Retrieval Success

The present project is a preregistered replication study examining whether successful retrieval causally enhances long-term retention when item memorability is controlled. Retrieval-practice effects are often interpreted as evidence that successful retrieval strengthens memory more than exposure to feedback alone. However, many paradigms confound retrieval success with item difficulty or memorability, making it difficult to isolate the causal role of retrieval itself. Participants study novel item–label pairs and undergo repeated learning rounds until retrieval success is achieved. Two learning conditions are compared: successful retrieval and retrieval attempts followed by feedback. Retention is assessed using a productive test. The design closely follows an initial study that observed small, non-significant differences between conditions, motivating the need for a higher-powered replication. The primary goal of the replication is to obtain a more precise estimate of the effect of learning condition on retention and to evaluate whether any difference between conditions can be ruled out within a prespecified equivalence bound. Data will be analysed using a mixed-effects logistic regression model with random intercepts for participants and items. An equivalence test will be conducted on the condition effect using bounds defined in advance. The study is designed to be informative regardless of outcome. Evidence for equivalence would suggest that successful retrieval does not confer a meaningful advantage over retrieval followed by feedback when memorability is controlled. Conversely, failure to demonstrate equivalence would indicate that effects of this magnitude cannot yet be ruled out and would motivate further investigation.