Relationship Education Research: Current Status and Future Directions

Abstract

The overarching aim of this paper is to review research on relationship education programs and approaches that have been published or accepted for publication since the last review article in 2002. This paper provides a critical overview of the relationship education field and sets an agenda for research and practice for the next decade. A theme weaved throughout the paper are the ways in which relationship education is similar and different from couples therapy and we conclude that there can be a synergistic, healthy marriage between the two. We then provide recommendations for future directions for research in the relationship education field. Finally, the co-authors comment on our experiences in both the relationship education field and couples therapy field as both researchers and interventionists.

The overarching aim of this paper is to review research on relationship education programs and approaches that have been published or accepted for publication since the last review article (Halford, Markman, Kline, & Stanley, 2003). We include research published from 2002 and beyond. Our overarching aim is to provide a critical overview of the relationship education field and set an agenda for research and practice for the next decade.

We start by discussing why we decided to use the term “relationship education” as compared to other terms such as marriage education, couples/marriage enrichment, couples/marriage enhancement, couples communication, and prevention of marital distress and divorce. We also provide an overview of the ways in which relationship education differs from couples therapy. Next, we briefly discuss the rationale for relationship education in terms of theory, research, and changing demographics. We then provide a review of the outcome research on relationship education during the past decade and highlight the key issues in the field based on this review. We highlight training issues for relationship education researchers (current and future) and provide recommendations for future research in the relationship education field. Finally, at the end of the paper each of us (the co-authors), comment on our experiences in marrying the relationship education and couples therapy fields.

Defining Relationship Education

For the purposes of the paper, we broadly define relationship education as efforts or programs that provide education, skills, and principles that help individuals (a person not in a relationship or a person without his or her partner) and couples (both partners participating) increase their chances of having healthy and stable relationships. Extensive overlap exists between relationship education and couples enrichment, communication, enhancement, and prevention programs. Enrichment and enhancement programs are examples of marital health promotion interventions (Van Widenfelt, Hosman, Schaap, & van der Staak, 1996) and are typically designed to help couples who wish to increase their level of relationship satisfaction (e.g., Silliman & Schumm, 2000). Communication programs, as their name suggests, are designed to teach couples effective communication skills. We divide prevention programs into three types based on principles of prevention science applied to couples (Markman, Stanley, & Kline, 2003). Universal prevention programs typically start with young couples who are happy and share the goal of keeping happy couples happy (Markman & Floyd, 1980). Selected intervention programs focus on couples at risk for distress or divorce. Finally, indicated programs target couples in the early stages of distress. We decided to use the term relationship education to cover all of these areas because most enrichment, enhancement, communication, and prevention programs include relationship education (and skill training) as a key component. While other authors may prefer the term couples relationship education (Halford, Markman, & Stanley, 2008; Hawkins, Blanchard, Baldwin, & Fawcett, 2008), we decided to omit the “couples” to include relationship education interventions delivered to individuals.

In the last comprehensive review paper, Halford, Markman, Kline, and Stanley (2003) defined best practices in relationship education as having seven key features: 1) assessment and measurement of variables associated with risk for distress or divorce (e.g., having divorced parents, negative communication), 2) encouragement of high-risk couples to participate, 3) assessment and education about relationship aggression, 4) the provision of relationship education at transition points (e.g., around the time of the birth of the first baby), 5) the provision of relationship education to “mildly distressed” couples early in the progression towards distress, 6) adaptation of programs for couples with diverse populations, and 7) increased accessibility of evidence-based relationship education. As we proceed, in our review of the recent literature we will examine the progress toward these goals that Halford et al. (2003) set forth for the relationship education field.

Relationship Education and Couples Therapy

Research and practice clearly indicate that couples who participate in relationship education programs sometimes report distress and may require therapy (Hawkins et al., 2008). As noted by Ooms (2010), there is substantial confusion amongst couples, service providers, and social policy makers concerning relationship education and couples therapy. For a comprehensive review of couples therapy programs and research see Lebow (2010). For a fuller discussion of the differences and similarities between couples therapy and relationship education, please see Ooms (2010). Here, we focus on the area where the boundaries are somewhat blurred, with the goal of elucidating the similarities and differences between the two fields.

We define therapy as interventions delivered to distressed couples with the goal of helping them become less distressed and increasing the health of their relationships. Usually, couples therapy involves one therapist and one couple. However, couples therapy can be offered to individuals with the goal of helping the couple (see more about this below in terms of individual interventions that are designed to impact a couple). Thus, one of the key distinctions between relationship education and couples therapy is that relationship education is typically delivered in workshop settings to a variety of couples, whereas couples therapy is typically delivered to distressed couples, one couple at a time. Another important distinction is that relationship education is typically manualized, whereas therapy, as delivered in the community, is typically not manualized. Without a manual, practitioners are prone to following their own experiences and instincts (Dimond, Havens, & Jones, 1978), which may reduce the effectiveness of interventions. Relationship education’s manualized approach may make delivering research-based interventions more accessible to practitioners (especially those with little prior experience with couples.)

Nonetheless, as noted earlier, the distinctions between relationship education and couples therapy do blur at times. For example, relationship education can be part of therapeutic approaches (e.g., Birchler, Fals-Stewart & O’Farrell, 2008) and most research-based therapies combine education with counseling. For example, Cognitive-Behavioral Couples Therapy has skill training at its core, due to its origins in the late 1970’s with the work of Stuart (1969) and Weiss (e.g., Birchler, Weiss, & Vincent, 1975; Weiss & Heyman, 1997), followed up by the work of others such as Gottman, Notarius, Markman, Jacobson, Margolin, Baucom, Epstein, Christensen, Halford, Hahlweg, Snyder, and Stanley. In our own research-based couples therapy practices, we often teach the Speaker-Listener Technique, which involves relationship education and skills training (Markman, Stanley, & Blumberg, 2010). Similarly, the empirically-supported Integrative Behavioral Couple Therapy approach includes some communication skills training (Jacobson, Christensen, Prince, Cordova, & Eldridge, 2000).

One important difference between therapy and education is the approach to couples’ discussions. Therapists often interrupt an interaction between two partners in therapy to make their own interpretations and comments about the content of the discussion. This kind of direct intervention with a couple regarding the content of their interaction is much less common in relationship education. Relationship education tends to focus on teaching couples the skills needed to work out issues in general. In contrast, practitioners in couple therapy tend to help partners work on specific issues while using communication skills in the process. Thus, one way to illuminate differences between relationship education and couples therapy is to clearly indicate to relationship education participants that they will be receiving education and skill training. Participants in relationship education should also be informed that, they will not receive attention on specific issues or specific skills sets, as they would in couples therapy.

Another important distinction between relationship education and couples therapy is that therapy approaches are better equipped to handle the most serious problems in relationships. These problems can include affairs, aggression, and co-morbid problems such as with mental health, pornography, or substance abuse. Relationship education programs need to have a plan of action to address these more serious problems. For example, in our own relationship education work we give every attendee a referral document called Getting More Help When Needed (please email authors for a generic version). Finally, therapists and relationship educators can share clients, with each referring couples to the other. Based on our own work, we have the impression that going to a relationship education program opens the door to seeking more services when needed, however, data are needed to support these impressions.

