Statistical Analysis

A paired-samples t-test was used to determine whether the Cognitive Load manipulation was successful in increasing cognitive demand, as determined by RSME scores. Effect size is reported as Cohen’s d.

Separate paired-samples t-tests were used to determine whether there was a significant decrease in raw performance scores for either the cognitive task or the motor task (stepping accuracy) during Dual-Task trials, when compared to single-task. A Wilcoxon test was used to determine whether there was a significant decrease in raw performance scores for time to complete the walking trial during Dual-Task trials. The use of a non-parametric test was deemed necessary here and elsewhere in the paper if data were non-normally distributed. Separate paired-samples t-tests were then used to determine whether any DTCs observed for either the cognitive task or the motor task (stepping accuracy) were significant, when compared to zero (which represented identical Single- and Dual-Task performance). A Wilcoxon test was used to determine whether any DTCs observed for time to complete the walking trial were significant, when compared to zero. Effect size is reported as Cohen’s d, unless the assumption of normality is violated, whereby effect size is reported as r = Z/√N [44].

A paired-samples t-test was used to investigate the effect of Cognitive Load on the number of task-relevant ‘inside’ fixations. These data were normalized to trial length, with the number of fixations presented as the average number of fixations per second. Separate Wilcoxon’s tests were used to investigate the effect of Cognitive Load on the duration (as a percentage of overall fixation durations) of task-relevant ‘inside’ fixations, and on both the number (/per second) and duration (as a percentage of overall fixation durations) of task-irrelevant ‘outside’ fixations, as these data were non-normally distributed.

A repeated measures ANOVA was used to investigate the percentage of time spent fixating task-irrelevant areas during each temporal time-bin. Effect size is reported as partial eta squared (ƞp²). Any significant effects were followed up by pairwise comparisons with Bonferroni adjustments.

Separate bivariate correlations were run between three measures of trait-reinvestment (trait-MSRS; trait-CMP; trait-MSC) and each of the aforementioned variables. As some variables of gaze data were not normally distributed (the duration of task-relevant ‘inside’ fixations for Baseline trials; the number of task-irrelevant ‘outside’ fixations for Baseline trials, and; the duration of task-irrelevant ‘outside’ fixations for Baseline trials), Spearman’s correlation were used for these comparisons. All other data were analyzed using Pearson’s correlation.