The Evidence on Teaching

Manipulatives and concrete-representational-abstract (CRA) sequences

Manipulatives help when bland, guided, and aged ~7–11 — a guided-representation effect, not 'hands-on learning.' Rich, toy-like objects hurt transfer.

moderate supportconf: mediumgc: low

math · ages 514

Effect summary

Manipulatives help — conditionally: with HIGH instructional guidance, perceptually BLAND objects (rich/toy-like ones hurt transfer), ages ~7-11, and especially fractions (meta ~d 0.3-0.4, exact values unverified). The CRA sequence reliably produces acquisition of trained computation skills in struggling learners when delivered with explicit instruction. This is a guided-representation effect, not a 'hands-on learning' effect.

Practical takeaway

Use plain, boring manipulatives (base-ten blocks, fraction bars) inside explicitly guided lessons, and FADE concrete → pictorial → abstract deliberately. Don't buy the 'hands-on = learning' theory: unguided play with rich attractive objects distracts and hurts transfer.

Who this applies to

Not yet assessed. Nobody has recorded the group size, dose, delivery, or boundary conditions for this decision, so it should not be recommended for a specific situation yet — only read. That is a gap in this record, not a claim that it applies everywhere.

Verdict

Manipulatives occupy a strange place in the math wars: beloved by reform pedagogy as "hands-on learning," but the evidence supports them only as a guided representation tool — and the moderators invert the folk theory. The canonical meta (Carbonneau 2013) finds a small-to-moderate overall benefit over symbols-only instruction that is conditional on high instructional guidance, larger with perceptually bland objects (attractive, toy-like ones hurt distal outcomes — the Kaminski/Sloutsky distraction effect), concentrated in ages 7–11 and fractions, and small for problem solving and transfer. For struggling learners, the concrete-representational-abstract (CRA) fading sequence delivered with explicit instruction reliably produces acquisition of the trained skill (special-ed evidence base; WWC 2021 rates concrete/semi-concrete representations Strong), though on researcher-built probes with no standardized-outcome or durability evidence.

Confidence is capped at medium: the flagship meta's exact effect sizes are paywalled-unverified, the special-ed base is single-case designs where study quality didn't moderate results, and no group-design RCT tests CRA on independent standardized outcomes.

Hereditarian-lens assessment

Risk: low (experimental comparisons). The practical point is about mechanism: the benefit comes from guidance and structured representation, not from the physical objects or child-led exploration — consistent with everything else in the guidance-dose literature.

Practical guidance

  • Plain objects, explicit guidance, deliberate fading: concrete → pictorial/diagram → abstract notation. The fading is the point — the goal is abstract fluency, not manipulative dependence.
  • Prioritize for fractions and place value, ages ~7–11, and for struggling learners (CRA).
  • Skip rich, attractive, toy-like manipulatives and unguided exploration time.
  • Virtual manipulatives work about as well as physical ones — convenience can decide.

Evidence (3 sources)

Export all: BibTeX · RIS

← Back to explore