Refixation Patterns of Mind-wandering during Real-world Scene Perception

Abstract

Recent studies have shown that mind-wandering (MW) is associated with changes in eye movement parameters, but have not explored how MW affects the sequential pattern of eye movements involved in making sense of complex visual information. Eye movements naturally unfold over time and this process may reveal novel information about cognitive processing during MW. The current study used Recurrence Quantification Analysis (Anderson, Bischof, Laidlaw, Risko, & Kingstone, 2013) to describe the pattern of refixations (fixations directed to previously-inspected regions) during MW. Participants completed a real-world scene encoding task and responded to thought probes assessing intentional and unintentional MW. Both types of MW were associated with worse memory of the scenes. Importantly, RQA showed that scanpaths during unintentional MW were more repetitive than during on-task episodes, as indicated by a higher recurrence rate and more stereotypical fixation sequences. This increased repetitiveness suggests an adaptive response to processing failures through re-examining previous locations. Moreover, this increased repetitiveness contributed to fixations focusing on a smaller spatial scale of the stimuli. Finally, we were also able to validate several traditional measures: both intentional and unintentional MW were associated with fewer and longer fixations; Eye-blinking increased numerically during both types of MW but the difference was only significant for unintentional MW. Overall, the results advanced our understanding of how visual processing is affected during MW by highlighting the sequential aspect of eye movements.

Scene meaningfulness guides eye movements even during mind-wandering

“…Indeed, Zhang et al (2020), using the same dataset, showed that scanpaths during unintentional MW were more repetitive compared to on-task episodes but a similar pattern was not observed for intentional MW. However, because there was no significant difference between the intentional vs. unintentional MW on the normalized meaning value, the results cannot point to a clear distinction between them.…”

“…After discarding data from 7 participants who had low tracking ratios (< 75%, a threshold set a priori), the final sample consisted of 57 participants. Note that analyses performed with the same sample, but which examined distinct research questions from those in the current study, are reported elsewhere (Zhang, Anderson, & Miller, 2020).…”

“…Table 1 summarizes the results of several behavioral and global eye movement measures. Because these measures are not of primary interest to our hypotheses and have been tested elsewhere (Zhang et al, 2020), we simply report descriptive statistics without performing significance testing. and tracking loss are reported on a 0-to-1 scale.…”

“…and tracking loss are reported on a 0-to-1 scale. More details about these measures can be found in Zhang, Anderson, and Miller (2020).…”

What do Scanpaths Tell Us About Cognitive Processes? An Investigation in a Problem Solving Domain

“…The results showed that when mentally recalling an image, our eye movements tend to re-fixate earlier and more often compared to actual scene viewing. Zhang, Anderson, and Miller (2021) also used the RQA measures to examine the effect of mind wandering on visual attention while examining different imagery. The results showed that during mind wandering participants' revisit earlier fixations as if to review stimuli, resulting in duplicated scanpaths.…”