The troubling news about couples therapy is that relatively few couples seek couples therapy, despite its advantages. For example, in one study of married couples, 19% of them received some form of couples therapy. Of those who had filed for divorce, 37% received therapy services beforehand (Johnson et al., 2002). Clearly, practitioners need to find a more effective way to reach more couples and reach them sooner, before the most serious and difficult problems to treat develop. In addition, when couples eventually do receive therapy, they typically do not receive evidence-based services (Johnson et al., 2002).

Finally, and perhaps most importantly, compared to couples therapy, more couples tend to receive relationship education. For example, a recent survey indicated that among those married since 1990, 44% of couples had received some form of premarital relationship education, typically provided in a religious organization (Stanley, Amato, Johnson, & Markman, 2006). Importantly, those entering second marriages are less likely to receive such services (Doss, Rhoades, Stanley, Markman, & Johnson, 2009). As Markman et al. (2009) conclude, relationship education provides the best way to reach diverse (including high risk) couples with evidence-based services.

Why is Relationship Education Important?

Although divorce rates have slowly declined in recent years (to a low point of 3.4/1000 in 2009), they continue to be high, with nearly 45% of first marriages ending in divorce (Raley & Bumpass, 2003; Tajada-Vera & Sutton, 2010). The negative impact of family instability is well documented and we refer readers to other reviews (e.g., Halford et al., 2003; Lebow, Current Volume). In brief, research indicates that adults who are divorced or unmarried tend to experience worse mental and physical health (e.g., Whisman, 2008). Furthermore, children who experience high levels of parental conflict or the dissolution of their parents’ relationship tend to fare worse on a range of outcomes from infant development to later adolescent social adjustment (Amato, 2001; Cummings & Davies, 1994; Grych & Fincham, 1990).

Demographic shifts

Because various forms of family instability are increasing (Lebow, 2010), we anticipate that over the next decade the negative effects of relationship distress may expand, affecting an even larger number of people. For example, although cohabiting relationships tend to be much less stable than marriage, more couples are choosing to live together before marriage, or not marry at all (Stanley, Whitton, & Markman, 2004). Furthermore, many children will experience significant family instability, given that 40% of children are born to unmarried parents (Ventura, 2009). Many other children will witness the divorce of their parents, and many will live with a parent who is cohabiting at some point while growing up (see Bumpass & Lu, 2000). Further, children whose parents are unmarried are exposed to a higher risk for abuse (Turner, Finkelhor, & Ormrod, 2007) and tend to have poorer outcomes (Brown, 2004).

Relationship education’s response to demographic shifts

The relationship education field has, in some ways, lagged behind these shifts in demographics and concerns about the negative effects of relationship distress and instability. Traditionally, relationship education services were generally limited to engaged couples or newlyweds. Recent meta-analyses of relationship education confirms that research on relationship education has also tended to focus on these groups (Hawkins et al., 2008; Blanchard, Hawkins, Baldwin & Fawcett, 2009). More recently, programs have been extended to unmarried couples who may or may not be planning marriages (such as the Building Strong Families project mentioned later in the paper) and to individuals who may or may not be in relationships (e.g., Antle et al., in press; Pearson, Stanley, & Rhoades, 2008). By reaching couples and individuals who would otherwise not have had access to relationship education these programs have helped address some of the changing family demographics, however, much more work needs to be done. The changes in the number of couples who live together or bear children without being married indicate that many adults form significant romantic unions long before they marry. If relationship education only reaches those in committed, premarital relationships, we miss opportunities to help some people make important relationship decisions. Thus, relationship education should be extended to individuals and couples earlier in their relationships.

Cohabitation

Additionally, it may be important to address cohabitation more directly in relationship education, as this is a fast growing demographic and an area that most relationship education programs do not address. Rhoades, Stanley, and Markman (2009) offer a detailed discussion of cohabitation and relationship education. Briefly, they suggest that relationship education with cohabiting couples needs to focus not only on teaching good communication skills, but also on helping couples consider and discuss their expectations for the future as well as specific psychoeducational material on commitment. They also suggest that relationship education delivered to individuals or youth should include information on cohabitation and, more broadly, about how to make healthy decisions in relationships.

Children

Similarly, given the increase in the nonmarital birth rate, increasingly more couples have children from previous relationships. As noted in an excellent review article by Higginbotham, Miller, and Niehuis (2009), traditional relationship education programs generally do not address issues specific to these couples who face unique issues such as children by previous partners, co-parenting, issues related to money and paying child support, and jealousy regarding communication with ex-partners. In a review of premarital preparation for couples entering remarriages, normalization and social support were two of the major factors couples found to be useful. Research-based programs specifically for remarital couples, building in part on the work of Higginbotham and colleagues, are now underway (Whitton, Nicholson & Markman, 2008, for a review). Thus, while the first decade of the new millennium has seen some major expansion in the reach of relationship education, both the content and the reach need even greater expansion to keep up with the changing makeup of families in the United States.

Before reviewing the studies that have been published from 2002 to the present, we first provide an overview of the conceptual model underlying some of the evidence-based relationship education programs.

Theoretical Rationale for Major Models of Intervention

In brief, the major theoretical rationale for most evidence-based relationship education programs (drawing from behavioral exchange and social learning theories) is that the quality of communication and conflict management early in a relationship is associated with the quality and the health of the relationship over time. Over the last several decades, several long-term studies following couples before marriage indicate that the seeds of future marital problems often present themselves in early interactions (e.g., Markman, Rhoades, Stanley, Whitton, & Ragan, 2010, also see Wadsworth & Markman, 2010). Thus, evidence-based relationship education programs focus on modifying interactions associated with future relationship outcomes (Halford, Markman, & Stanley, 2008).

The conceptual base of the research extends beyond the simple association between communication, conflict management, and future outcomes. For example, in the past decade the field has increasingly focused on both research and theory regarding the following: 1) positive factors as protective factors in relationships as well as the role of transformative processes in relationships, such as commitment, sacrifice, and forgiveness (see Fincham, Stanley, & Beach, 2007), 2) the link between marital quality and a variety of individual functioning dimensions, as reviewed earlier (e.g., Whisman, 2008), and 3) how parenting and child functioning are related to marital quality and vice versa (e.g., Cowan, Cowan, Pruett, Pruett, & Wong, 2009). Finally, as noted elsewhere (Markman, Rhoades, Delaney, White, & Pacific, 2009) evidence-based relationship education programs often follow a prevention science model that focuses on intervening on risk factors associated with future problems, as well as protective factors associated with future relationship health (see Coie et al., 1993).

Relationship Education Research Review

Here we provide a comprehensive review of the research published (or accepted for publication) since the last decade review article (2002-present). The bulk of the information from our review is included in Table 1. In this table we list the authors, date of publication, sample characteristics, sample size, intervention program, research design (i.e., control group, random assignment), assessment points, and outcomes.

Table 1.

