OBJECTIVE: To describe the effect of design associated with a sample survey carried out to estimate the distribution of occupations within an economic active population. METHODS: A cluster sample of households, chosen from a comprehensive directory via systematic random sampling, was performed including 4,782 of all residences in Botucatu, Brazil, between June and July 1997. RESULTS: Of the 4,782 households, 17,219 subjects were assessed. Due to the loss of distribution heterogeneity of the occupations within the households, the effect of […]