Sampling couples to understand them: Mixing the theoretical with the practical

The project described here is one of the first to obtain a probability sample of couples based on characteristics of both partners. The undertaking is presented as a mix of theoretical and practical concerns. First, I present the theoretical background justifying the need for such a sample and outline the methods of the longitudinal study of couples. Then the steps taken to recruit the sample (238 couples) are traced, including the process of screening for eligible respondents. Several unforeseeable problems occurred that required more time, and thus more money, than were originally budgeted. This paper exposes these difficulties in a straightforward manner not often seen in journal articles that typically cover up or smooth out all blemishes. I also offer suggestions for ways to diminish the effects of such problems. The hope is that other researchers who wish to follow a similar path will find this paper helps make their journeys smoother than the one described here.