Review of 2002–2009 Published Studies on Relationship Education

Study	Sample Characteristics	n	Intervention	Design	Assessment	Outcomes
Alqashan (2008)	Married Kuwaiti couples	32	5 4-hour sessions on communication and love	1 intervention group, 1 control group	2 time points (pre, post)	Experimental group displayed significant gains in marital communication and marital adjustment compared to control group post-intervention. No significant differences on affectional expression between control and experimental group post-intervention.
Antle et al. (in press)	Individuals with low income levels	202	Within My Reach (15-hour program delivered by community center staff)	Intervention group only	3 time points (pre, post, 6 month follow-up)	Participants reported high levels of satisfaction with the program, increases in knowledge, and improvements in communication, conflict resolution skills, relationship quality, and aggression. No significant differences observed between gender or racial groups.
Baucom et al. (2009)	Couples, wives with breast cancer	14	6 75-minute individual sessions, curriculum developed by authors	RCT, 1 intervention group, 1 no treatment control group	3 time points (pre, post, 12-month follow-up)	Medium effect sizes for improvements in relationship and individual functioning across measures for breast cancer patients and their partners at post-intervention and 1-year follow-up compared to controls. Cancer patients in intervention group reported significant improvements on medical symptoms at post-intervention and follow-up. Psychological distress decrease for cancer patients only at 1-year follow-up only (not at post-intervention).
Bodenmann et al. (2006); Pihet et al. (2007)	Middle-aged married or cohabiting couples	118	Couples Coping Enhancement Training (18 hours)	Matched control group design (not RCT)	5 time points (pre, post, 6-month follow-up, 1-year follow-up, 2-year follow-up)	Improvements in marital quality displayed across all time points for intervention group compared to controls. Stress communication was not significantly different between groups across time points. Improvements in supportive dyadic coping (by oneself and by one’s partner), negative dyadic coping, and common dyadic coping observed in intervention group compared to controls across time points. In general, effects partially attenuated by 2-year follow up.
Bouma et al. (2004)	Couples with hazardous drinking	37	Controlling Alcohol and Relationship Enhancement Program (6 90-minute sessions with therapists)	RCT, 1 intervention group, 1 active control group	3 time points (pre, post, 6-month follow-up)	Intervention group displayed observational (pre-post only) and self-report improvements in communication compared to controls. No significant differences in alcohol use or relationship adjustment reported between groups.
Braithwaite & Fincham (2007)	College students in dating relationships	91	ePREP (1-hour computerized version of PREP)	RCT, 1 ePREP intervention group, 1 mental health intervention group, 1 control group	2 time points (pre, post)	ePREP and mental health intervention groups showed improved relationship functioning and reductions in anxiety and depressive symptoms compared to controls post-intervention. ePREP group demonstrated reductions in negative affect scores, with no significant differences observed in positive affect. No significant differences were found between gender or ethnic groups.
Braithwaite & Fincham (2009)	College students in dating relationships	77	ePREP (1-hour computerized version of PREP)	RCT, 1 Intervention group, 1 placebo control group	3 time points (pre, post, 10-month follow-up)	ePREP condition produced better outcomes in anxiety measures, psychological aggression, and physical assault compared to control group. Findings suggest ePREP group displayed better relationship outcomes and is durable to relationship dissolution. ePREP group had no significant effect on depressive symptoms. ePREP group appeared to get worse at post-intervention regarding anxiety, physical assault, psychological aggression and constructive criticism and subsequently improved by 10-month follow-up.
Braukhaus et al. (2003)	Married couples who completed EPL-II (German version of PREP)	62	2 booster sessions	Couples self selected booster or no booster	3 time points (pre, post, 1-year follow-up)	All couples displayed improvements in communication from pre-post time points. Couples who participated in booster session displayed significantly higher marital satisfaction and fewer problem areas than couples with no booster sessions at 1-year follow-up.
Busby et al. (2007)	Dating couples, mostly in college	79	6 sessions, Relationship Evaluation (RELATE)	RCT, 1 workbook-only group, 1 workbook plus therapist group, 1 RELATE plus facilitator group	3 time points (pre, post, 6-month follow-up )	Workbook-only group displayed relatively stable relationship functioning scores across all time points. Therapist-directed group displayed similar scores from pre to post time points, followed by drops in relationship functioning measures by the 6-month follow up. RELATE group displayed scores indicating improvement across all time points. By 6-month follow-up RELATE group had better scores on relationship satisfaction and communication compared to therapist-group, and better scores on problem areas compared to therapist-group and workbook-only group.
Carson et al. (2004)	Married or cohabiting couples	44	8 sessions of mindfulness-based relationship education	RCT, 1 intervention group, 1 waitlist control group	3 time points (pre, post, 3-month follow-up)	Intervention group displayed better scores on relationship satisfaction, autonomy, relatedness, acceptance of partner, and relationship distress compared to controls. Intervention group displayed improvements in optimism, individual relaxation, and psychological distress compared to controls. Of participants in intervention group, those who practiced more mindfulness displayed greater improvements in happiness, relationship stress, stress coping efficacy, and overall stress. Intervention group maintained benefits at 3-month follow-up.
Cordova et al. (2005)	Married couples, 77% of whom were at-risk for distress	74	Marriage Check-Up (4 hours of assessment and feedback with a therapist)	RCT, 1 intervention group, 1 control group	2 time points (pre and post)	Intervention group displayed improvements in relationship distress, acceptance, intimacy, and motivation to change compared to control group at post-intervention.
Duffey et al. (2004)	Married couples	108	Dream sharing (4-hour workshop), event sharing (4-hour workshop)	RCT, 2 intervention groups, 1 control group	2 time points (pre, post)	Dream and event sharing groups reported improvements in intimacy and marital satisfaction compared to controls at post-intervention.
Duncan et al. (2009)	Married couples, mostly members of the Church of Latter-day Saints	100	12 hour curriculum developed by the authors	RCT, 1 website intervention group, 1 in person intervention group, 1 control group	3 time points (pre, post, 3-month follow-up)	Website and in-person groups reported higher relationship satisfaction and increases in empathetic communication than control group. No significant differences in magnitude of change found between website and in-person intervention groups. No significant results reported on changes in love, criticism, contempt, stonewalling, flooding, or soothing.
Einhorn et al. (2008)	Men and women in prison	254	Prison PREP (12 hours)	Intervention group only	2 time points (pre, post)	Intervention group reported improvements in dedication, confidence, friendship, satisfaction, and self-report communication compared to controls at post-intervention. Intervention group displayed decreases in negative interactions and feelings of loneliness at post-intervention compared to controls.
Feinberg & Kan, (2008)	Expecting couples	155	Family Foundations (8 group sessions)	RCT, 1 intervention group, 1 control group	2 time points (pre, post)	Intervention group reported more positive coparenting, decreases in maternal depression, and decreases in parent–child dysfunctional interactions compared to controls at post-intervention. No significant effects on coparental undermining, parenting-based closeness, anxiety, or paternal depression were observed.
