Reliability of cross-seasonal memory of environmental conditions and symptoms

In some indoor environment surveys, respondents are asked to recall conditions acrossseasons (e.g. to recall summer conditions during a winter survey). This study assessed thereliability of such recall, based on a survey of 728 people in 12 UK office buildings. In bothwinter and summer, the questions addressed both summer and winter conditions, andbuilding-related symptoms. Correlations were calculated between equivalent responses ineach season (e.g. odour in winter, as rated in the summer and winter surveys). The bestcorrelated IAQ questions were those concerning specific smells (e.g. musty, sharp) or generalsatisfaction with IAQ. Non-specific IAQ factors (e.g. smelly, irritating, stuffy) were less wellcorrelated. Correlations for symptoms also ranged from high to not significant. Buildingsymptom indices, integrating data from all symptoms, were well correlated between seasons;such indices therefore provide a relatively stable measure of the building.