Abstract
The executive-resource account (Smallwood & Schooler, 2006) proposes that mind wandering (MW) and attention compete for the same executive resources. We tested whether this account holds when executive resources vary due to WM maintenance demands, time-on-task, or the interaction of time-on-task and demand. Participants (N=74) completed a delayed-recognition working memory task, with mnemonic load manipulated across trials (1 item=low load, 2 items=high load) and subjective MW assessed intermittently using a self-report Likert scale (1 = on-task, 6 = off-task). On average, WM task accuracy (% correct) was lower and MW was higher for high vs. low load trials. Results of a multilevel model showed Working Memory Capacity (WMC) and time-on-task moderated the relationship between MW and task-accuracy. Participants reported more MW at the beginning of the experiment for low vs. high load trials but as the experiment progressed, this pattern was reversed, reflecting overall experimental averages of more MW reported during high vs. low load trials. These results suggest that the executive-resource account of MW is lacking in instances when the availability of resources is altered due to mnemonic demands, time-on-task, and the interaction of demand and time.