Search Type Title Author/Investigator Search Advanced Search

Table of Contents | Previous | Next

Introduction

Family structure in the United States changed rapidly in the second half of the twentieth century. A wide variety of family forms increasingly replaced the two-parent family norm. In 2001, 69 percent of children lived in two-parent families, down from 77 percent in 1980 (Federal Interagency Forum on Child and Family Statistics, 2002). Divorce is common. About half of all recent first marriages are expected to end in divorce (Ooms, 2002). Of children born into two-parent families, 34 percent will experience a disruption of their parents’ union by age 16. One-third of all births are out-of-wedlock. And couples opting to cohabit rather than marry is becoming an increasingly common phenomenon. Forty percent of all births occur within cohabiting unions rather than marriages (Bumpass & Lu, 2000). Some European countries also experienced a precipitous decline in marriage rates but have recently seen those rates level and even rise (Ford, 2002).

A vast accumulation of research suggests that children do not fare as well in these alternative family structure forms as children living with their two married biological parents. Numerous studies indicate that children growing up in single-parent families experience worse outcomes than children growing up in two-parent families (Acs & Nelson, 2001; Amato & Keith, 1991; McLanahan & Sandefeur, 1994; Wu & Martinson, 1993). And many studies show that divorce, specifically, is correlated with negative effects on children’s well-being (Amato, 1993; Amato & Keith, 1991; Chase-Lansdale, Cherlin, & Kiernan, 1995; Chase-Lansdale & Hetherington, 1990). Even when parents remarry, a synthesis of the research suggests that this does not appear to improve outcomes (Amato, 1993).

Recent research also suggests a relationship between marriage and positive outcomes for adults. Married couples build more wealth on average than singles or cohabiting couples, while divorce and unmarried childbearing increase the risk of poverty for children and mothers (Lupton and Smith, 2002). Individuals who are married are found to have better health and longer life expectancies than similar singles (Lillard and Waite, 1995). Married mothers have lower rates of depression than cohabiting or single mothers (Brown, 2000). Research also shows that unhappily married adults who divorce or separate, on average, are no happier than unhappily married adults who stay together (Waite et al., 2002).

Relationship Programs
A diverse set of relationship programs currently exists to improve relationships and marriages. Marriage and relationship education programs vary by sponsoring organization, curricular focus, client learning style, and target population. They are developed from government, research, or faith-based initiatives, or they may operate privately for profit. Programs may operate in mental health centers, hospitals, public assistance offices, churches, or universities, among other places. Curricula deal with topics such as communication, parenting or finances. Providers utilize formats that may be instructive, group-oriented, or analytic, and programs may operate with different group sizes and treatment dosage amounts. The programs reach many populations, including individuals (e.g. youth, fathers, mothers), couples (e.g. pre-marital, married), and families.

While the interpretation of outcomes research on marriage and relationship programs is complex, experts in the field suggest generally there is promising evidence that couples can learn specific skills to improve their relationships (Stanley, Markman, & Jenkins, 2002). According to these experts, couples can learn to reduce patterns of negative interaction and maintain higher levels of relationship satisfaction. They noted that in some studies, higher-risk couples show the strongest program effects, and some studies have found that the beneficial effects appear to last up to five years after the training.

U.S. Policy Response
Recently marriage has become a national issue of public policy in the United States. The Bush Administration has proposed that the Federal government dedicate $300 million a year as part of the Temporary Assistance for Needy Families (TANF) program to "help couples form and sustain healthy marriages." Proposed legislation focuses on eight allowable activities:

1. Public advertising campaigns on the value of marriage and the skills needed to increase marital stability and health.
2. Education in high schools on the value of marriage, relationship skills, and budgeting.
3. Marriage education, marriage skills, and relationship skills programs, which may include parenting skills, financial management, conflict resolution, and job and career advancement, for non-married pregnant women and non-married expectant fathers.
4. Pre-marital education and marriage skills training for engaged couples and for couples or individuals interested in marriage.
5. Marriage enhancement and marriage skills training programs for married couples.
6. Divorce reduction programs that teach relationship skills.
7. Marriage mentoring programs, which use married couples as role models and mentors.
8. Programs to reduce the disincentives to marriage in means-tested aid programs, if offered in conjunction with any activity described in this subparagraph.

