The Evidence on Teaching

Implicit and explicit instruction in L2 learning

Goo J, Granena G, Yilmaz Y, Novella M · 2015

grade Cmeta-analysisindependentnot-applicable
Sample
34 unique sample studies (11 carried over from Norris & Ortega 2000, 23 published 1999-2011)
Population
Second-language learners across the included studies; adult and university learners dominate
Design
An explicit update of Norris & Ortega restricted to studies that DIRECTLY compared explicit and implicit instructional treatments, which is a cleaner contrast than the original meta-analysis's mixture of treatment-versus-control and treatment-versus-treatment comparisons. Published as a book chapter in Studies in Bilingualism.
Key findings
Explicit instruction was found to be more effective than implicit instruction overall, with the difference examined across several moderator variables. The direction of Norris & Ortega's central claim survives a fifteen-year update on a partly new corpus.
Genetic confound
Low as a threat to the within-study comparisons.
Replication notes
This IS the replication of Norris & Ortega on the explicit-implicit contrast, and it succeeds in direction. Together with Spada & Tomita and Kang et al. it makes "explicit instruction beats purely implicit exposure for target structures" one of the better-replicated claims in applied linguistics - on treatment-aligned measures.

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
Explicit vs implicit L2 instructionpooled effect size comparisonexplicit more effectiveresearcher-designedpost-testactive-alternativeend-of-treatmentdomain-skill

Cited by