Moving beyond averages: variations in reported thermal comfort

Thermal comfort research characterises group thermal perception using averages. This approach overlooks the value of analysing variation as a dependent variable characterising groups’ state of comfort. In this paper, we reviewed the results of 219 surveys carried out in five schools in the UK and in Sweden between 2011 and 2016. Results show that pupils’ thermal sensation and preference votes varied more at moderate indoor operative temperature. This result suggests that pupils may have a greater range of adaptive opportunities, including clothing, in moderate environments. Substantively, reviewing the spread of the thermal comfort is critical to unpick behavioural, psychosocial and physiological mechanisms. Furthermore, results are significantly different while analysing the central tendency or spread of comfort votes. For example, there is no difference in comfort votes’ central tendency between surveys carried out during the heating seasons and the non-heating seasons but there is a significant difference in the spread, indicating the need for multilevel analysis. Methodologically, reviewing the spread of thermal comfort is also critical to establish the data analysis method. With recent advances in surveys’ tools allowing larger datasets to be gathered at individual and group levels, it is essential to review the range of analysis methods