Plancher, G. , Stocker, C. & Barrouillet, P.
Developmental Cognitive Psychology Team. University of Geneva. Geneva, Switzerland
The Time-Based Resource-Sharing model was designed to account for the relationships between working memory functions that are processing and storage. It has been shown that storage capacity is a function of the cognitive load (CL) involved by processing. Adapting to a complex span task the DRM paradigm known to provoke false memories, we wonder whether true and false memories are similarly influenced by CL. Participants studied lists of 6 words, all associated with a non-presented critical item. Between each word, they performed intervening activities varying in cognitive load (high or low) and nature (articulatory suppression or attentional capture). For one group, lists were semantically related (bed, rest, pillow, ... for Sleep), for another, phonologically related (rat, fat, hat, … for Cat). Immediate serial recall followed each list presentation and delayed recognition ended the experiment. While high CL leads to more forgetting of true memories, with articulatory suppression more deleterious at immediate recall and attentional capture at delayed recognition, false memories remain uninfluenced by these factors. Our results suggest that false memories do not appear to rely on working memory mechanisms, but probably rather on long-term memory processes.