Calibration of subjective scaling responses

Abstract

Subjective scaling is being used more frequently to assess the lighting characteristics of interiors, although there has been little attempt to validate the responses that subjects provide. It is argued that investigators using subjective scaling should calibrate their subjects in some way because idiosyncratic responses can alter averaged scaling estimates. In this experiment task contrast was varied and estimates of visual performance (speed and accuracy) were collected simultaneously with subjective responses on seven-point semantic scales. Subjects exhibited sterotypical behaviour on the performance measures; their scaling responses were quite idiosyncratic, although some subjective responses to the experimental parameters varied like the performance responses. It is argued that visual performance measures could be used to calibrate subjective scaling responses.