Source: Personal Responsibility, Work, and Family Promotion Act of 2003. H.R.4.IH

Current Review
This review is designed to inform U.S. policymakers in their effort to provide funding for marriage and relationship programs and will examine specifically how these programs impact measures of relationship satisfaction and communication. However, the greater policy question surrounding the effects of marriage programs relates to the well-being of children and adults, particularly in low-income families. The administration is interested in promoting these programs with the hope that they will lead to an improvement in marital satisfaction and ultimately child well-being outcomes for their participants’ children. It is especially interested in the ways that programs could effectively serve low-income couples. This review can provide an answer to the most immediate policy question: What do published and unpublished evaluations of marriage and relationship programs indicate about the impact of interventions on the satisfaction of couples? Figure 1 below shows this logic model.

Figure 1

General policy question; black box indicates question answered by the current review. [D]

More specifically, this review is a systematic review of evaluations of marriage and relationship programs, which are defined as those that aim to improve the relationship between two people involved romantically. Reviewers performed a systematic search of literature sources and obtained studies for this review that met the following criteria: first, each study had at least one treatment group as well as a no-treatment or wait-list control group; second, each demonstrated that these two groups were created by random assignment or high-quality quasi-experimental methods; third, each presented results on relationship satisfaction, communication, or both that could be converted to standardized effect sizes; and fourth, each study demonstrated that at least 40 percent of its original sample was assessed at pre- and post-test. This review analyzes a final set of 39 studies that met these inclusion criteria. Reviewers find an average effect size of .68 for relationship satisfaction and .26 for relationship communication.

Funding
The U.S. Department of Health and Human Services, Administration for Children and Families (ACF), funded this review to inform policymakers of the existing research on the effectiveness of marriage programs as Congress debates TANF reauthorization. ACF also funded this review to guide implementation of new legislative mandates related to promoting healthy, stable marriages.

Past Reviews
Many reviews have examined the evidence on the effects of marriage programs. Some of these are narrative reviews, which differ from the current review because they do not synthesize results quantitatively. Rather, they present a reviewer's summary of a group of selected studies. One such narrative review examined relationship satisfaction rates and concluded that marital and family enrichment programs produce positive results, leading to significant improvements in pre-marital, marital, and family capabilities (Guerney & Maxson, 1990).

There are also several reviews that have examined the effects of marriage programs using meta-analysis to synthesize findings. In a meta-analytic review, each study included in the review is treated as a data point. Reviewers create an effect size for each study that is based upon results presented by authors of the studies. Reviewers then derive an average impact and associated confidence interval of the interventions from these data.

These meta-analytic reviews are similar to the current review because they use quantitative methods to synthesize results but they differ in that they do not always employ a systematic approach. One meta-analysis examined marital therapy programs and concluded that they are effective, (Shadish et al., 2000). Another meta-analysis examined the effectiveness of two types of therapy programs— behavioral marital therapy and behavioral pre-marital intervention programs. The review indicated that both are more effective than no treatment (Hahlweg & Markman, 1988). Another meta-analysis examined 85 studies of pre-marital, marital and family enrichment programs and also concluded that these programs were effective (Giblin et al., 1985). And a meta-analysis of 16 studies of one program, Couples Communication, concluded that positive couple benefits can be anticipated from the program (Butler, 1999). Finally, one meta-analysis that focused on therapy programs and employed a systematic search found a relationship between the effectiveness of the therapy programs and different types of formats used in family and marital psychotherapy (Shadish, 1993).

