Implications of attitude-behavior studies for population research and action.

Results of many family planning surveys do not suggest that attitudes are strongly predictive of family planning practices. An inquiry into methodological issues suggests two recommendations for improvement of research procedures. First, more attention must be paid to the reliability of the questionnaires by which attitudes and practices are measured. Second, questions on attitudes and practices should refer to the same action or object and to the same period of time. The available evidence suggests that implementation of these procedures can lead to identification of stronger linkages between family planning attitudes and practices, and to survey results having greater practical utility.