Issues in combining the categorical and visual analog scale for the assessment of perceived thermal sensation: methodological and conceptual considerations.

Typically, the scales for the measurement of thermal sensation have been formatted as categorical scales (CS). Emerging is the use of CS combined with visual analog scale (VAS) for the measurement of thermal sensation to improve the sensitivity of scales. However, reports are rare comparing the typical CS, standard VAS, and combined CS with VAS. Methodological and conceptual issues are arising with the combining of scales, but there are insufficient reports about the advantages and limitations of different scales. The present study compared 9-points categorical scale (9pts CS), VAS, and CS combined with VAS (graphic CS) through a questionnaire survey (n=988) and a controlled experiment during exercise (17 male subjects). Our results showed that graphic CS was more closely related to indoor air temperature for resting residents rather than VAS or 9pts CS. Around thermal neutral zone indoor environments, sensitivity to discriminate thermal sensation was the greatest for graphic CS. In particular, questionnaire responses to VAS showed a remarkable clustering around the thermal neutral zone. For dynamic exercising subjects, mean skin temperature was more closely related to graphic CS than 9pts CS. Our results indicated that graphic CS seemed to be more valid and sensitive than 9pts CS or VAS for the measurement of thermal sensation, but there are many issues to be considered when combining CS and VAS from the methodological and conceptual view points: definitions of terms, verbalizing with descriptors, number of category, scoring length, unipolar/bipolar construction, language translation, central terms, both anchor terms, orientation, color, etc. The above methodological and conceptual issues were discussed.