dialogue/debate between self-concept and self-efficacy researchers (Marsh, Roche, Pajares, & Miller, 1997) regarding the distinctions between these 2 constructs. The study is a substantive-methodological synergy, bringing together new substantive, theoretical, and statistical models and developing new tests of the classic jingle-jangle fallacy. We demonstrate that in a representative sample of 3,350 students from math classes in 43 German schools, generalized math self-efficacy and math outcome expectancies were indistinguishable from math self-concept, but were distinct from test-related and functional measures of self-efficacy. This is consistent with the jingle-jangle fallacies that are proposed. On the basis of pretest variables, we demonstrate negative frame-of-reference effects in social (big-fish-little-pond effect) and dimensional (internal/external frame-of-reference effect) comparisons for three self-concept-like constructs in each of the first 4 years of secondary school. In contrast, none of the frame-of-reference effects were significantly negative for either of the two self-efficacy-like constructs in any of the 4 years of testing. After controlling for pretest variables, each of the 3 self-concept-like constructs (math self-concept, outcome expectancy, and generalized math self-efficacy) in each of the 4 years of secondary school was more strongly related to posttest outcomes (school grades, test scores, future aspirations) than were the corresponding 2 self-efficacy-like factors. Extending discussion by Marsh et al. (1997), we clarify distinctions between self-efficacy and self-concept; the role of evaluation, worthiness, and outcome expectancy in self-efficacy measures; and complications in generalized and global measures of self-efficacy.

School-average achievement is often reported to have positive effects on individual achievement (peer spillover effect). However, it is well established that school-average achievement has negative effects on academic self-concept (big-fish-little-pond effect [BFLPE]) and that academic self-concept and achievement are positively correlated and mutually reinforcing (reciprocal effects model). We resolve this theoretical paradox based on a large, longitudinal sample (N = 14,985 U.S. children) and improved methodology. More appropriate multilevel modeling that controls for phantom effects (due to measurement error and preexisting differences) makes the BFLPE even more negative, but turns the peer spillover effect from positive to slightly below zero. Thus, attending a high-achieving school has negative effects on academic self-concept and a nonpositive effect on achievement. The results question previous studies and meta-analyses showing a positive peer spillover effect that do not control for phantom effects, along with previous policy and school selection decisions based on this research. Educational Impact and Implications Statement: Counter to a widely held belief that being in a high-achieving school has a positive effect on student's achievement (peer spillover effect), the present findings suggest that this effect is actually slightly negative. When using stronger, more appropriate statistical methodology, the apparent peer spillover effect disappeared, suggesting that positive effects are a phantom. Furthermore, the negative effect of school-average achievement on academic self-concept (big-fish-little-pond effect) turned out to be even more negative when using more appropriate methodology. Thus, the findings indicate that attending a high-achieving school has a negative effect on self-concept and no positive effect on achievement. These results call into question prior research that did not control for phantom effects, and challenge policy and practice decisions that promote selective schooling.

Research suggests that a country does not need inequity to have high performance. However, such research has potentially suffered from confounders present in between-country comparative research (e.g., latent cultural differences). Likewise, relatively little consideration has been given to whether the situation may be different for high- or low-performing students. Using five cycles of the Programme for International Student Assessment (PISA) database, the current research explores within-country trajectories in achievement and inequality measures to test the hypothesis of an excellence/equity tradeoff in academic performance. We found negative relations between performance and inequality that are robust and of statistical and practical significance. Follow-up analysis suggests a focus on low and average performers may be critical to successful policy interventions.

As the online world plays an increasing role in young peoples' lives, research on compulsive internet use (CIU) is receiving growing attention. Given the social richness of the online world, there is a need to better understand how CIU influences adolescents' social support and vice versa. Drawing on ecological systems theory, we examined the longitudinal links between adolescents' CIU and perceived social support from three sources (parents, teachers, and friends) across 4 critical years of adolescence (Grades 8–11). Using random intercept cross-lagged modeling, we found that CIU consistently preceded reduced social support from teachers, whereas social support from parents preceded increases in CIU over time. We discuss the implications of our findings for parents and schools seeking to support young people experiencing CIU. [ABSTRACT FROM AUTHOR]