Subgroups holding different conceptions of scales rate room temperatures differently

Publication date: 15 January 2018
Source:Building and Environment, Volume 128
Author(s): Xaver Fuchs, Susanne Becker, Karin Schakib-Ekbatan, Marcel Schweiker
Current thermal comfort models are fairly inaccurate at predicting occupants' thermal comfort from parameters of indoor environment. The predicted variables, thermal sensation and thermal comfort, are commonly measured using scales. These scales might themselves contribute to the problem of poor prediction due to inter- and intra-individual differences in respondents' interpretation of scales. Until now, it is unclear whether variation in scale use is systematic and potentially statistically controllable. This study investigated (1) whether there are subgroups holding different conceptions of scales and (2) whether these conceptions are associated with different ratings of sensation and comfort under experimental conditions. Sixty-three participants completed a free positioning task that assessed the relative distances between labels belonging to the ASHRAE thermal sensation scale and their distribution along various dimensions (sensation, preference, comfort, pleasantness, acceptability, and tolerability). Subsequently, the participants rated office rooms at cool, neutral, and warm conditions regarding the same dimensions. Latent class regression on the free positioning task revealed subgroups showing distinct and interpretable patterns such as preferences for different temperature ranges. Remarkably, these patterns were mirrored in the participants' ratings under the experimental thermal conditions. The results suggest the existence of different conceptions concerning the relationships between the labels of the ASHRAE scale. The prediction of participants' ratings of thermal conditions in concrete situations can be significantly improved when taking these conceptions into account.