Working memory is (almost) perfectly predicted by g

Working memory is (almost) perfectly predicted by g
Roberto Colom et al., 2004 

http://jtoomim.org/brain-training/working%20memory%20is%20almost%20perfectly%20predicted%20by%20g.pdf

"The central factor in the human-abilities model is g, not fluid intelligence, spatial visualization, spatial relations, or perceptual speed, while the central factor in the theories of human information processing is WM (Jensen, 1998; Lohman, 2000). An adequate representation of psychometric g requires a number and variety of tests (Jensen & Weng, 1994) . The only two studies relating WM with g are the ones performed by Ackerman et al. (2002) and Sueb et al. (2002). The former study found a correlation of .70 between WM and g, while the later study found a correlation of .38 between spatial WM and g, and of .58 between verbal-quantitative WM and g. However, these two studies did not test the predictive power of g over WM and against other first-order latent factors. CFA models are informative with respect to the predictive power of g over WM. Considering the three reported studies in the present article, the loading of g over the WM latent factor averages .96. Therefore, WM is (almost) perfectly predicted by g (92% of explained variance)."