QUESTIONNAIRE DESIGN FOR SICK BUILDING SYNDROME - AN EMPIRICAL-COMPARISON OF OPTIONS

This study investigated the effects of questionnaire design on reports of sick building syndrome (SBS) symptoms and environmental discomfort , in order to make informed decisions on the standardisation of SBS qu estionnaires. Twenty-five variants of a questionnaire design were comp leted by 5566 occupants (one variant per occupant) of it office buildi ngs in the UK. A design using separate questions to report symptoms pr oved better than a table (of symptoms and responses) for collecting sy mptom data. Separate questions produce more accurate and complete answ ers at the cost of more space on the questionnaire. Symptom reports we re also affected by details of the description of the symptoms, but no t by the question used to establish whether the symptoms were building -related. Both symptom reports and environmental discomfort ratings we re affected by the frequency scale used to report them, in a manner wh ich is only partly consistent with expectation and a magnitude that is not predictable from a logical comparison of the scales. Thus, differ ent questionnaires may all have been measuring SBS, but by different m ethods and hence obtaining different results. If there is to be greate r comparability of SBS research in the future, then there needs to be a standard questionnaire: a proper calibration of instruments.