The Evidence on Teaching

Effects of First-Grade Number Knowledge Tutoring With Contrasting Forms of Practice

Fuchs, L. S., Geary, D. C., Compton, D. L., et al. · 2013

grade Brctdeveloper-ledunreplicatednumbers spot-checked
Sample
591 at-risk first graders (195 speeded / 190 nonspeeded / 206 control)
Population
At-risk US first graders; 16 weeks, 3x/week tutoring.
Design
Both arms got identical 25-min conceptual number-knowledge tutoring; the contrast is 5 min of SPEEDED retrieval practice vs nonspeeded principle-based practice games. The single best causal test of timed vs untimed practice.
Key findings
Adding 5 minutes of speeded retrieval practice to conceptual tutoring beat identical tutoring with untimed practice (d=0.51 head-to-head), with NO word-problem cost and no attitude/motivation differences — and it specifically helped weak reasoners compensate. Direct causal counterevidence to the anti-timed-practice claims (anxiety unmeasured).
Genetic confound
Randomized. The ATI: timed practice reduced outcome dependence on (heritable) reasoning ability — a compensatory, gap-narrowing interaction.
Replication notes
Numbers verified (N=591 corrected). Direction converges with Powell 2009 and the Number Rockets replication; no independent replication of the specific speeded-vs-nonspeeded contrast.
DOI / URL
10.1037/a0030127

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
Arithmetic: speeded vs nonspeeded (head-to-head)d0.51 (0.87 vs control; nonspeeded 0.38 vs control)researcher-designedend of treatmentactive-alternativeend-of-treatmentdomain-skill
Word problems: speeded vs nonspeededd0.04 (ns) — NO cost of timed practiceresearcher-designedend of treatmentactive-alternativeend-of-treatmentdomain-skill
Aptitude-treatment interactionpatternspeeded gains UNIFORM across reasoning ability; nonspeeded gains reasoning-dependent — speeded practice lets weak reasoners compensateresearcher-designedend of treatmentactive-alternativeend-of-treatmentdomain-skill
Attitudes/motivation/effortcondition differencesnoneresearcher-designedend of treatmentactive-alternativeend-of-treatmentnon-cognitive

Cited by