How Teachers Shape Teen Achievement: Two Kinds of Self-Belief Carry the Effect

Every teacher knows the moment: a student stares at a difficult problem, and something invisible decides whether they push forward or give up. Psychologists have long argued that this invisible something is self-efficacy, the belief a learner holds about their own capacity to succeed. A new study published in Current Psychology by Zelin Su and colleagues at Universiti Putra Malaysia, together with Mahmoud Danaee of the University of Malaya, now offers one of the most finely grained pictures yet of how the support teachers give in the classroom travels through these beliefs to shape real academic outcomes. Working with 916 junior high school students, the research team found that three distinct flavors of teacher support, learning support, emotional support, and ability support, were each positively linked to achievement, and that two separate dimensions of academic self-efficacy carried part of that effect.

The study’s central innovation lies in its refusal to treat either teacher support or self-efficacy as a single lump. Drawing on the social cognitive tradition founded by Albert Bandura, the researchers split teacher support into three theoretically meaningful components. Learning support refers to the instructional side of teaching: explaining material clearly, giving useful feedback, and helping students master content. Emotional support captures the warmth of the relationship, the sense that a teacher cares, listens, and provides a safe psychological base. Ability support, the third and least studied component, describes the ways teachers signal confidence in a student’s competence, for example by communicating high expectations and affirming that the student is capable of demanding work. Each of these, the team hypothesized, might feed differently into the beliefs students hold about themselves.

On the self-efficacy side, the researchers made an equally consequential distinction. Rather than measuring one global score of academic confidence, they assessed two dimensions separately: perceived learning ability, which reflects a student’s belief in their own competence to understand and perform academic tasks, and perceived learning regulation, which reflects confidence in the capacity to plan, monitor, and control one’s own studying. This second dimension connects directly to decades of work on self-regulated learning, notably Barry Zimmerman’s models of how learners set goals, track their progress, and adjust their strategies. The distinction matters because a student can feel smart yet helpless to manage their time, or feel average in ability yet superbly disciplined; the two beliefs are related but not identical, and they may respond to different kinds of teaching.

To test how these pieces fit together, the team used structural equation modeling, a statistical framework that allows researchers to estimate a network of simultaneous relationships among latent variables, that is, constructs such as emotional support or regulation self-efficacy that cannot be observed directly but are inferred from questionnaire responses. The modeling approach, following the partial least squares tradition described by Hair and colleagues, let the investigators estimate both the direct paths from each form of teacher support to academic achievement and the indirect paths that run through the two self-efficacy dimensions. In such models, a significant indirect path is the statistical signature of mediation: it suggests that the support does not merely accompany achievement but plausibly contributes to it by changing something in the student’s mind.

The results were strikingly consistent across the board. All three forms of perceived teacher support were positively associated with both perceived learning ability and perceived learning regulation. In other words, it did not matter whether a teacher was explaining a concept, offering encouragement, or expressing faith in a student’s potential; every one of these behaviors was linked to stronger beliefs, both about competence and about the capacity for self-directed study. This convergence suggests that students read a broad range of classroom signals as information about themselves, a process consistent with Bandura’s account of how efficacy beliefs are built from social persuasion, vicarious experience, and mastery experiences scaffolded by more capable others.

Just as important, the two efficacy dimensions were each independently associated with academic achievement. Perceived learning ability and perceived learning regulation each contributed their own association with performance, indicating that they are not redundant proxies for a single general confidence factor. A student who believes she can regulate her learning appears to gain something over and above the benefit of believing she is capable, and vice versa. This finding gives empirical weight to a growing movement in educational psychology that treats self-regulatory efficacy as a distinct lever for intervention, one that can be strengthened through explicit training in planning, monitoring, and strategy use, as recent intervention studies such as the college learning-to-learn course evaluated by Wolters and colleagues have explored.

The mediation analysis delivered the study’s headline conclusion: the indirect paths were significant, indicating partial mediation. Teacher support was associated with achievement both directly and indirectly through academic self-efficacy. Partial mediation is a nuanced result. It means the belief pathway explains a meaningful share of the support-achievement link, but not all of it. The direct paths imply that supportive teaching may also improve achievement through channels the model did not measure, such as increased engagement, better classroom behavior, reduced anxiety, or simply more time on task. Earlier meta-analytic work by Tao and colleagues in 2022, which synthesized studies of perceived teacher support, engagement, and achievement, had already pointed toward engagement as one such channel; the new study adds the belief pathway as a parallel route rather than a replacement for it.

For anyone designing educational interventions, the parallel structure of these pathways is the most actionable insight. If teacher support worked on achievement through a single psychological bottleneck, then boosting that one bottleneck would suffice. Instead, the findings suggest that teachers and schools should aim at both targets simultaneously. Programs that build students’ confidence in their competence, for instance through mastery-oriented feedback of the kind John Hattie and Helen Timperley analyzed in their influential work on feedback, address the ability dimension. Programs that explicitly teach study strategies, goal setting, and self-monitoring address the regulation dimension. And the finding that all three forms of teacher support fed both dimensions implies that the everyday repertoire of good teaching, clear instruction, warmth, and communicated high expectations, is not a menu of alternatives but a combined investment in student belief.

The study also carries implications for equity, a theme that runs through the broader literature the authors cite, including OECD analyses of learning disparities and research on immigrant adolescents by Juang, Schachner, and others. If a portion of the socioeconomic and demographic gaps in achievement operates through students’ beliefs about themselves, then classroom environments rich in ability support, the deliberate communication that every student is seen as capable, become a concrete, low-cost equity tool. Beliefs are malleable in a way that family circumstances are not, and the evidence here suggests they are responsive to what teachers say and do daily. The research was approved by the Human Subjects Research Ethics Committee of Universiti Putra Malaysia, with parental consent obtained for all participants, and the authors report no funding or competing interests.

As with any cross-sectional design, caution is warranted in reading the arrows. The model specifies support leading to beliefs leading to achievement, and the theoretical grounding in Bandura’s social cognitive theory and Vygotsky’s account of scaffolded development supports that direction, but longitudinal and experimental data would be needed to confirm the causal ordering. Self-reported perceptions of teacher support also reflect the student’s viewpoint, which is itself psychologically meaningful but may differ from externally observed teaching quality. Still, the study’s fine-grained architecture, three supports, two efficacies, one achievement outcome, mapped with formal structural modeling, gives researchers and practitioners a sharper map than the blunt question of whether teachers matter. The answer now on offer is more interesting: teachers matter partly by building two different kinds of belief, and those beliefs, competence and control, each carry their own share of the weight.

Subject of Research: The relationship between differentiated teacher support, academic self-efficacy, and academic achievement in junior high school students

Article Title: Differentiated teacher support and academic achievement: parallel mediation through ability and learning regulation self-efficacy

Article References: Su, Z., Azam, N. D., Arshat, Z., & Danaee, M. (2026). Differentiated teacher support and academic achievement: parallel mediation through ability and learning regulation self-efficacy. Current Psychology, 45(20), Article 1613. https://doi.org/10.1007/s12144-026-10184-8

Image Credits: AI Generated

DOI: 10.1007/s12144-026-10184-8

Keywords: teacher support, academic self-efficacy, learning regulation, perceived competence, academic achievement, structural equation modeling, educational psychology, self-regulated learning, adolescents, classroom interactions, mediation analysis, student motivation