Halford et al. (2010)	Expecting couples	71	Couple CARE for Parents (CCP) (workshop, home visit, self-directed combination) or Becoming a Parent (BAP) (maternal parenting educational program)	RCT, 1 CCP group, 1 BAP group	3 time points from last trimester to 12 months postpartum	At post-intervention, the CCP group displayed more decreases in negative couple communication, with results partially attenuating by 1 year follow-up. CCP prevented declines in relationship adjustment and self-regulation in females. Positive adjustment to parenting observed in both groups with no significant differences. CCP reported higher consumer satisfaction than BAP.
Halford et al. (2004)	Non-distressed married or cohabiting couples	59	Couple CARE (self-delivery with therapist telephone call)	RCT, 1 intervention group, 1 waitlist control group	2 time points (pre, post)	Compared to controls, CARE group displayed improvements in partner and self-regulation, but only for women. No significant changes in self-regulation were observed in men. CARE group displayed increases in relationship satisfaction and stability. No significant differences in negative couple communication were reported between groups.
Hawkins et al. (2006)	Expecting couples	155	Marriage Moments (delivered during child birth classes)	RCT, 2 intervention groups, 1 control group	4 time points from second trimester to 9 months postpartum	Participants reported high satisfaction with the programs. No significant effects observed between control group and intervention groups on marital virtues, marital satisfaction, or the transition to parenthood.
Holt-Lunstad et al. (2008)	Married couples	34	Education about warm touch	RCT, 1 intervention group, 1 control group	2 time points (pre, post)	Intervention group displayed enhanced salivary oxytocin levels and reduced alpha amylase at post-treatment. Husbands in intervention group displayed significantly lower 24-hour systolic blood pressure than controls. No significant difference in systolic blood pressure was reported between intervention group wives and control group wives at post-intervention.
Kotrla & Dyer (2008)	United States Air Force, married couples	23	Skill trainign for Active Military (weekend retreat format)	Intervention group only	3 time points (pre, post, 2-month follow-up)	Improvements in satisfaction, communication, conflict resolution, negative interactions, and commitment at post-intervention and 2-month follow up compared to pre-intervention measures. No significant changes in hope for success of present relationship at post-intervention or 2-month follow up.
Larson et al. (2007)	Premarital couples	39	Online Relationship Evaluation (RELATE)	RCT, 2 intervention groups (Self-interpretation group and Therapist-Assisted group), 1 control group	2 time points 60 days apart	Interpretation of results by a therapist had significant positive effect on relationship satisfaction, commitment, and opinions/feelings/readiness for marriage from post-intervention to 2 month follow-up. Increased awareness of strengths and challenges, communication, and expectations of prevention of future relationship problems. Intervention group without therapist resulted in initial drops in relationship satisfaction post-intervention followed by improvements by 2 month follow-up.
Laurenceau et al. (2004)	Premarital couples	217	PREP	RCT, 1 intervention group delivered by university staff, 1 intervention group delivered by religious organization staff, 1 treatment-as-usual group	3 time points (pre, post, 12-month follow-up)	No changes in marital satisfaction over post and 12 month follow-up, and no significant differences between groups on marital satisfaction trajectories. Improvements in observed negative communication for all groups, with wives in religious-organization group showing largest improvements. Largest decreases in negative communication observed between post and follow-up for all groups. Naturally occurring group and university group displayed decreases in positive communication, most noticeably between post and follow-up time points. Religious organization group did not display significant changes in positive communication over time.
Owen & Rhoades (in press)	Court-ordered parents	20	Working Together Program (12 hours on conflict management and co-parenting)	Intervention group only	3 time points (pre, post, 2-month follow-up)	Participants displayed improvement in co-parent relationship functioning and confidence from pre-post and maintained improvements at 2-month follow-up. reductions in conflict in front of children. Over hostility decrease from pre-post and was maintained at 2-month follow-up. Negative communication did not significantly change from pre-post, but improved by 2-month follow-up. Only women reported reductions in negative communication with co-parent from pre to 2-month follow-up.
Schilling et al. (2003)	Premarital couples	39	PREP (weekend retreat format)	Intervention group only	Pre, post, annual follow-ups for 5 years	Pre-post increases in positive communication and decreases in negative communication. Couples maintained premarital satisfaction levels during first 3 years of marriage. Male improvements in communication predicted decreases in female distress, however female increases in positive communication predicted post intervention distress in both genders. Higher female assertive negative communication associated with better marital outcomes.
Schulz et al. (2006)	Expecting couples	66	P. A. Cowan & C. Cowan (2000) psychoeducational groups	RCT, 1 intervention group, 2 control groups	5 time points between last trimester to 66 months postpartum	Intervention group showed less decline in marital satisfaction than controls across all time points.
Shapiro & Gottman (2005)	Expecting couples	38	2-day workshop developed by authors	RCT, 1 intervention group, 1 control group	3 time points from last trimester to 12 months postpartum	Intervention group displayed increases in marital quality, decreases in postpartum depression, and decreases in observed hostile affect for both genders by 1 year follow-up. Workshop group displayed worsening in marital quality, postpartum depression and wives’ hostility at 3 month follow-up before improving by 1 year follow-up.
Stanley et al. (2005)	Married couples in which at least one member was active-duty in the U.S. Army	230	PREP delivered by Army chaplains (two days of workshop)	Intervention group only	2 time points (pre, post)	Improvement in relationship satisfaction and self-reported communication. Changes over time were not significantly different between gender, racial, or income groups.
Stanley et al. (in press)	Married couples in which at least one member was active-duty in the U.S. Army	476	PREP delivered by Army chaplains (14 hours)	RCT, 1 intervention group, 1 control group	3 time points (pre, post, 1-year follow-up)	Lower divorce rate for intervention group at 1-year follow-up
Van Epp et al. (2008)	Single Army soldiers	272	PICK (5-hour program delivered by Army Chaplains)	No randomizatio n, 1 intervention group, 1 comparison	Post survey only	High satisfaction with the program. Increases in the importance intervention group participants placed on getting to know their partners. Some perceived increase in knowledge about relationships.
Wood et al. (2010) (Oklahoma Study)	Low-income couples who are expecting or recently had a child, mostly unmarried.	1,010	Oklahoma site of Building Strong Families Study, Family Expectations delivered by trained leaders (20 hours of group sessions), individual support from family coordinators, referral to support services	RCT, 1 intervention group, 1 control group	2 time points (baseline, 15-month follow up)	Increased relationship length, happiness, support, and better conflict control. Improved relationship quality in African American couples.
Wood et al. (2010)	Low-income couples who are expecting or recently had a child, mostly unmarried.	5,102	Building Strong Families Study, 8 US locations, delivered by trained leaders (30–42 hour group sessions) individual support by family coordinators, referral to support services	RTC, 1 intervention group, or 1 control group	2 time points (baseline, 15-month follow up)	Overall results: No significant effects on relationship length, relationship quality, relationship satisfaction, or conflict behaviors. BSF fathers more likely to receive support services. Increased relationship quality among African Americans in 6 BSF Programs. Specific Locations: Baltimore location negatively affected relationship length and co-parenting quality. See “Oklahoma Study” for positive results.

