Spontaneous adaptation of maintenance strategies during a complex span task

This experiment is part of a series of five previous studies. The general objective of this set of experiments is to investigate the influence of the free-time distribution on the type of maintenance strategies that participants spontaneously use during a complex span task. The first four studies have already been published (Leproult et al., 2024). The five previous studies were consistent in showing an increase in working memory (WM) recall performance when participants had few long free times (i.e., condition with massed free times) compared to many short free times (i.e., condition with distributed free times), despite equivalent total free time between experimental conditions. The initial hypothesis was that more free time at once would promote the use of elaborative strategies, and thus improve recall performance. Nevertheless, a number of arguments against this explanation emerged (e.g., lack of such a beneficial effect for the delayed recall, lack of an effect of free-time distribution effect on the type of strategies reported; see Leproult et al., 2004 for more details). Consequently, we are now considering an alternative hypothesis to explain the boost in immediate recall during longer uninterrupted free-time periods, focusing on direct semantic maintenance in WM (Martin et al., 1996). Beyond phonological maintenance, this is thought to be a memory maintenance mechanism that directly targets semantic representations of items and actively keeps them accessible in WM. Research suggests that this mechanism is distinct from other WM maintenance mechanisms like phonological rehearsal, and it is not mediated by recovery in long-term memory (LTM; Acheson et al., 2010; Martin et al., 1996, 1999; Shivde & Anderson, 2011). Therefore, we propose a new study in which we manipulate the availability of semantic features in WM through the concreteness of memoranda. The rationale is that semantic features are more available for concrete words than for abstract words (e.g., Bourassa & Besner, 1994; Acheson et al., 2010; Plaut & Shallice, 1993). More specifically, researchers suggest that concrete words are associated with more representations in LTM (De Groot, 1989) and with richer representations (i.e., including sensorial components; dual-coding theory, Paivio, 1971). Consequently, we propose that if the effect of free-time distribution is due to an increased mobilization of semantic representations in WM, then it should be accentuated for concrete words compared to abstract words.