In prior work, one of us has suggested that WM is represented by both primary and secondary memory components (Unsworth and Engle, 2007a and Unsworth and Spillers, 2010a). Primary memory reflects both the number of items that can be distinctly maintained and attention control

processes that actively maintain those items and prevent attentional capture. Secondary memory reflects the need to retrieve items that could not be maintained in primary memory as well as the need to retrieve other relevant information from secondary learn more memory. According to this multifaceted model of WM, there are multiple sources of variance within WM measures, and multiple sources of variance that account for the relation between WM and gF (Unsworth, in press, Unsworth and Spillers, 2010a and Unsworth et al., 2009; see also Conway, Getz, Macnamara, & Engel de Abreu, 2011). Likewise, Cowan et al. (2006) suggested that both capacity and attention control would be important sources of variation. The current study represents a direct test of this multifaceted view of WM and its relation to gF. In particular, although prior work has suggested that each of these factors (attention control, capacity,

and secondary memory retrieval) are important, no study has simultaneously examined all three to determine if they will jointly mediate the relation between WM span and gF. As noted previously, WM always seems to have a residual relation with gF, even after controlling for other factors. However, this could be due to the fact that no prior study selleck inhibitor has jointly examined all three factors. In one prior study, both attention control and secondary memory were examined, but WM still predicted gF after controlling for these other two factors (Unsworth & Spillers, 2010a). This suggests that WM is composed of distinct processes and these processes independently

contribute to individual differences in gF. If the multifaceted view of WM is correct, then we should see that WM is related to all three factors, all three factors are related to gF, and importantly all oxyclozanide three factors mediate the relation between WM and gF, with little to no residual relation between WM and gF. Furthermore, given that in most prior studies the storage score from complex span tasks was used to index WM, we also examined measures of processing (specifically processing time) from the complex span tasks. As mentioned previously, prior work has suggested that WM represents resource sharing between processing and storage and it is this resource sharing ability that leads to variation in WM and accounts for its relation with higher-order cognition (Case et al., 1982, Daneman and Carpenter, 1980, Daneman and Tardif, 1987 and Just and Carpenter, 1992). However, other research suggests that processing and storage make independent contributions to performance and to the relation with gF (Bayliss et al., 2003, Logie and Duff, 2007, Unsworth et al., 2009 and Waters and Caplan, 1996).