Usefulness of Another Review?
This review builds on these previous reviews and confirms their findings that marriage programs lead to positive effects for couples. Still, it differs from each review in at least one way. It is different from the Guerney & Maxson (1990) review because it synthesizes results quantitatively, examines marital enrichment programs and other program types (therapy, education, pre-marital preparation, and counseling), and employs a systematic review process. Following a systematic review process means that prior to conducting the review, reviewers planned exactly how they would search for evaluations and how they would screen studies for inclusion, and documented these search and inclusion criteria prior to the start of the review in a protocol. These methods contribute to transparency of the review methods and remove one possible source of reviewer bias. Though the Shadish et al. (2000) and Hahlweg & Markman (1988) reviews are also meta-analyses, the current review is distinguishable by the fact that it goes beyond reviewing evaluations of marital therapy programs to include studies of other program types. It is also a systematic review. The review differs from the Giblin et al. (1985) and Butler (1999) reviews because it goes beyond evaluations of enrichment programs and communication skills programs, respectively, and is also systematic. The current review differs from a systematic meta-analysis, the Shadish (1993) review, because it covers studies of other types of programs besides therapy programs. And finally, this review is unique because it specifically aims to inform policymakers at the U.S. Department of Health and Human Services.

The review will also be submitted to the Campbell Collaboration’s Social Welfare Review Group to become an approved Campbell review. The Campbell Collaboration—modeled after the Cochrane Collaboration for reviews of medical research trials— is a group that promotes the creation and dissemination of systematic reviews in the fields of social welfare, education, and crime and justice. The group has strict guidelines for the reviews it approves and mandates that reviewers include all obtainable high quality evidence available on a given topic. The review must also be done systematically to eliminate any reviewer bias. The current review has been conducted in parallel with several other Campbell-style reviews that are funded by the Smith Richardson Foundation.

Criteria for study inclusion
This review includes rigorous evaluations of programs designed to improve the relationship between two people involved romantically. Specifically, several inclusion criteria were used.

1. Goals of the intervention: The intervention must serve both members of a couple that is romantically involved. A principal component of the intervention must be relationship improvement, but there may be other goals of the intervention.
2. Counterfactual treatment: The study must contain a control group that does not receive the treatment evaluated by the study between the time of pre- and post-intervention measurements, or by the time follow-up measures are taken if applicable. However, this group would be able to receive treatment from an outside source during the evaluation if the group members chose to do so.
3. Target population: The focus of the study analyses must be on the individuals who are part of a couple relationship.
4. Outcomes measures— type: The study must include measures of relationship satisfaction or communication. Specific examples of measures found by the studies are detailed later in the "characteristics of studies/results presented by studies" section.
5. Outcomes measures— unit: The measures may be taken at the individual or couple level.
6. Outcomes measures— reporting: The study must present sufficient data from which to calculate standardized effect sizes and weights based upon the inverse of the variance of the effect size.
7. Study quality: The study design must be random assignment, matched random assignment, or a high quality quasi-experimental design (meaning that evaluators created a control group using rigorous statistical techniques). In addition to this, the evaluation report must indicate that the attrition rate between the time of random assignment and the time the post-test measures were taken was no higher than 40 percent. Reviewers will assess independently those studies that demonstrate an attrition rate of 20 percent or lower.
8. Year of publication: The evaluation must have been published or completed since 1960.

Search strategy
The reviewers’ search strategy for identifying relevant studies included four components: database searches of published literature; internet searches for published and unpublished research; manual searches of journals, books, and other reviews on relationship program evaluations; and professional contacts.

Reviewers limited the search to literature published or completed since 1960. Although the bulk of research on this topic occurred between 1977 and 1982, experts in the field cite studies as far back as 1962 (Hunt et. al., 1998). Moreover, a change arguably came about in western society as a result of the sexual revolution of the 1960s, which brought about changes in the family form. Reviewers included both published and unpublished work from any country and in any language.

The reviewers used the following method to conduct a search for evaluations. They searched sources, such as databases and Internet sites, for records, which in most cases were titles and abstracts of reports or journal articles. If there were fewer than 100 records available from a specific source, reviewers selected all records for screening. In the interest of time, it did not make sense to implement a specific search "strategy" for such a small number of records. If there were between 100 and 500 studies available from a source, the reviewers keyed in “ marriage OR marital OR pre-marital OR relationship OR couple OR premarriage OR newlywed ” and screened all records retrieved. If there were between 500 and 2,000 records available, reviewers searched using the following phrase ( “marriage” OR “marital” OR “pre-marital” OR “relationship” OR “couple” OR “premarriage” OR “newlywed”) AND (“program” OR “satisfaction” OR “quality” OR “stability” OR “enrichment” OR “education” OR “therapy” OR “counseling” OR “learning” OR “outcome” OR “communication” OR “treatment”).

