The Evidence on Teaching

Does Calculation or Word-Problem Instruction Provide a Stronger Route to Pre-Algebraic Knowledge? (modularity RCT)

Fuchs, L. S., et al. · 2014

grade Brctdeveloper-ledreplicatednumbers spot-checked
Sample
1,102 students / 127 classrooms (3-arm cluster RCT)
Population
US elementary students.
Design
Calculation instruction vs word-problem/schema instruction vs control.
Key findings
The cleanest demonstration that math sub-skills are MODULAR: fluency does not create problem-solving, and problem-solving does not create computation. Both folk theories fail — each strand must be taught explicitly.
Genetic confound
Randomized.
Replication notes
Numbers verified. The modularity finding replicates across the Fuchs program.
DOI / URL
10.1037/a0036793

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
Calculation instruction -> word problemsES0.07-0.14 (null transfer)mixedend of treatmentbusiness-as-usualend-of-treatmentnear-transfer
Word-problem instruction -> calculationsES0.14 (null transfer)researcher-designedend of treatmentbusiness-as-usualend-of-treatmentnear-transfer
Word-problem/schema instruction -> pre-algebraic knowledgeES1.36 proximal / 0.22 distal (calculation arm: 0.01-0.04)researcher-designedend of treatmentbusiness-as-usualend-of-treatmentdomain-skill

Cited by