Generative processing and emotional false memories: a generation "cost" for negative false memory formation but only after delay.

Previous research shows that manipulations (e.g. levels-of-processing) that facilitate true memory often increase susceptibility to false memory. An exception is the generation effect. Using the Deese/Roediger-McDermott (DRM) paradigm, Soraci et al. found that generating rather than reading list items led to an increase in true but not false memories. They argued that generation led to enhanced item-distinctiveness that drove down false memory production. In the current study, we investigated the effects of generative processing on valenced stimuli and after a delayed retention interval to examine factors that may lead to a generation effect that increases false memories. At the immediate test, false recognition rates for both negative and neutral valanced critical lures were similar across read and generate conditions. However, after a one-week delay, we saw a valence differentiation, with a generation effect for false recognition but only for negative stimuli. The roles of item-specific and relational processing during encoding and their interaction with long-term retention are discussed.

Reconsolidation or interference? Aging effects and the reactivation of novel and familiar episodic memories.

We examined aging effects in reconsolidation and interference in episodic memory by reactivating memories for well-learned items in young and healthy older adults while controlling memory strength and the degree semantic processes contributed to memory. In Experiment 1, young and old adults learned pairs of real words and images to a strict criterion. After 24-hours, half of the images were reactivated and new words were paired with the images and learned to criterion. Following a 1-week delay, recognition and source monitoring were measured for both sets of pairings. Experiment 2 was a replication of Experiment 1, but using previously unknown novel words and unusual images. As predicted, older adults needed more trials to learn both the A-B and A-C pairings. Older adults required more trials to learn the new associations for reactivated than the not reactivated pairs, although there was no main effect of reactivation and no Age x Reactivation interaction for measures of recognition one-week later. These results are inconsistent with previous findings concerning age differences in reactivation effects in episodic memory. Instead, they suggest that once memory strength and input from semantic memory are better controlled, young and old adults perform similarly on tests of long-term recognition memory.