Does class size affect student achievement?
The sources here don't directly answer whether smaller classes raise achievement; class size only appears as a non-significant moderator in one math project-based learning meta-analysis.
Covers: This page examines the evidence on whether smaller classes improve student achievement, focusing on experimental and quasi-experimental studies in primary and secondary education. It does not cover higher education, online learning, or the effects of class size on non-academic outcomes such as student well-being.
3 free full reads left this month. Join or upgrade
The short answer
Interpretation AI-prepared starting mapThe sources available here do not directly answer whether smaller classes raise achievement. The closest direct signal comes from a three-level meta-analysis of project-based learning in mathematics: class size was tested as a moderator and showed no significant effect on outcomes, alongside group size, educational level, mathematics domain, publication type and technology use. That meta-analysis found a significant overall effect for project-based learning (Cohen's d = 0.529, 95% CI [0.38, 0.68]) but with signs of publication bias (trim-and-fill corrected estimate = 0.275), so its moderator results should be read cautiously. Other sources address different levers on achievement: student characteristics (cognitive abilities, positive personality traits, positive emotions) show a medium-level overall relationship with achievement, strongest in kindergarten and moderate at elementary and secondary levels, and perceived teacher support in reading works indirectly through reading self-concept and reading engagement rather than directly.123
- Evidence 11
- Interpretation 6
In brief
The sources here do not directly test whether smaller classes raise achievement; class size appears only as a non-significant moderator in one meta-analysis of mathematics project-based learning.1
InterpretationProject-based learning in mathematics showed a significant overall effect (d = 0.529), but publication-bias correction reduced it to 0.275, so the authors call the support moderate rather than robust.1
Evidence-backedStudent characteristics show a medium-level overall relationship with achievement, strongest in kindergarten and moderate at elementary and secondary levels, with cognitive abilities correlating most strongly.2
Evidence-backedPerceived teacher support affected reading literacy indirectly through reading self-concept and reading engagement, not directly, in a study of 12,058 Chinese students.3
Evidence-backed
At a glance
The picture in numbers
Live · updated just now
12,058 students
29 meta-analyses
The evidence behind it
3 sources- Reviews of many studies2
- Other studies and data1
Published in 2026
| Source | Kind | Year |
|---|---|---|
| Can project-based learning improve students' mathematics learning outcomes: a three-level meta-analysis. | Reviews of many studies | 2026 |
| Thinking and feeling matter: a second-order meta-analysis of student characteristics across developmental domains and their effect on academic outcomes. | Reviews of many studies | 2026 |
| The Impact of Perceived Teacher Support on Reading Literacy: The Mediating Roles of Reading Self-Concept and Reading Engagement. | Other studies and data | 2026 |
The community around it
- Contributions
- 0
- People
- 0
- Following
- 0
Nobody has added anything yet. Experience, evidence or a different view would show up here.
What it means for you
Which fits you?
Pick the situation closest to yours. Each answer says what it rests on.
If you are weighing smaller classes against other changes to raise achievement
the evidence here cannot confirm a class-size effect; the one synthesis that tested class size found it a non-significant moderator, while instructional approach, student characteristics and teacher support showed measurable associations with achievement.123
InterpretationIf you are considering project-based learning in mathematics
the pooled effect was significant (d = 0.529) but corrected downward to 0.275 for publication bias, and effects were strongest over 1 to 8 weeks, so treat the approach as moderately supported rather than proven.1
Evidence-backedIf you teach in early years or kindergarten
student characteristics showed their strongest correlation with achievement at kindergarten level, with moderate associations at elementary and secondary levels.2
Evidence-backedIf you want to support reading achievement through teacher support
the evidence points to an indirect route: teacher support was linked to reading self-concept, which was linked to greater reading engagement, which in turn related to reading literacy, rather than a direct effect.3
Evidence-backedIf you are reading findings from Western education systems and applying them elsewhere
the reading study notes that most prior work was Western and that extrapolating to Eastern contexts such as China is difficult, and the project-based learning synthesis found cultural background to be a significant moderator.13
Evidence-backedThe full story · 2 chapters
01
What the evidence shows about class size
AI summary:Class size was a non-significant moderator in a math project-based learning meta-analysis, while other sources cover student characteristics and teacher support.
Evidence-backed: In a three-level meta-analysis of project-based learning in mathematics, class size was examined as a potential moderator of learning outcomes and was not a significant one. The same analysis found no significant moderators for educational level, group size, mathematics knowledge domain, publication type, or number of technology types used, which the authors read as broad generalizability of project-based learning across mathematics teaching contexts. The overall effect of project-based learning was significant (Cohen's d = 0.529, 95% CI [0.38, 0.68]), with the strongest effects reported within a 1-to-8-week timeframe.1
Evidence-backed: That same meta-analysis cautions against taking its estimates at face value: publication-bias analyses suggested potential overestimation, with a trim-and-fill corrected estimate of 0.275 against the uncorrected 0.529, and the authors describe the support for applying mathematics project-based learning as moderate rather than robust. Non-cognitive abilities showed a numerically larger effect than cognitive abilities, but the difference was not statistically significant (p = 0.083).1
Evidence-backed: A second-order meta-analysis of 29 first-order meta-analyses (71 effect sizes) found a medium-level overall relationship between student characteristics and academic achievement. Effect sizes varied by educational level, with the strongest correlations in kindergarten and moderate associations at elementary and secondary levels. Among domains, cognitive abilities correlated most strongly with achievement, followed by positive personality traits and positive emotions, both showing moderate effects.2
Evidence-backed: A study of 12,058 Chinese students using PISA 2018 data found that perceived teacher support did not directly affect reading literacy, but did affect it indirectly through reading self-concept and reading engagement, including a chain mediation effect. The authors note that most prior work on these links was conducted in Western cultures, which limits extrapolation to Eastern contexts such as China.3
In your experience, how much does class size affect student learning?
