Structured Peer Tutoring Outperforms Unstructured¶

Claim · All claims
Evidence · 2 studies · 2 quant-synthesis r? · q3
Currency · newest study 2006 · 1 of 2 entries checked against their sources (automated) · not reviewed by a person

Peer tutoring produces larger learning gains when the tutoring relationship is structured — with defined roles, training for tutors, scripted or scaffolded interaction protocols, and regular monitoring — than when students are simply paired and told to help each other.

Subclaims¶

q3 i? Across a meta-analysis of 52 tutoring-achievement studies, structured tutoring programs produced significantly larger achievement gains than nonstructured tutoring programs, though no numeric effect-size split was reported for this specific moderator. → Cohen et al. 1982

q3 i? Across 36 elementary-school peer-assisted-learning studies, structured student roles were one of the program components the abstract reports as related to the size of social, self-concept and behavioral effects; the abstract gives neither the direction nor the magnitude of that relationship, so this entry does not by itself show that structured tutoring outperforms unstructured tutoring. → Ginsburg-Block et al. 2006

Evidence¶

Cohen et al. 1982¶

Cohen, P. A., Kulik, J. A., & Kulik, C.-L. C. (1982). Educational outcomes of tutoring: A meta-analysis of findings. American Educational Research Journal, 19(2), 237–248. doi:10.3102/00028312019002237

q3 · meta-analysis of 65 school-tutoring evaluations · i? · significant moderator, magnitude not numerically reported · n=52 studies (tutee achievement) · quant-synthesis · r?

A meta-analysis of 65 evaluations of school tutoring programs (52 reporting tutee achievement) found an overall achievement effect of tutoring of d = .40. Six study features were significantly related to the size of this effect, and whether the program was structured or nonstructured was one of them: tutoring effects were larger in the more structured programs. The authors state this held even though the less-structured programs still produced a real benefit — the difference was one of degree, not of whether unstructured tutoring worked at all. (Read in full as the 1981 AERA conference version of this synthesis, ERIC ED204416, which reports the same 65 studies; the journal version was not reachable.)

Ginsburg-Block et al. 2006¶

Ginsburg-Block, M. D., Rohrbeck, C. A., & Fantuzzo, J. W. (2006). A meta-analytic review of social, self-concept, and behavioral outcomes of peer-assisted learning. Journal of Educational Psychology, 98(4), 732–749. doi:10.1037/0022-0663.98.4.732

q3 · meta-analysis, moderator analysis (abstract only) · i? · no effect size reported for this moderator · n=36 studies · quant-synthesis · r?

A meta-analysis of 36 elementary-school peer-assisted-learning studies examined social, self-concept, and behavioral (not just achievement) outcomes. Program components including student autonomy, individualized (ipsative) evaluation, structured student roles, interdependent group rewards, and same-gender grouping were each related to the size of these effects. This is consistent with Cohen et al.'s finding for achievement outcomes, but only the abstract was available, and it says only that these components "were related to effect sizes": it does not state the direction or magnitude of the relationship for "structured roles", so it cannot show that structured tutoring outperformed unstructured tutoring on these outcomes.

Discussion¶

What "structured" means here. Structure in peer tutoring typically includes: (a) explicit tutor training before tutoring begins, (b) clear role differentiation (tutor/tutee, sometimes alternating), (c) scripted or semi-scripted interaction routines — e.g., prompting, questioning, and feedback protocols rather than answer-giving, and (d) ongoing teacher monitoring and feedback on tutoring quality. Unstructured arrangements lack these supports and tend to drift toward answer-sharing or off-task talk.

Why structure matters mechanistically. Structured protocols push tutors toward explanation, questioning, and elaboration rather than simply supplying answers, which benefits both tutor and tutee — the same elaborative-explanation mechanism that underlies broader collaborative learning benefits [+M]. Structure also manages the working-memory and skill demands placed on tutors, who are themselves novices — consistent with Cognitive Load Theory and the broader claim that cognitive overload degrades learning [+M]. Without training and scaffolds, tutors often model unproductive help-seeking and help-giving behaviors [-M].

Boundary conditions and constraints. The advantage of structure is likely strongest for younger learners and low-ability tutors, who have fewer self-regulation resources to organize an interaction on their own [-M]. Highly able, well-motivated dyads may extract benefit from looser arrangements — a possible manifestation of the expertise reversal effect, in which scaffolds that help novices become redundant or constraining for the more competent [-M]. Overly rigid scripting can suppress the productive negotiation of meaning that makes peer tutoring valuable in the first place, and can reduce tutoring to recitation rather than dialogue [~M]. Unstructured pairings without tutor training frequently devolve into answer-giving or off-task interaction, eroding the gains peer tutoring is meant to deliver [-M]. Role reversal (reciprocal tutoring) may add benefits by giving both partners the chance to explain, but this is an open design question rather than a settled finding [~W].

Open questions. The relative contribution of each structural component (training vs. scripting vs. monitoring) is not well separated in the literature; most effective implementations bundle them. Whether reciprocal role reversal adds gains beyond fixed roles, and how much structure to fade as tutors gain experience, remain unsettled design questions.

Note on evidence. This page has two Evidence entries, both meta-analyses (Cohen et al. 1982; Ginsburg-Block et al. 2006); neither reports a numeric effect size for the structured-versus-unstructured moderator, and the mechanism and boundary-condition points above are not sourced on this page.

Related Claims¶

• In school studies, cooperative and peer-learning methods that reward groups on every member's individual learning raised achievement more than methods without such rewards — the general claim this finding bears on, with the studies behind it

• Collaborative situations often fail to be productive in practice, so orchestrating collaboration is not simple — the general claim this finding bears on, with the studies behind it

• Collaborative learning improves outcomes. — peer tutoring is a structured form of collaborative learning; conditions for effective collaboration apply here

• Active learning improves exam performance. — peer tutoring is one active-learning structure whose benefits depend on how interaction is organized
• Cognitive Load Theory — explains why novice tutors need scaffolds and protocols to avoid overload
• Expertise reversal effect — the amount of structure that helps novice tutors may need to fade as tutors gain competence
• Mastery Learning Improves Outcomes — related
• Peer-assisted learning interventions improve achievement for elementary students — related
• A descriptive report defines peer assisted learning and asserts, without a tested result, that it promotes knowledge construction, collaboration ability and interpersonal interaction — a broader claim this one bears on
• Peer Tutoring Improves Achievement — related
• Learning By Teaching Improves Tutor Learning — related
• Tutoring is highly effective, and hiring older peers as tutors supports learning gains for both older and younger students — related
• Unstructured expert interviews miss up to 70% of the subtasks needed to describe a cognitive task — related
• Turn-taking test review makes each member's expertise visible, enabling peers to locate and draw on community knowledge — related
• Structured criteria in peer assessment are not shown to improve its effect on academic performance over freeform review — reports the opposite