Table of Contents | Previous | Next |
I. The Research Foundations for the Supporting Healthy Marriage Project
The SHM project is motivated by three related but distinct bodies of research: Studies indicating that children in low socioeconomic status (SES) families spend less time in two-parent families (and that this gap between lower- and higher-income children continues to grow); studies suggesting that children benefit from growing up with two parents who are in a stable, low-conflict relationship; and random assignment evaluations demonstrating that, at least for some target groups and some outcomes, marriage education interventions can have positive effects on couples and their children.
Trends in Marriage and Divorce
An important literature in recent years has examined differences in marriage and divorce by socioeconomic status. Married couples with low education levels are more likely to divorce (Fein, 2004; Martin, 2006), and when low-income married couples split up, their children are likely to become poor and dependent on public assistance (Behrman and Quinn, 1994). Furthermore, socioeconomic inequality in children’s access to two parents has increased in recent decades. Low-income individuals have become increasingly likely to have children before marrying, and, for those who marry, dissolution rates have remained very high while declining among better-off couples (Ellwood and Jencks, 2004; Martin, 2006; Raley and Bumpass, 2003). If publicly funded interventions like SHM can promote marital quality and stability, they may be able to contribute to more equitable prospects for children.
Influences of Family Structure and Family Relationships on Child Well-Being
Reviews in other recent volumes have summarized past research on the effects of family structure on children’s well-being (Amato, 2005; Ribar, 2004). Individual studies, as well as meta-analyses, have attempted to disentangle the effects on children of living in different types of households by comparing outcomes for children in a variety of family configurations. Additional research has attempted to uncover the potential causal contributions of specific experiences of children who live in different family structures. These several decades of nonexperimental studies have provided considerable evidence that children benefit when they have access throughout their childhoods to adequate financial resources; effective parenting; a stable household; and minimal exposure to parental conflict.
Over the last two decades, the evidence has strengthened that children who grow up with two parents who are in a stable relationship are more likely to receive the parenting and financial supports that promote their well-being (Amato, 2000; McLanahan and Sandefur, 1994). At the same time, there are many uncertainties about exactly what inputs matter most, the optimal role for public policy, and whether and why marital relationships may be particularly advantageous for children (Acs, 2007; Kamp Dush, this volume; Sweeney, this volume).
Although carefully designed comparisons between children in different circumstances have contributed to our understanding of the processes by which families support children’s development, nearly all of these prior studies are nonexperimental and vulnerable to potential bias due to parents’ self-selection into different family structures and other circumstances (Ribar, 2004). Thus, although the primary function of the SHM project is to test the effectiveness of marriage education for disadvantaged married couples, the experimental design of the SHM study also affords a unique opportunity to contribute to existing basic research: The proximate targets of the intervention (parents’ handling of conflict, positive relationship qualities, and/or marital stability) will have been influenced by an exogenous influence — random assignment to either a group that receives marriage education or a group that does not. By analyzing both these proximate impacts of this intervention and the more distal effects on aspects of family life, such as parents’ mental health, parenting and co-parenting, family income, and children’s well-being, SHM has the potential to shed new light on the causal links between individual and family characteristics and child well-being that are difficult to identify definitively in a nonexperimental framework.
Effectiveness of Marriage Education Programs
A last important thread in prior research is a literature suggesting that psycho-educational interventions can be effective at improving couples’ relationships. A common characteristic of these programs is their prevention outlook: At the same time that basic research on marriage was learning much about early communication and other behaviors associated with long-term relationship distress, practitioners were becoming increasingly convinced that marital therapy is likely to be only modestly successful, given that couples often are seriously distressed by the time they seek help (Bradbury, Johnson, Lawrence, and Rogge, 1998; Christensen, 1999). Substantial enthusiasm thus greeted the idea that it might be possible to teach couples how to recognize and practice positive interaction and avoid negative exchanges while their relationships are still in good shape (Halford, 2001; Halford, Markman, Stanley, and Kline, 2003; Markman and Floyd, 1980; Silliman et al., 2002).
