ABSTRACT
Indoor air pollution, caused by the indoor burning of biomass fuels, has been associated with an increased risk of child acute respiratory infections in developing countries. The amount of time that children spend in proximity to fires is a crucial determinant of the health impact of indoor air pollution. Researchers are reliant on social scientific methods to assess exposure based on child location patterns in relation to indoor fires. The inappropriate use of methods could lead to misclassification of exposure. The aim of this paper is to compare two methods (observations and questionnaire interview) with video analysis (which is thought to offer a more accurate assessment of exposure) in rural South African villages. Compared to video analysis, results show that observations may underestimate the amount of time that children spend very close (within 1.5 m) to fires. This is possibly due to reactivity caused by the presence of an observer. The questionnaire interview offers a more accurate assessment of the amounts of time that children spend within 1.5 m of fires at the expense of a detailed behavioural analysis. By drawing on the strengths and weaknesses of each, this paper discusses the appropriateness of methods to different research contexts.
