Apt pupils: Investigating what’s going on when nothing is happening

Description

This paper investigates the utility of pupillometry in Violation-of-Expectations (VOE) paradigms. We showed infants videos of toy trains traveling on a circular track on which there were two tunnels. In Experiment 1, 10-month-olds (n = 24) were familiarized to plausible events where trains were visible in the gap between tunnels and tested on a combination of plausible or implausible (i.e., the train was not visible in the gap) events involving familiar- or novel-coloured trains. Infants looked longer at the gap when a train was visible, but no effect of condition was observed using looking times. Pupil diameter analyses reveal an interaction between plausibility and familiarity when a train ought to appear in the gap. In Experiment 2, 12-month-olds (n = 24) were familiarized to implausible events and tested on combinations of plausibility and familiarity. Looking time analyses do not distinguish between the test trials, although infants look longer at the gap when a train was present. Pupil analyses reveal
an effect of plausibility on the second passage of the train through the gap; pupil diameter was largest when the train failed to appear in the gap. In both experiments, the repeated occurrence of an implausible event within a trial has a seemingly priming effect on pupil dilation. In Experiment 1, which uses a typical procedure of familiarizing to plausible events, the implausible test trial is absolutely novel. In Experiment 2 however, infants are familiar with implausible events yet also exhibit within trial priming. Although these results are primarily driven by the absence of stimulation (i.e., there is no train to be seen), they have general implications for the design of VOE experiments. Moreover, they highlight the advantage of fine, time-locked measures like pupil diameter over the more common cumulative looking time which remains the typical dependent measure in such studies.

Period	24 Aug 2022
Event title	Lancaster Conference on Infant and Early Child Development
Event type	Conference
Location	Lancaster, United KingdomShow on map