Couple Relationships over the Transition to Parenthood: Methodological Issues in Testing for an Intervention Effect

Abstract

A conceptual framework using transition in meaning/identity, roles, and relationship/interaction was used to develop and test a three-class, prenatal couple relationship intervention. A nonrandomized convenience sample (treatment group [TG] = 43, control group [CG] = 62 prenatal class couples) was compared preclass and postclass and at mean 10.5 months postbirth. Couples with lower relationship quality self-selected into TG. Both TG and CG showed a rise in relationship satisfaction from preclass to postclass and a pattern of returning to baseline postbirth. ANCOVA (Time 1 children and parenting as covariate) found no treatment effect; TG declined in conflict resolution from Times 1 to 3. Methodological issues in intervention research that may have impeded a teamwork effect are discussed in this article, including randomization, intervention length, unintended intervention and control group effects, content “fit” and need for piloting, attrition, sample size, recruitment, tool length, and participant fatigue.