A comment on method

Before I describe the data that I rely on it is important to make a comment on methodological issues associated with the measurement of ‘unmet need’. I have used surveys to assess reproductive processes in my own work, however, because of my experiences with surveys, I find it important that careful consideration be given to the usefulness of surveys (alone) for the measurement of processes in­volved in (reproductive) decision-making. While surveys are useful for providing cross-sectional data (and even longitudinal data), they have limitations when it comes to providing insights into what goes on between couples and how. The dis­crepancies between survey and interview responses among the same sample high­light these limitations. In the ensuing discussion I use the example of my own sur­vey data, modelled along the lines of the DHS, as well as case stories, to show how surveys can both over — and understate the extent of ‘unmet need’, as well as re­verse the identity of the partner who actually has the ‘unmet need’. The case of