Covenant Marriage: The Movement to Reclaim Tradition in America

Appendix A: Sampling and Research Methods

153 Appendix A: Sampling and Research Methods This appendix describes the methods and goals of the various studies used in this project. We explain the sample design, recruitment and retention procedures, and goals of our core panel study of newlywed couples and the various implementation and program evaluation studies. Our project is unusual because it relied on multiple methods. Our approach used multiple ways of framing the core questions and various methods to address different audiences affected by covenant marriage. Briefly, we have both assessment and policy-implementation components, which include studies of the general population’s attitudes about the necessity of marriage-law reform and receptiveness to and knowledge about covenant marriage, an assessment of the critiques made by relevant political constituencies with contrasting stakes in covenant marriage, and an assessment of the quality of the implementation of the various legal features of covenant marriage by state actors. We also have two quantitative studies that directly explore the effect of covenant marriage on the lives of couples considering and entering marriage. First, we explore how engaged couples move toward marriage and select or reject the option of covenant marriage as they transition to marriage. Second, we use a combination of qualitative methodologies and a three-wave quantitative panel study of newlywed couples who either chose the new covenant or established standard marriage laws for their unions. In the following pages, we provide a thumbnail sketch of each of these projects. We present only the methodologies for the components presented in this book. We begin with the core study of covenant as compared to standard marriages, the heart of this larger project. Marriage Studies Marriage Matters Panel Survey of Newlywed Couples, 1998–2004 The key part of our project on covenant marriage is to study the innovators who availed themselves of the new covenant marriage law. These data are from three 154 Appendix A waves of a survey study of newlywed couples over the course of the first five to seven years of their marriage. The first wave was administered over a 1998–2000 time span, marking the first few months after the August 1997 passage of covenant marriage into the first few years of its inception. We recruited couples throughout this time span and surveyed them, on average, three to six months after their weddings . We administered the second and third waves approximately eighteen months after the first, and twelve to twenty-four months after the second wave, respectively. We left the field in late spring 2004. The sample selection criteria consisted of two steps. First, seventeen of sixty parishes were selected randomly and proportionate to size. Second, from these seventeen parishes, all covenant marriage licenses and the standard marriage licenses filed next to the covenant licenses were drawn. We purposefully over-sampled standard licenses because of our expectation, later confirmed, that our covenant marriage innovators would be more committed to the rigors and duration of the study. Of the 1,714 licenses that were validly part of our sampling frame, we eventually confirmed 1,310 couples for a confirmation rate of 76.4 percent. This confirmation rate is notable because our research teams found these couples via the briefest information listed on the marriage licenses. The licenses listed the names of the couple, the couple’s parents, and the witnesses to their marriage. Without address or telephone information, we conducted multiple phone book and Internet searches to find someone who could provide contact information for these newly married, highly mobile couples. From these confirmed couples, a total of 2,589 initial surveys were mailed, 975 for covenants and 1,614 for standards. An administrative error at a mailing facility resulted in the mis-mailing of an undetermined number of these questionnaires. They were sent as bulk rather than first-class mail. This meant that the U.S. Postal Service did not attempt to forward to new addresses. Once we discovered this error via an assessment of return rates, we immediately corrected the mailing problem. This initial setback resulted in a 49 percent overall response rate for the first wave. Covenant couples overall had a 57 percent first wave response rate, standard couples a 44 percent response rate. Between surveys, we worked actively to maintain our sample’s integrity. We telephoned the couples occasionally to confirm their addresses. If we lost them, we contacted the family and friends they had listed as contacts on their questionnaires . We sent birthday and anniversary cards...