Descriptive Summary of What We Found (See Table 1)

We found 30 studies which evaluated a total of 21 programs. The results generally indicated that relationship education programs as compared to a variety of control groups were successful in improving outcomes ranging from communication to marital satisfaction (adjustment) to individual function. We do not elaborate on the specifics of the findings here because they are reviewed in detail in the table and two recent meta-analyses also provide an excellent overview of findings (Blanchard et al., 2009; Hawkins et al., 2008). These meta-analyses report, for example, an effect size of .44 for communication measures at post–test as well as follow-up. These effect sizes, though somewhat lower than those that are found for treatment studies, are impressive given that the majority of the studies are prevention studies. Interestingly, stronger findings emerged for more rigorous designs (e.g., randomized controlled trials).

Overall, while the outcome findings were both interesting and important, we feel that the most important set of findings in the past decade emerged from two large-scale projects. First, a study on the effectiveness of the Prevention and Relationship Enhancement Program (Markman et al., 2010) in the U.S. Army indicated an effect of relationship education for divorce. Divorce is rarely studies in research on relationship education because of the necessity of large samples and long-term data. In that study, relationship education reduced the risk for divorce two-fold, as measured one-year after the intervention (Stanley, Allen, Markman, Rhoades, & Prentice, 2010).

One other particularly important study conducted during the past decade is the federally-funded Building Strong Families study (Wood, McConnell, Moore, Clarkwest, & Hsueh, 2010). In the first report on this large-scale (N = 5,102), eight-site randomized controlled trial of relationship education provided to unmarried couples who are having a child together (see www.buildingstrongfamilies.org). Preliminary results from the Building Strong Families study are generally disappointing, with very few positive effects of relationship education and even some negative effects (Wood et al., 2010). When sites were analyzed separately, only one site showed significant positive findings across many of the relationship quality variables. This one site, located in Oklahoma City and called Family Expectations, integrated PREP with a program called Becoming Parents (Jordan, Markman, Stanley & Blumberg, 1999). The results showed that participating couples were more likely to stay together; had higher levels of happiness, support, affection, and fidelity, and were better at parenting than couples in the control group.

One reason Oklahoma’s program may show positive outcomes while other do not is that they have a higher-than-average rate of intervention completion. This high rate of completion may be partly attributable to the fact that Oklahoma uses material incentives for program participation. For example, couples earn “Crib Cash” that they can spend at an on-site store for session attendance (see http://www.familiesok.org/thecrib.html).

Another possible reason for the success of the Oklahoma’s Family Expectations program may be their innovative implementation of post-workshop services. These services included booster sessions as well as planned activities for couples and families, such as a holiday party every December for program participants. Such services should be strongly considered in future relationship education programs and their effects should be evaluated in research over the next decade.

In addition, the very strong organizational structure supporting the delivery of Family Expectations likely plays an important part in the outcomes achieved. The program is well-staffed and run professionally. More generally, stronger administration of large scale community-based relationship education efforts will likely be associated with stronger outcomes. For more discussion of the Family Expectations program and findings from the Building Strong Families evaluation specific to Oklahoma, see Devaney and Dion (2010).

Methodological Strengths and Weaknesses

In this section we use the format in Table 1 as much as possible to review the strengths and weaknesses of the studies conducted on relationship education from 2002 to 2009.

Sample characteristics

The samples described in Table 1 included a mix of premarital and marital couples, with most studies either not distinguishing between the two or combining the two groups in analyses. Many of the studies included cohabiting participants, but most did not indicate how many, nor did they look at differences between married and cohabiting participants. When information on couples’ satisfaction was provided, many studies used cutoffs to eliminate distressed couples from the samples.

More generally, the couples included in the research reflect the increasing diversity of the field. Studies now include evaluations of interventions for couples with low income levels, unmarried couples with a child together, couples in which one partner has a medical problem, military couples, foster and adoptive parents, step-families, and couples with children. Thus, good progress continues on the aforementioned recommendations by Halford et al. (2003) in the last review paper of increasing the accessibility of services for high risk and diverse populations. There have been major strides in the last decade to offer and evaluate services to couples with low-income levels, including the large scale, multi-site evaluation mentioned earlier (Wood et al., 2010; for a full review see Hawkins & Ooms, 2010). Also consistent with the recommendations from the last review paper, we found evaluations of programs that offer services during transition periods.

Nevertheless, wide gaps exist between the diversity of those who are participating in research and whom we are serving in practice. That is, the participants receiving services are not generally included in the studies we have reviewed. Many populations exist that are either underserved or not served at all including: older couples; gay, lesbian, or transgendered couples; separated and divorced people; cohabiting couples; and single people searching for a relationship. Recently, Whitton and her research team have started to evaluate a relationship education program for gay couples (e.g., Whitton & Buzzella, 2011). We hope that the field will continue to reach out to diverse populations in terms of both delivery of research-based services and the evaluation of these services using rigorous research design in the next decade.

Interventions

Several different intervention approaches were evaluated in the last decade. Although increasing diversity in the content of and approach to relationship education may be positive in some ways, some of these programs are not grounded in theory and/or research. Furthermore, many of the studies in Table 1 were evaluated by researchers who developed, and sometimes delivered, their own interventions. These kinds of research designs may lead to bias in the research. One solution to this potential bias is to have other research teams evaluate programs.

In addition, despite the strong recommendation made by Halford et al. (2003) in the previous decade review programs explicitly target the prevention of relationship aggression, most interventions fail to do so. Teaching couples conflict management skills in an education program is not to be confused with “anger management” treatment programs, which are explicitly designed to treat intimate-partner violence. There is one study which suggests that learning conflict management may reduce the risk of later aggression (Markman & Hahlweg, 1993) but relationship education hoping to prevent or reduce aggression need to go beyond teaching conflict management skills (e.g., helping those in unsafe relationships end them, teaching anger management skills, changing attitudes about women, and teaching attendees about healthy relationships). Some of these interventions will need to focus on partners individually, rather than having them attend together (see Rhoades & Stanley, in press for more discussion).

A larger problem for this field is that most relationship education services that are delivered are not being evaluated at all. Thus, linkages between practices as defined by Halford et al. (2003), current practice, and research on current practice, remain separated by wide gaps that need to be bridged in the next decade.

Design

As indicated in Table 1, there is much diversity in the methodological quality of the studies. On the positive side, there have been several randomized clinical trials (RCTs) implemented in the last decade. This kind of design is the strongest in terms of establishing effects of an intervention (more on RCTs below). Though we did not examine effect sizes, meta-analyses conducted by others (Blanchard et al., 2009; Carroll & Doherty, 2003; Hawkins et al., 2008) indicate that the most powerful effect sizes emerge from the best methodologically-designed research. The majority of the studies conducted in the last decade used weaker designs. In terms of the type of control group, most studies used a no-intervention control group, and the second most frequently used control group was an alternative intervention. There were no placebo-control group designs to control for attention and expectations, which could provide alternative explanations for positive effects. In research conducted prior to the studies reviewed in the current paper, several studies used placebo controls such as reading a relationship book (e.g., Halford, Sanders & Behrens, 2001) or watching and discussing a movie (Rogge, Cobb, Johnson, Lawrence, & Bradbury, 2002). In general, relationship education programs have outperformed placebo control groups.

Measuring outcomes

The majority of the studies revealed positive effects of relationship education on key indicators of relationship quality including communication quality, conflict management skills and relationship satisfaction. However, there were a paucity of studies that measured communication using observational measures, which are the gold standard in terms of assessing communication (Markman & Notarius, 1987; Weiss, Heyman, Halford, & Markman, 1997). In addition, very few studies measured other important dimensions of relationship quality including measures of protective factors, such as commitment, friendship, and passion.