If the search engine was not sophisticated enough to allow reviewers to use a complex search string, reviewers entered the phrases “marriage* and program*,” “marriage* and satisfaction,” “marriage* and quality,” and so on. If there were more than 2,000 studies returned from this search, then reviewers crossed this search with an “AND” statement and the following words “evaluation* OR impact* OR experiment* OR intervention* OR random* OR control*.”

Reviewers implemented this search strategy on several academic databases:

Databases— Cochrane Central Register of Controlled Trials (CENTRAL); Dissertation Abstracts International; EconLit; ERIC; LILACS; Mental Health Abstracts; POPLINE; Population Index; PsychInfo; Sage Family Studies Abstracts; Social Science Citation Index; Social Services Abstracts; Social, Psychological, Educational, and Criminological Trials Register (SPECTR); Sociological Abstracts; Substance Abuse and Mental Health Data Archive.

Reviewers searched websites of government agencies, research organizations, professional associations, information services, university policy and social research programs, relationship programs, faith-based groups, and foundations. Specifically, reviewers searched the following sites for documents:

Government agencies— Commonwealth Department of Family and Community Services (Australia); Department of Health and Human Services— Administration for Children and Families (ACF, USA); Department of Health and Human Services— Substance Abuse and Mental Health Services Administration (SAMHSA, USA); Federal Statistical Office (Germany); INSEE (French government’s statistics bureau, France); main government websites for Australia, Canada, and the EU (as well as individually: the United Kingdom, Sweden, Norway, Denmark, Spain, and France); National Institute of Mental Health (NIMH, USA); Oklahoma Marriage Initiative (USA); Statistics Norway (Norway); Swiss Federal Statistics Office (Switzerland); and state government websites for Arizona, Arkansas, Florida, Louisiana, Maryland, Minnesota, Michigan, Oklahoma, Utah (USA).

Research Organizations— Alan Guttmacher Institute (USA); Alternatives to Marriage Project (USA); American Association for Marriage and Family Therapy (USA); Australian Institute of Family Studies (Australia); Brookings Institution (USA); Austrian Institute for Family Studies (Austria); BC Council for the Family (Canada); Building Family Strengths Parent Information Center (USA); Center for Law and Social Policy (USA); Child Trends (USA); Council of Contemporary Families (USA); European Observatory on the Social Situation, Demography, and Family (Europe); Family Research Council (USA); Heritage Foundation (USA); Human Science Research Council (South Africa); Institute for American Values (USA); Interdisciplinary Centre for Comparative Research in the Social Sciences (ICCR) (Austria and France); International Academy for Marital Spirituality (Belgium); National Center for Children in Poverty (USA); National Center for Policy Research for Women and Families (USA); National Council on Family Relations (NCFR) (USA); National Fatherhood Initiative (USA); National Governor’s Association (USA); National Institute of Relationship Enhancement (USA); National Marriage Program (USA); National Marriage Week (U.K.); One Plus One (U.K.); Penn Council for Relationships (USA); Tavistock Marital Studies Institute (UK).

Professional associations— American Bar Association; American Psychological Association (APA, USA); American Psychological Society (APS, USA); Australian Psychological Society; British Psychological Society; Canadian Psychological Association; Children, Youth, and Families Education and Research Network (USA); Directory of Scholarly and Professional E- Conferences; Stepfamily Association of America (USA); European Association for Advancement of the Social Sciences; European Health Psychology Society; European Sociological Association; Family Law Section (USA); International Association of Applied Psychology; International Society of Health Psychology Research; Marriage and Relationship Educators’ Association of Australia Inc. (Australia).

