Affective state influences retrieval-induced forgetting for integrated knowledge

Background: Selectively testing parts of learned materials can impair later memory for nontested materials. Research has shown that such retrieval-induced forgetting occurs for low-integrated materials but may be prevented for high-integrated materials. However, previous research has neglected one factor that is ubiquitous in real-life testing: affective state. Methodology/Principal Findings: We investigated whether affect influences the resistance of integrated materials to retrieval-induced forgetting by inducing neutral, positive, or negative affect immediately before selectively testing previously learned textbook passages containing interrelated facts and concepts. As negative affect is known to promote a detail-oriented local processing style, we hypothesized that experiencing negative affect during testing may decrease the protective effects of integration and lead to reoccurrence of forgetting. By contrast, as positive affect is known to promote a relation-oriented global processing style, we hypothesized that experiencing positive affect may support effects of integration and prevent forgetting. Our findings are consistent with these predictions. No subsequent forgetting occurred when testing memories for integrated text materials in affectively neutral and positive states, whereas forgetting occurred when testing in negative states. A correlation analysis showed that forgetting decreased with higher positive affect, with participants experiencing high positive affect even showing facilitation instead of forgetting. Conclusions/Significance: These findings indicate that affect can moderate the memory consequences of test taking and suggest that educators should use testing as a tool to improve memory with care.