Component processes in task switching: cue switch costs are dependent on a mixed block of trials

People are slower when shifting than repeating tasks (switch cost). A considerable portion of the switch cost is due to the possibility of a shift in a mixed block where switches are possible (mixing cost), and to processing a cue that signals a task change (cue switch cost). We use an online sample (n = 12,533) and double cuing paradigm to examine the independent and interactive effects of cue switch costs and mixing costs. All effects were significant, with medium effects for cue changes (eta(2 )(p)= 0.06) and task changes (eta(2 )(p)= 0.10), a large effect for block context (eta(2 )(p)= 0.37), and a small block by cue interaction (eta(2 )(p)= 0.04) indicating that the role of the cue depends on the possibility of a switch. These findings offer empirical completeness by measuring the cue change in both blocks of a switching paradigm. The integrative approach quantifies how separable empirical components contribute to the overall switch cost.