Concreteness and levels of processing: a test of the dual-coding hypothesis using dynamic visual noise.

Concreteness and levels of processing (LOP) effects have been attributed to the differential availability of visual images for concrete words, and at deeper levels of processing, respectively. Interestingly, the concreteness effect has been shown to disappear under conditions involving dynamic visual noise (DVN), which is thought to suppress the generation of visual images from long-term memory. The present study further investigated the role of visual imagery in concreteness and LOP effects. Across four experiments, DVN was manipulated during study, and participants' memory for concrete and abstract words was measured using recall and recognition tests. Although some support for dual-coding was found, concreteness and LOP effects were not fully explained by visual imagery because they were present under DVN conditions. We conclude that concreteness and LOP effects may be better explained by an "extended dual-coding theory" that incorporates the role of context availability in accounting for this pattern of results.