Information services— Americans for Divorce Reform (USA); National Research Council Research Information Service (USA); Research Engine for the Social Sciences (USA); Smart Marriages (USA); UK Online

University Policy and Social Research Programs— Center for Family Research; University of Cambridge (UK); Center for Marital and Family Studies; Center for Research and Child Well-Being at Princeton University (USA); Center for Marital and Family Studies, University of Denver (USA); Center for Research on Family, Kinship, and Childhood, University of Leeds (UK); Family Action Centre; University of Newcastle (Australia); Family Centre, University of Queensland (Australia); Family Research Unit, University of Jyvaskyla (Finland); Gottman Marriage and Family Institute (USA); Institute of Family and Sexuality Studies at the School of Medicine, Catholic University of Leuven (The Netherlands); Institute for Family Research and Counseling at the University of Fribourg (Germany); Institute for Social and Economic Research at the University of Essex (U.K.); Newcastle Center for Family Studies; University of Newcastle (Australia); Rutgers University (The National Marriage Project); School of Higher Studies (Paris); University of Washington (USA).

Relationship Programs— Fragile Families (USA), Marriage Alive (USA), Marriage Savers (USA), PAIRS (USA), PREP (USA).

Faith-based Groups— Catholic Society for Marriage Education (Australia)

Foundations— Annie E. Casey Foundation (USA)

News/media— The American Prospect (USA)

Reviewers also conducted manual searches of journals by searching tables of contents from these journals since 1960 or the inception of the journal if it began after 1960. These journals included American Journal of Family Therapy, Family Relations, Journal of Consulting and Clinical Psychology, Journal of Marital and Family Therapy, and the Journal of Sex and Marital Therapy. Reviewers chose these journals because they published a large number of the evaluations originally returned by the search. Reviewers also screened lists from 53 other reviews of marriage and relationship program evaluations. They obtained all studies that fit the inclusion criteria for the current review.

Reviewers contacted experts in the field of marital and family therapy— including Alan Hawkins (Brigham Young University), Bill Coffin (U.S. Department of Health and Human Services), Thomas Bradbury (UCLA), Benjamin Karney (University of Florida), and Bill Doherty (University of Minnesota)— to determine if any studies were missing from the list of included studies for this review. Of these experts, two (Karney and Bradbury) responded and suggested several titles beyond the current group of 39. Reviewers determined that all of these studies had been part of the original set of 514 evaluations screened previously in the review but were eliminated because they did not contain a no-treatment control group.

Contacting authors of studies for missing data
Reviewers contacted study authors if a study had insufficient information— either relating to attrition rates or to results. Study authors either did not present information on attrition rates for their sample, or they presented results in a way that made it impossible for reviewers to create an effect size. Unfortunately, some of these authors could not be found. Of those who reviewers contacted successfully, none were able to produce the missing data.

Outcome of the Search
Reviewers obtained 12,828 abstracts from the aforementioned sources. They immediately screened all abstracts that were clearly not evaluations of programs. Most of these citations were articles related to marriage or marriage programs, but were not actual program evaluations. For those of a somewhat unclear nature, reviewers extracted full texts for further analysis. They obtained 514 full-text articles in this stage of the review, and most of these were program evaluations.

Reviewers screened the 514 full-texts based upon the inclusion criteria, and the majority of the studies did not meet the criteria. Fifty-nine of these were other reviews of marriage evaluations, which were used to further inform the reviewers of the current review. Of the citations deleted from the group 514, some were not evaluations of marriage programs (179) or available in full text (46). There were studies without random assignment (54), with poor random assignment (3), or with poor quality quasi-experimental designs (1). Other studies did not contain a control group or control group data (129), presented results that could not be converted into effect size form (14), or had attrition that exceeded 40 percent (4). Reviewers retained 39 evaluations of marriage programs for analysis.

Figure 2

[D]

Thirty-nine studies remained after reviewers checked the evaluations against the inclusion criteria. Table 1 (in attached documents) provides descriptive information on each of these studies. The table highlights the following characteristics: program type, publication year, location, program intensity, total length of time in session, number of couples in each session group, sample characteristics, distress level, and follow-up assessment.

 

 

Table of Contents | Previous | Next