Combining Methods:

Abstract

In this article, the author looks at the processes involved in constructing concepts such as family and household using different methods. The author argues that the choice of method does matter but that there is no a priori link between method and construct or between method and epistemological position. However, the extent to which different methods can be used in a complementary way depends on the epistemology that the researcher adopts. If households or families are seen as objectively independent of the researcher's interpretations of them, then merging findings may be seen as unproblematic. If, however, concepts such as family are seen as the joint production of researcher and subject, then the production process itself must become the focus of analysis before data set up using different methods can be merged. The extent of overlap between constructs produced by different methods must be investigated rather than assumed.