Your individual response is private. Only totals are shown.
02
How to read these findings together
AI summary:The class-size question stays open here, and effects across these sources depend on context and mechanism rather than one structural change.
Interpretation: The evidence gathered here points to instructional approach, student characteristics and the relational side of teaching as measurable influences on achievement, while class size appears only as a moderator that did not reach significance in one mathematics project-based learning synthesis. That is not the same as showing class size is irrelevant: a non-significant moderator in one meta-analysis of a different intervention cannot settle a question usually studied through class-size reduction experiments. Readers should treat the class-size question as open on this page until direct experimental or quasi-experimental evidence is added.123
Interpretation: The pattern across the sources is that effects depend on context and mechanism. Project-based learning effects varied with cultural background and intervention duration, and were strongest over 1 to 8 weeks. Student-characteristic effects were strongest in kindergarten and weaker at later levels. Teacher support operated through self-concept and engagement rather than directly. Any claim that a single structural change such as smaller classes will raise achievement sits awkwardly with this context-dependence.123
Ask this Sylo
Still wondering about something?
Answers come only from this page's reviewed material, with citations, and say plainly when the page doesn't cover it yet.
Behind this page
Who's adding to it, where it comes from, how it changed and what would make it better. Always open to everyone.
Discussion
Sources
Numbers match the citations in the article. A working link isn't proof that a page supports a claim; check the quoted passage and date.
- 1Can project-based learning improve students' mathematics learning outcomes: a three-level meta-analysis.Frontiers in psychology (Wang et al.)Published Sep 4, 2026Checked Oct 4, 2026
“The results indicated a significant overall effect size for PBL (Cohen's d = 0.529, 95% CI [0.38, 0.68]), but publication bias analyses suggested potential overestimation (trim-and-fill corrected estimate = 0.275); therefore, findings should be interpreted cautiously. Moderator analyses revealed that cultural background and intervention duration significantly influenced outcomes. Notably, PBL demonstrated strong effectiveness within a 1-to-8-week timeframe. Non-cognitive abilities showed a numerically larger effect than cognitive abilities, but the difference was not statistically significant (p = 0.083). No significant moderators were found for educational level, class size, group size, mathematics knowledge domain, publication type, or the number of technology types used, suggesting that PBL is broadly generalizable across various mathematics teaching contexts. Given the presence of publication bias, these results provide moderate rather than robust support for the localized application of mathematics PBL.Systematic review registrationhttps://www.crd.york.ac.uk/PROSPERO/view/CRD420251105265, identifier CRD420251105265.”
- 2Thinking and feeling matter: a second-order meta-analysis of student characteristics across developmental domains and their effect on academic outcomes.BMC psychology (Erdem et al.)Published Jun 20, 2026Checked Oct 4, 2026
“Student characteristics are widely recognized as significant correlates of academic achievement. Among these, attributes related to developmental domains-such as cognitive abilities, psychomotor skills, positive emotions, and positive personality traits-have been the focus of numerous meta-analyses. This study employed a second-order meta-analytic approach to synthesize findings from 29 first-order meta-analyses selected based on predefined quality criteria using the R-AMSTAR checklist, yielding a total of 71 effect sizes. Results revealed a medium-level overall relationship between student characteristics and academic achievement. Effect sizes varied significantly by educational level, with the strongest correlations observed in kindergarten, and moderate associations found at the elementary and secondary levels. Among the domains, cognitive abilities demonstrated the highest correlation with academic achievement, followed by positive personality traits and positive emotions, both showing moderate effects. The findings are discussed in light of existing literature, and practical implications for educational policy and practice are presented.”
- 3The Impact of Perceived Teacher Support on Reading Literacy: The Mediating Roles of Reading Self-Concept and Reading Engagement.Behavioral sciences (Basel, Switzerland) (Wang et al.)Published Aug 24, 2026Checked Oct 4, 2026
“To date, limited research has systematically explored the interconnections between perceived teacher support, self-concept, engagement, and academic achievement specifically regarding the reading domain. Most previous studies have been conducted in Western cultures, making it difficult to extrapolate the findings to Eastern cultures, such as China. Therefore, this study investigated the relationship between perceived teacher support, reading self-concept, reading engagement, and reading literacy in a Chinese educational context. A structural equation modeling analysis was conducted using 12,058 Chinese students from the Programme for International Student Assessment 2018 data. The results revealed that perceived teacher support did not directly affect students' reading literacy. However, it could indirectly affect reading literacy through the independent mediation effects of reading self-concept and reading engagement as well as their chain mediation effect. Hence, the results suggest that perceived teacher support can increase students' sense of reading self-concept, resulting in greater reading engagement, which may lead to improved reading achievement.”
How it changed
Published 1 time since Oct 4, 2026.
- Version 2Oct 4, 2026Live now
AI-prepared Starting Map from live research.
- First published version.
Help improve it
The brief is open about what's uncertain. These are the specific gaps that new material would fill.
“How to read these findings together” has no evidence or firsthand experience yet
It's a synthesis for now. Evidence or experience would show whether it holds.
Open questions
What do randomized or quasi-experimental class-size reduction studies in primary and secondary schools find for achievement, and how large are the effects?
No answers yet
If class size matters, does it work through more individual attention, changed teaching practice, or something else?
No answers yet
Are there class-size thresholds below which achievement gains appear, and do they differ by subject, age or student background?
No answers yet
How does class size compare with instructional approach, student characteristics and teacher support as a lever on achievement?
No answers yet
Around this topic
Sylos connect: narrower topics report up to broader ones, so what's learned in one place shows up where it matters.