The reliability of measuring embarrassment: A generalizability theory approach

To assess subjective feelings of embarrassment, researchers often employ scenarios designed to evoke embarrassment in participants, along with questions aimed at capturing their emotional responses. However, the reliability of this measurement approach remains uncertain, particularly regarding the influence of different scenarios. Through an online survey with a crowdsourced sample, this study aims to explore the impact of both choice of questions and scenarios on reliability by adopting a generalizability theory perspective. Participants were presented with various scenarios depicting both public and private situations and were subsequently asked to rate their emotional responses. Results underscore the importance of incorporating multiple scenarios and questions to achieve adequate generalizability across both public and private situations. Furthermore, a greater number of scenarios and questions are particularly beneficial in private situations than in public ones. Correlation analysis involving personality traits indicates that incorporating several scenarios helps prevent the attenuation of correlation coefficients. This study concludes with practical recommendations aimed at enhancing the quality of embarrassment measurements from a generalizability perspective.