Identification and Compensation of Aberrant Response Patterns: Quality of Comfort Assessments in Human Subject Trials

Abstract

Whenever human samples are used to assess a set of different items by using self-reports, concerns regarding the quality of data can arise because participants potentially reply without paying sufficient attention to the contents. The present study reviews different indices of aberrant response patterns and investigates this phenomenon by applying some indicatorson subjective assessments of thermal comfort of N = 160 subjects within the DLR-project Next Generation Train (NGT). Based on this approach, a full sample was compared with a cleaned sample to examine whether aberrant responses have significantly biased the results. Overall, the differences were negligible, which is a proof for the data quality and the utility of human subject trials in comfort assessments. However, the usefulness of indices of aberrant response behaviour is still demonstrated and further suggestions to prevent and reduce aberrant responses in similar studies are discussed.