Follow-ups

The absence of long-term follow-up is notable. More studies need to include longer-term follow-ups, as the vast majority of studies assess outcomes only at post-test. This is a major problem, because the goals of prevention programs are by definition long-term in nature. In addition, there need to be at least three data points to apply state-of-the-art growth curve analyses to evaluate change over time. Without more assessment points, non-linear effects in relationship education cannot be captured. Perhaps even more importantly, future research needs to document how long the effects of relationship education last and if “sleeper effects” emerge over time. In general the data that do exist show that long-term follow-up trajectories display a tendency toward attenuation (e.g., Bodenmann, Pihet, Shantinath, Cina, & Widmer, 2006), highlighting the need for booster sessions and other post-intervention services.

Only one study evaluated the use of booster sessions (Braukhaus, Hahlweg, Kroeger, Groth, & Fehm-Wolfsdorf, 2003), and it found promising results. However, these findings require replication with a control group and in a randomized clinical trial. In general, although most researchers recommend booster sessions, few programs use them, fewer studies evaluate them, and those that do have trouble persuading couples to participate (e.g., Pregulman, Rienks, Markman, Wadsworth, Einhorn & Moran, in press).

Suggested Program of Research and Interventions with Diverse Populations

Many of the studies reviewed in Table 1 are not incorporated into a systematic program of planned research. Here, we offer a model for researchers who want to develop a program of research consistent with the best practices for relationship education research. The first step is to start with a population of interest (preferably a high-risk population) and then develop, use, or adapt a relationship education program for these couples and/or individuals. The next step is to pilot the intervention and then use a pre-post, no control group design, to see if there are effects over time and if it is acceptable to the population of interest (see Markman et al., 2004, for an example). Then, research teams can move first to quasi-experimental studies, then to randomized clinical trials, and finally to dissemination trials. It is also important to have other researchers cross-validate findings to ensure that results are not biased.

A stage that is often missed and that is relatively easy to implement is to conduct “effectiveness” studies where real world use of the intervention is evaluated through assessments on outcomes of interest from clients, service providers, and organizations. Some research teams have been successful in moving studies in a laboratory setting (i.e., efficacy studies) to testing disseminations of intervention with different populations in real-world settings (Hahlweg, Grawe, & Baucom, 2010). While more research is needed, evaluations of dissemination studies have yielded promising findings (e.g., Markman et al., 2004).

Thus, there is emerging evidence from the past decade that research-based programs developed in university settings are transportable to a number of community settings (see also Wood et al., 2010). The transportability of relationship education constitutes one of the major accomplishments in the field because we have now reached hundreds of thousands of people since the last review, including close to a quarter of a million people in Oklahoma alone (Ooms, 2010).

Mechanisms of change

Progress has also continued to be made in investigating mechanisms of change for relationship education interventions (see Wadsworth & Markman, 2010, for a review). To show that a mechanism of change is operating, evaluations first must show that the targets of the intervention (e.g., communication quality) change from pre to post and then must show that changes in the target variable(s) are associated with other changes of interest (e.g., improved relationship satisfaction, lower divorce rate). The findings presented in Table 1 and meta-analyses (e.g., Hawkins et al., 2008) indicate that the target variables of interventions do typically change. Most of the studies reviewed in Table 1 (as well as studies from prior decades) show that couples improve their ability to communicate from pre to post assessment compared to control groups. However, mixed findings emerge when researchers examine the extent to which changes in such target variables are associated with other outcomes of interest. In at least one case, the findings were not in the predicted direction (see Schilling et al., 2003). However, these results may be attributable to correlations between husband and wife interactions that were not taken into account (Stanley et al., 2007). Nonetheless, the existing body of research on mechanisms of change in relationships leaves much to be discovered. Future work should continue to attempt to identify why programs are successful and what the key ingredients are to helping couples and individuals improve their relationships. The relationship education (and marital therapy) field can benefit from following the lead of the conduct disorder and substance abuse fields both of which as demonstrated stronger evidence for mechanism of change of therapeutic interventions in these areas.

Moderators

Moderators are also very important to consider when evaluating relationship education programs. Here, the questions revolve around the extent to which a program works differently and produces different outcomes for one group versus another. A major issue in the field that has emerged in the last decade, for example, is whether relationship education programs work better for high risk (e.g., children of divorce and aggression) versus low risk couples (Halford et al., 2008). The Building Strong Families study and the related report specific to Oklahoma’s site also identified some potential moderators of effects, particularly race (African American or not) and education (see Devaney & Dion, 2010; Wood et al., 2010). The answers to questions about which groups benefit more or less are decidedly mixed and moderators remain a burning topic for the next decade to tackle. Wadsworth and Markman (in press) provide a more detailed discussion of moderator effects.

Issues in Conducting Randomized Clinical Trials (RCTs)

Conducting a RCT is easier said than done and here we very briefly cover some of the issues in designing and implementing a RCT in the relationship education field. One of the most interesting research issues in the field is how to design a randomized clinical trial. We next describe the approach to RCTs that the Administration for Children and Families has taken in their two large-scale RCTs of relationship education. We focus on these projects because their very large sample sizes and in-depth measurement mean that they have the power to make a significant impact on this field in the coming decade. Just as there are many ways to have a good relationship, there are many ways to do a very good randomized clinical trial.

Both of these studies focus on the effectiveness of relationship education delivered by community organizations to couples with low-income levels. Instead of focusing on the evaluation of one specific program or intervention, they each use a multisite design in which each site chose it’s own program. Thus, these projects are evaluations of relationship education in general, not a particular approach to relationship education (see www.buildingstrongfamilies.org and www.supportinghealthymarriage.org for more details).

With regard to the design of these RCTs, both projects include random assignment of couples to either relationship education or a no-treatment control group. The Building Strong Families RCT (Wood et al., 2010) described earlier used an “intent-to-treat” approach to analyses that is different what is often used in this field. An intent-to-treat approach is more common in studies evaluating the implementation of a program (rather than the effectiveness of a specific curriculum) and it involves analyzing data from all participants as they were randomized, regardless of whether they completed the intervention. Importantly, across the eight sites involved in this study, 39% of couples who entered the study did not attend any sessions at all, which may have reduced the impact of the program.

Another characteristic that different about these studies versus most studies in this field is that neither of them conducted what are typical pre assessments, wherein which key outcome variables are measured before the intervention. Instead, they collected baseline information on demographics and are now collecting data at planned intervals post-intervention.

The rationale for not conducting pre-assessments is that random assignment of couples to conditions assures that all pre-test variables, if collected, would be equal across groups. Avoiding a comprehensive pre-assessment also decreases the risk of the assessment serving as an unmeasureable intervention in itself. Under these assumptions, comparing groups at follow-up assessment points will answer the research questions effectively. Although this makes sense, uncertainty remains because pre-test differences may exist due to unplanned selection effects or chance occurrences (e.g., when a coin is tossed 10 times, sometimes it comes out to 8 heads and 2 tails). In addition, because there are no pre-post data, questions regarding mechanisms of change cannot be directly investigated.

