Effects of Image and Verbal Generation on False Memory

Abstract

The effects of image and verbal generation on false memories in the Deese–Roediger–McDermott (DRM) paradigm were investigated by comparing three experimental conditions (description, image, and image description). False memory rates were reduced when participants received an explicit imagery instruction rather than when they were asked to describe the characteristics of word referents or both imagine and describe referents. Verbal generation is thought to promote the associated processing of list items and thus increase the probability of activating lures during encoding and of falsely remembering. Conversely, self-generated imagery is expected to decrease the probability of falsely recalling and recognizing the lures. Imagery encoding should be used as a distinctiveness heuristic at retrieval for impeding the false recalls/recognitions of the critical items. Nevertheless, when imagery encoding was coupled with a verbal generation task, mental images were no longer sufficiently stringent criteria for reducing false memories. These findings could challenge the conditions of using imagery techniques when people have to remember an event while evoking it in a mental image so as to describe it in greater detail.