Until recently, this literature was comprised mainly of experiments involving small samples of mostly middle- and upper-class couples and conducted mainly by researchers and clinicians who developed the interventions. Meta-analyses of varying subsets of programs over the past two decades suggest that preventative psycho-educationally oriented programs on average have produced moderate positive effects on relationship satisfaction and communication (Butler and Wampler, 1999; Carroll and Doherty, 2003; Giblin, Sprenkle, and Sheehan, 1985; Hawkins, Blanchard, and Fawcett, 2007; and Reardon-Anderson, Stagner, Macomber, and Murray, 2005).
The earliest programs studied, such as the Couple Communication program (Wampler, 1990) and the Relationship Enhancement program (Guerney, 1977), focused on communication and problem-solving skills. A landmark 1985 meta-analysis of 85 studies (Giblin, Sprenkle, and Sheehan, 1985) found average effect sizes of .3 (men) and .5 (women) for relationship satisfaction and of .6 for communication skills (for both men and women).
A second generation of interventions expanded the skills taught beyond communication and conflict resolution to include new insight into the role of couples’ expectations and attitudes; emotions (both positive and negative); the meaning that might be derived from recurrent conflicts; and the importance of nurturing the positive side of the relationship, including fun, friendship, emotional supportiveness, and intimacy. The most widely disseminated of these is the Prevention and Relationship Enhancement Program, PREP (for example, Hahlweg et al., 1998; Halford, Sanders, and Behrens, 2001; Markman and Hahlweg, 1993; Stanley et al., 2001). Recent meta-analyses — which include second-generation evaluations in addition to earlier evaluations — have reported effect sizes roughly comparable with those limited to earlier programs (Carroll and Doherty, 2003; Hawkins, Blanchard, and Fawcett, 2007; Reardon-Anderson, Stagner, Macomber, and Murray, 2005).
Collectively, a series of sampling and technical limitations in these earlier studies makes it difficult to predict confidently the impacts of more intensive programs such as SHM, developed for low-income couples. As mentioned, nearly all of the studies cited involved pre-dominantly white, middle-class couples. Furthermore, though a good number of studies utilized experimental designs, many of them measured outcomes only for nonrandom subsamples of subjects originally assigned to experimental and comparison groups. Such selection occurred through a variety of mechanisms: Researchers decided to study only those treatment group members who participated in or completed the program; there was attrition when marriages dissolved; or researchers could not locate original sample members (Carroll and Doherty, 2003; Reardon-Anderson, Stagner, Macomber, and Murray, 2005). Very few studies employed statistical techniques such as regression-adjustment that might help control for such sources of imbalance in the designs. Finally, these evaluations tended to follow couples for only a very short time — in many cases extending no more than several months after the program ended — and rarely measured impacts on marital stability or child outcomes.
A third set of marriage education studies focused on couples experiencing stresses related to child-rearing has begun to address some of these limitations. Common to this newer class of interventions is the notion that parents may be especially open to participating in marriage education if it is presented in a way that supports their roles as parents or addresses specific transition points in family life. Until recently, these studies again involved primarily middle-class couples, although they also tended to be based on somewhat larger samples, more careful designs, and longer follow-up.
The first important evaluation in this category, of a transition-to-parenthood program called Becoming a Family, used a group discussion format with skilled clinicians and found effects on a range of outcomes, including positive effects on marital satisfaction (but not stability) at five and a half years after random assignment (Cowan and Cowan, 1992). More recently, the same investigators have reported impacts on marital satisfaction, parenting, and children’s test scores and behavior in Schoolchildren and their Families, a study that recently completed a 10-year follow-up analysis (Cowan and Cowan, 2006a). The ongoing Supporting Fathers’ Involvement Study is finding impacts from a similar, discussion-based model for a mostly low-income Hispanic population (Cowan, Cowan, Pruett, and Pruett, 2006). Other higher-quality experiments that reported promising findings in this category include tests of the Bringing Baby Home program (Shapiro and Gottman, 2005) and an adaptation of the Incredible Years that addressed relationships between parents who are coping with children with serious behavior problems (Webster-Stratton and Taylor, 2001).
Thus, work to date provides many reasons to test more intensive,
large-scale psycho-educational models and address the substantial
deficit in knowledge about how marriage education may affect low-income
couples. In the remainder of this chapter, we describe the conceptual
model, program design, and evaluation approach for the SHM demonstration,
which is testing relatively intensive and comprehensive services in
eight sites around the country.
| Table of Contents | Previous | Next |