In our own work, even with random assignment, we sometimes have significant pre-test differences. However, we know these differences exist and we have the option of dealing with the differences in our analyses. Using propensity scores to handle group differences in quasi-experimental designs may also be used in RCTs when pre-test differences exist (David Atkins, personal communication, May, 2010). Further discussion as how best to deal with such differences is beyond the scope of this paper.

Another issue in RCTs involves the recruitment of couples. Three strategies have been used in the field: 1) Recruit couples for basic research and after couples complete the basic research (which provides pre-assessment data), then randomly assign them to intervention or control group(s) and tell them about the intervention. In other words, couples do not know anything about the intervention until after completing the basic research. Researchers using this strategy tend to have a large “decliner” group since couples did not volunteer to be in the research knowing there was an intervention option and thus many will not be interested. 2) Recruit couples for intervention and then tell them about the research after random assignment. In this scenario, you might get more attrition from couples in the control group since they were denied an intervention in which they were interested. 3) Recruit couples for an intervention study and tell participants about all the research conditions so that they agree to participate in whatever group to which they are assigned. In this approach, assignment to a group happens after pre-assessment. In all of these approaches, pre-assessment must occur before condition assignment so that researchers avoid biases associated with knowing one’s condition. However, under the third approach, participants may still slant responses one way or the other, depending on which group they would like to get into. True placebo control designs may mitigate this issue if all interventions appear equally plausible.

Researchers must remember that random assignment does not guarantee representation of the population. Clearly, the strongest studies randomly select from the population of interest (e.g., all couples who live in the United States) and then randomly assign participants to intervention and control conditions. To our knowledge, such a study has not yet been done, but it would be a highly valuable contribution to this field.

How the Field has Changed Since the Last Review

In reviewing how the field has changed in the last decade, we focus on funding and social policy initiatives, the general practice of relationship education, and the research we have reviewed.

Funding and social policy

A significant development in the last decade has been the creation of several large-scale programs that have led to the widespread use of relationship education programs in the community. The first of these programs is the federal Healthy Marriage and Fatherhood Initiative, funded by the Administration of Children and Families (ACF), which has provided over $150,000,000 for services in the form of 5-year grants beginning in 2006 and ending in 2011. (Note: New funding for healthy marriage programming in 2011 was announced in December 2010 with the reauthorization of the Temporary Assistance for Needy Families program, although, at the time this paper went to press, the ACF has not yet indicated how these funds will be allocated.) This federal Healthy Marriage and Fatherhood Initiative also funded the two large-scale evaluation studies mentioned earlier.

A second development in the last decade is that several states now have healthy marriage initiatives of their own (see the National Healthy Marriage Resource Center’s website for more information: www.healthymarriageinfo.org/programs/index.cfm). The state that has devoted the most energy toward the goal of increasing access to relationship education is Oklahoma. Since that state’s marriage initiative began in 1999, over 240,000 residents of Oklahoma have received relationship education, representing the largest dissemination project related to relationship education this field has ever seen. For more information about this program, see Dion (2006) and Hendrick (2008). For a discussion of related policy implication of initiatives like Oklahoma’s, see Amato and Maynard’s work (2007).

A third large-scale community program is the U.S. Army’s Strong Bonds program that offers relationship education throughout the Army. Services within this program are delivered by Army clergy members to all levels of Army personnel and their families (see: www.strongbonds.org). These two programs have greatly increased the visibility of relationship education for policy makers and they have led to very widespread dissemination of relationship education activities.

One issue related to these new, large-scale dissemination programs is whether programs are research-based and empirically-tested. PREP remains the only program that has been rated as empirically-based and tested by the Substance Abuse and Mental Health Services Administration (SAMHSA; see http://www.nrepp.samhsa.gov). Evaluations of programs that are being widely used, including those funded under federal and state initiatives, are needed and should be a high priority in the field because many couples are now receiving services that have not been adequately tested.

Innovations in research and practice

There has been an increase in the diversity of programs offered to couples and individuals, including many based research-based relationship education programs like Couples Communication (Miller, Nunnally & Wackman, 1976) Relationship Enhancement (Guerney, 1977), Couple Care (Halford, 2011), and PREP (Markman et al., 2010). There are also new programs that represent creative variations on traditional relationship education themes, such as a mindfulness-based intervention (Carson, Carson, Gil, & Baucom, 2004) and individual interventions designed to both enhance couples relationships (Markman et al., 2009), as well as to help partners exit unhealthy relationships and improve their chances for successful mate selection (Rhoades & Stanley, 2009). A fuller discussion of the important trends in the relationship education field with regards to providing programs for individuals (both in and out of relationships) is included below.

We noted in our research review that there was an increase of evaluations on the impact of relationship education services as delivered to couples who are more diverse than the White middle-class couples who were typically the focus of earlier studies. At the same time, services are being provided to many other groups and these services have not been evaluated. Such groups include: partners in same-sex relationships, individuals in prisons, students in middle school and high school, refugee couples, and couples in other countries. Thus, more evaluations of the services now being delivered need to be a high priority for the field.

In terms of diversity in modes of service delivery, in the last decade, the relationship education field has provided services over the telephone and over the Internet, as well as the provision of relationship coaching over the phone (e.g., Halford, 2010; Loew et al., 2011), We have also increased diversity in where services are delivered, with the most important changes being in the provision of services in the communities in which people live. We have increased the diversity of service providers, such that service providers are often now trained to provide services in a particular setting while delivering a specific program (Markman et al., 2009). We suggest elsewhere that the “messenger matters” and that service deliverers who know the couples and their experiences and backgrounds may do a better job than even trained professionals (Markman et al., 2004). For example, in our work with the U.S. Army evaluating PREP for Strong Bonds (e.g., Stanley et al., 2010), the service providers are chaplains, and many go to war with the soldiers in the program. Such connections of service providers to participants in a program likely contributes to higher levels of alliance with leaders, and higher alliance has been linked with more positive outcomes (Owens, Rhoades, Stanley, & Markman, in press).

Thus, the field is beginning to provide services to couples who desire and need our services, regardless of ability to pay. The services are delivered in their community by a service provider trained specifically for the program and who is either a community member or knows the audience well. All of the above ideas are consistent with the philosophy and principles of the Community Mental Health Center Movement that was fueled by the passage of the Community Mental Health Centers Act in 1963. Interestingly (and sadly) the Community Mental Health Center Act was the last piece of legislation John F. Kennedy signed before he was tragically assassinated less than a month later. A more detailed review of the links between the relationship education and the Community Mental Health Center Act is provided by Markman et al. (2009).

Issues and Recommendations

In this section, we highlight the major issues from our review and make recommendations for the next decade of research and practice. We use a best practices framework to compare what we recommended 10 years ago to the current state of the field and future goals.

