Comparing Methods for Analysis of Pupillary Response.

Abstract

Changes in eye-pupil size index a range of cognitive processes. However, variations in the protocols used to analyze such data exist in the psychological literature. This raises the question of whether different approaches to pupillary response data influence the outcome of the analysis. To address this question, four methods of analysis were compared, using pupillary responses to sexually appetitive visual content as example data. These methods comprised analysis of the unadjusted (raw) pupillary response data,z-scored data, percentage-change data, and data transformed by a prestimulus baseline correction. Across two experiments, these methods yielded near-identical outcomes, leading to similar conclusions. This suggests that the range of approaches that are employed in the psychological literature to analyze pupillary response data do not fundamentally influence the outcome of the analysis. However, some systematic carryover effects were observed when a prestimulus baseline correction was applied, whereby dilation effects from an arousing target on one trial still influenced pupil size on the next trial. This indicates that the appropriate application of this analysis might require additional information, such as prior knowledge of the duration of carryover effects.