Schedule of Events | Search Abstracts | Invited Symposia | Symposia | Rising Stars | Poster Sessions | Data Blitz

Retrieval Practice Drives Integration Across Related Memories: A Candidate Mechanism for the Testing Effect

Poster Session A - Saturday, March 7, 2026, 3:00 – 5:00 pm PST, Fairview/Kitsilano Ballrooms

Lauren Homann¹ (lauren.homann@mail.utoronto.ca), Morgan Barense^1,2; ¹University of Toronto, ²Rotman Research Institute

Retrieval practice may strengthen targets by integrating them with related memories, yet empirical evidence for this mechanism is limited. We tested this account of the “testing effect” (enhanced memory after retrieval practice, versus restudy) in a two-day behavioural study (N=37). Participants encoded events in which one person interacted with one object in a specific location. Importantly, “related event pairs” constituted two different events sharing an identical object and scene-type (e.g., park). Each pair was assigned to one of three within-participant conditions, with events reviewed individually: retrieval practice (reimagining the event from a face cue, with feedback), restudy (full re-exposure), or control (no review). During retrieval and restudy, participants also answered questions probing each event’s location and object category (“directly-reviewed associations”). On Day 2, we tested memory for directly-reviewed associations (assessing the “testing effect”) and examined changes in associative structure across related event pairs (assessing integration). Although retrieval and restudy yielded similar accuracy for directly-reviewed associations (both outperforming no-review control), retrieval uniquely integrated related event pairs across multiple measures. Relative to restudy and control, retrieval promoted erroneous recombination of background details, enhanced the ability to infer links between individuals across related events, and increased co-dependence in memory for directly-reviewed associations of related events (i.e., associations were co-remembered or co-forgotten). Moreover, greater event integration predicted stronger retention of directly-reviewed associations. Results indicate that retrieval practice preferentially integrates reviewed content with related memories, suggesting that the “testing effect”, when observed, reflects not merely mnemonic enhancement, but a broader restructuring of relational memory networks.

Topic Area: LONG-TERM MEMORY: Episodic