Variation in working memory capacity and cognitive control: Goal maintenance and microadjustments of control

Abstract

Variation in working memory capacity (WMC) and cognitive control was examined in four experiments. In the experiments high- and low-WMC individuals performed a choice reaction time task (Experiment 1), a version of the antisaccade task (Experiment 2), a version of the Stroop task (Experiment 3), and an arrow version of the flanker task (Experiment 4). An examination of response time distributions suggested that high- and low-WMC individuals primarily differed in the slowest responses in each experiment, consistent with the notion that WMC is related to active maintenance abilities. Examination of two indicators of microadjustments of control (posterror slowing and conflict adaptation effects) suggested no differences between high- and low-WMC individuals. Collectively these results suggest that variation in WMC is related to some, but not all, cognitive control operations. The results are interpreted within the executive attention theory of WMC.

Keywords:

• Cognitive control
• Working memory capacity
• Individual differences

Notes

¹ In Experiment 1, WMC was based on two complex span tasks (operation span, Ospan; and reading span, Rspan), whereas in the other three experiments WMC was based on three complex span tasks (Ospan; Rspan; and symmetry span, Symspan). The change to three tasks reflects a change in our overall screening procedure in which we added symmetry span to ensure that visual spatial skills were being measured along with verbal skills. In a large sample of participants (N = 1,785) from our laboratory, the three tasks all correlate well with one another (Ospan–Rspan = .61, Ospan–Symspan = .43, Rspan–Symspan = .42). Given these strong correlations, excluding Symspan in Experiments 2–4 led to qualitatively identical results to those reported.

² Note that delta plots were examined in Experiments 3 and 4 because delta plots represent the difference between congruent and incongruent trials. Because there are not congruent and incongruent trials in Experiments 1 and 2 delta plots could not be constructed.