Fully Labeled Scales Are More Cognitively Demanding: A Multimodal Cognitive Load Approach

Abstract

This article investigates cognitive load in fully labeled and end-labeled Likert-type scales. We compared cognitive load between two conditions using different methods such as pupillometry, completion times, and subjective evaluation of mental effort. We conducted a laboratory experiment (N = 170) using eye-tracking technology to measure pupil dilation while respondents completed a web survey. The mean pupil diameter was larger in the fully labeled than in the end-labeled condition. Because the fully labeled scales provide more information and require respondents to read labels for all scale values, they increase neurophysiological load compared to the end-labeled scales. Though the effect was small, it was consistent. We found no difference in completion time and subjective rating of mental effort between the conditions. Despite higher cognitive load in the fully labeled condition, data quality benefits from the provision of clear verbal labels in the scale.