Pupil size

To evaluate whether physiological arousal varied based on the reaction time to the alternating target, we analyzed pupil size in the metronome task. Pupil size, recorded every 2 ms, was averaged over a 200 ms epoch following the end of a saccade, provided the participant remained fixated on that location for a minimum of 200 ms. This window length was selected to avoid any pupil size changes induced by the pupillary light reflex if the target appeared postsaccade (i.e., a source of light stimulates pupil constriction ∼300 ms following its appearance on a dark background; Ellis, 1981; Wang et al., 2018). Pupil trials that met the minimum fixation length criterion were excluded from analysis if a blink or saccade occurred during peripheral target fixation, if the timing of fixation onset exceeded the minimum gap (100 ms) between fixation and next target appearance, and if pupil velocity fell outside the range of −5000°/s to 5000°/s. This left 19175 viable pupil trials in the metronome task to analyze (2 Hz: 4093 trials; 1.33 Hz: 3904; 1 Hz: 3845 trials; 0.8 Hz: 3794 trials; and 0.66 Hz: 3539 trials), and these trials were divided into three categories based on SRT: predictive (6094 trials), express (2384 trials), and regular (10697 trials). To query whether pupil size varied based on prediction or reaction to the rhythmic target, two analyses were performed. First, a one-way ANOVA was computed to test for a main effect of saccade type (predictive, express, and regular) on pupil size. Second, saccade types were combined to test whether SRT correlated with pupil size using a Spearman rank correlation. To test whether pupil size varied across the developmental age range, baseline pupil size for predictive, express, and regular saccades were analyzed per individual subject.