Working memory is a system devoted to short-term storage and processing and is used in various cognitive tasks, such as reading, reasoning, and mental arithmetic. Throughout the past decennium, research into the role of working memory in mental arithmetic has flourished (for a review, see DeStefano & LeFevre, 2004) and has shown that solving both simple arithmetic problems (e.g., 8 5, 3 9) and complex arithmetic problems (e.g., 23 98, 12 35) relies on working memory resources. The present study further investigates the role of working memory in simple-arithmetic strategies, on the basis of the multicomponent working memory model of Baddeley and Hitch (1974). In this model, there is an attentional system (the central executive) that supervises a phonological subsystem and a visuospatial subsystem. The phonological subsystem guarantees short-term maintenance of phonological information, and the visuospatial subsystem guarantees short-term maintenance of visuospatial information.The role of executive working memory resources in simple arithmetic has been shown extensively (see, e.g., Ashcraft, 1995;De Rammelaere, Stuyven, & Vandierendonck, 1999Deschuyteneer & Vandierendonck, 2005a, 2005bDeschuyteneer, Vandierendonck, & Muyllaert, 2006;Hecht, 2002;Lemaire, Abdi, & Fayol, 1996;Seitz & Schumann-Hengsteler, 2000. The role of phonological working memory resources in simple arithmetic is less clear. In some studies, an effect of phonological load on simple-arithmetic problem solving was observed (see, e.g., Lee & Kang, 2002;Lemaire et al., 1996;Seitz & Schumann-Hengsteler, 2002), whereas, in other studies, it was not (see, e.g., De Rammelaere et al., 1999Seitz & Schumann-Hengsteler, 2000). Investigations of the role of the visuospatial "sketch pad" in simple arithmetic are scarce (but see Lee & Kang, 2002;Seitz & Schumann-Hengsteler, 2000), and the findings in the few studies conducted are equivocal.A drawback of all the studies mentioned above, however, is that none of them showed any distinction between retrieval and nonretrieval trials. Yet it has been shown that adults use several strategies to solve even the simplest arithmetic problems (see, e.g., Hecht, 1999;. For instance, although direct memory retrieval (i.e., knowing that 3 4 12) is the most frequently used strategy, nonretrieval strategies (or procedural strategies) are used as well. Examples of such procedural strategies are transformation (e.g., 96 (10 6) 6 60 6 54) and counting (e.g., 4 7 7 . . . 14 . . . 21 . . . 28). The studies mentioned above notwithstanding, it is impossible to discern the specific simple-arithmetic strategies in which executive and phonological working memory resources are needed.Investigations of the roles of executive and phonological working memory across different simple-arithmetic strategies have only very recently been conducted, beginning with Hecht (2002). In his study, simple addition equa-
1759Copyright 2007 Psychonomic Society, Inc.
Do multiplication and division strategies rely on executive and phonological working memory...