Investigating volitional attentional control during film viewing

Abstract

Film viewers’ eye movements seem largely disconnected from their comprehension. Viewers’ eye-movements rarely deviate from focal narrative elements, regardless of differences in their comprehension (Loschky et al., 2015; Hutson et al., 2017), suggesting bottom-up film features overwhelm top-down attentional control. However, viewers’ eye-movements did deviate from focal narrative elements when given a task irrelevant to film comprehension. This suggests viewers used volitional attentional control. However, do people commonly use volitional attention during film viewing? Volitional attention is cognitively demanding, as shown by anti-saccade task failure under executive working-memory (E-WM) load (Mitchell et al., 2002). Thus: 1) Is volitional attention during film viewing cognitively demanding? 2) Is this demand moderated by E-WM capacity? Participants viewed film clips with different task goals and levels of attentional demand while they were eye-tracked. Participants had a primary task of either watching a film clip for comprehension (Comprehension Condition) or drawing a map of the film space from memory (Map Condition). Participants had a secondary task (cognitive load) on half the trials to increase attentional demand. We then assessed their E-WM capacity. We measured viewers’ eye-movement deviation from screen center and used multilevel modeling techniques. We entered film, participant, and their interaction as random effects, and factorially combined cognitive load and condition as fixed effects. The Map Condition had significantly more deviation, showing volitional attention. The cognitive load trials had significantly less deviation, showing less volitional attention control. However, cognitive-load task performance showed two distinct groups of participants, high and low. When accounting for E-WM capacity we expect to find an interaction: specifically, high E-WM viewers may have higher deviation regardless of load, but low E-WM viewers may only achieve high deviation if they ignore the load task. If so, E-WM may moderate the relationship between volitional attention and task during film viewing.