[Method problems in the study of health status of domestic workers]

This feasibility study was undertaken to evaluate methodological problems in identifying and obtaining permission to interview domestic employees. Worksite, convenience and residential sampling strategies were compared. Worksite sampling proved to be a relatively efficient method if restricted to middle and upper socioeconomic status neighborhoods and yielded the most representative sample; however, non response rate was quite high. When reasons for non response were evaluated it was found that employer refusal and difficulty in locating daily workers were the most important ones. Residential sampling proved very inefficient. Convenience sampling proved to be highly efficient with no problem of non response but yielded a biased sample.