Engaging men

Research to date has suggested that virtually every group that has had an opportunity to participate in relationship education programs have expressed high levels of interest and participation, though there are barriers unique to each group (e.g., childcare for children). There are also common barriers for all couples, such as getting both partners “in the room, ” so that they can receive services at the same time. Many of the common barriers are discussed by Pregulman et al. (in press). Here we focus on one of biggest issue in the field --motivating men to participate in research and interventions. While relationship education has been particularly appealing to men given the focus on learning skills, ground rules, and talking without fighting (Ooms, 2010), there is still a lot of work to be done. For example, one barrier for men is their concern about sharing feelings and dealing with past issues. In our work in PREP, we tell partners that they are not expected to share personal information with the other couples in the room and that talking about personal issues is only done with each other. In addition we tell couples that we focus on current and future issues, rather than stirring up the past. Also appealing to some men is our approach of focusing on being aware of the importance of the relationship as well as deciding rather than sliding when it comes to relationship decisions. Finally, we have found that using male/female presenting teams provides another way of increasing men’s participation.

Enhance motivation and consider incentives

A key issue in the relationship education field involves motivating people who are not currently experiencing problems to seek services or take advantage of services. As we noted before, it is hard to convince couples who are experiencing problems to seek help, so it is often even harder to convince people not experiencing difficulty to seek services (also see Doss et al., 2009).

One very controversial solution to the motivation issue facing our field is the issue of providing incentives. Many people in the field think that it is inappropriate from a research or service perspective to pay couples for their participation in interventions. However, incentivizing couples seems to be working in Oklahoma (see earlier discussion and Devaney & Dion, 2010). While data on the cost-benefit analyses of incentives are not yet available, one could argue that for every dollar invested in having couples complete a relationship education program (program costs plus any incentives), many times this amount would be saved in terms of health care costs, divorce, child behavior problems, and work loss (Turvey & Olson, 2006 ). Thus, relationship education is an excellent investment for a community.

Study leader effects and the alliance between leaders and attendees

There is a plethora of research in the psychotherapy literature indicating that the quality of the relationship between clients and therapists determines outcomes (e.g., Pinsof, Zinbarg, & Knobloch-Fedders, 2008). However, we have a paucity of studies in the relationship education literature on the topic of alliance. One recent study indicates that the working relationship between relationship education providers and the couples they serve is important in understanding the impact of relationship education (Owen et al., in press). More work in this area in clearly needed.

On a related note, there are few studies that evaluate the impact of specific leader characteristics on outcomes. Owen et al. (in press) demonstrated that different leaders have different degrees of impact, but the specific characteristics that make individuals good leaders for relationship education are relatively unknown. There are a large number of questions here both theoretical and pragmatic. For example, how important is the match between leaders and attendees on factors such as gender, age, ethnicity, marital status, or sexual orientation? To what extent is co-leading relationship education better, worse, or similar, to it being delivered by one person? Do programs delivered by at least one male presenter promote men’s engagement in relationship education? Having said all of this, our experience leads us to conclude that the competency and enthusiasm of the presenter is more important than any specific leader characteristics, but this is an empirical question.

Study moderators and mediators of program impact

As noted earlier, we know very little about the degree to which the proposed mechanisms of change in relationship education programs (e.g., communication skills) are associated with longer-term outcomes of interest (e.g., divorce, child functioning). Similarly, we know very little about factors that moderate program effects. Research in the next decade should focus on these very important issues (see Wadsworth & Markman, in press, for a fuller discussion).

Employ observational measures

Only a handful of the studies reviewed included observational measures to assess intervention outcomes. Given that the goal of most interventions includes changing couples’ interactions, and that self-report measures of communication provide only one perspective on interactions, the omission of observational measures precludes fully answering basic questions about the effects of relationship education.

Evaluate dissemination efforts

There need to more studies of the effects of the dissemination of relationship education. One important metric of such evaluations would be the use of programs by those trained in them. For example, in one study, the majority of clergy trained in PREP were using at least parts of the program up to 5 years after they were trained (Markman et al., 2004).

In summary, the field of relationship education has substantially advanced in the last decade. In particular, the field has made strides toward achieving some of the “best practices” in relationship education recommended in the last decade review, including increasing access to programs, reaching out to high risk and diverse populations, and including information on healthy relationships. Paraphrasing Stanley (2001), we know a great deal about what works to help couples through relationship education, but we still need to know more. Hopefully this paper has set forth a roadmap for the field as we move forward.

The Integration of Relationship Education and Couples Therapy: A Conversation with the Authors

We conclude by providing our personal perspectives on the relationship education field, through the lenses of our dual roles as therapists and educators. We both are relationship education program developers, researchers, trainers, and service providers as well as practicing couples therapists and supervisors of trainees in couples therapy.

Who We Are

HM: I am a licensed clinical psychologist, and for 33 years I have had a relatively small private practice seeing couples and individuals using empirically-supported interventions. My primary clients include couples, but I also sometimes see individuals. More recently, I have begun providing additional therapy for men in more individually-oriented therapy in conjunction with therapy for the couple. I am also the founder of PREP and I direct and provide supervision for a couples therapy clinic at the University of Denver.

GR: In addition to my research work at the University of Denver, I also have a small private practice in which I see couples, individuals, and children. In my practice, I integrate relationship education into many of my couple, adult, and adolescent cases, while also using other empirically-supported treatments. I have also worked to develop several relationship education curricula including one focused on individuals (rather than couples) called Within My Reach. Additionally, I provide supervision to doctoral students on their therapy cases at the University of Denver.

How we Integrate Relationship Education into Clinical Practice

HM: In my own work as a therapist, after an assessment of issues and history, the early sessions with a couple tend to be very consistent with an educational approach. I teach couples communication and conflict management skills and then teach three core principles from our PREP family of programs: 1) to help them decide rather than slide in their relationship during both major life transitions and on a daily basis, 2) to have them work individually and together to make their relationship safe for positive connections, and 3) to do their part to implement the tools they learn on a daily basis in their relationship. Our focus on “do your part,” coincides with the recommendations of many relationship experts (e.g., Hollis, 2003), suggesting that each partner needs to work on him or herself in order to increase their chances of having a healthy, happy relationship. This focus on self-improvement often helps couples improve in terms of managing negative affect, increasing positive connections, and understanding the meaning of some of the deeper themes in relationships such as forgiveness, sacrifice, and commitment (Fincham et al., 2007). As therapy progresses, I integrate the basic skills and principles of relationship education with more traditional cognitive-behavioral couples therapy, as well as other forms of couples interventions.

GR: I integrate relationship education into my therapy practice by teaching concepts, skills, and principles from relationship education curricula to therapy clients. In particular, I use concepts from PREP and Within My Reach most commonly because these are the two programs with which I am most familiar and because they have received the greatest amount of research support in the literature. Not surprisingly, many of the communication skills that couples learn in relationship education are also useful to couples in therapy. Indeed, as we discussed earlier, an overlap exists between behavioral couples therapy and relationship education in terms of the skills and principles couples learn.

Another way that I integrate relationship education includes referring clients to local relationship education workshops. As in many states, Colorado offers free relationship education workshops for couples and individuals through a grant from the Administration for Children and Families (www.marrywell.org). I find that both couples and individuals in therapy can often benefit from attending a relationship education workshop because they learn new skills and come back to therapy thinking about their relationships in new ways. As a clinician trained in relationship education, I can then support the new skills they have learned and help them to continue using them well.

HM: I also find it best in my practice to have couples who come to therapy to attend a weekend couples workshop (e.g., www.loveyourrelationship.com) as a prelude to therapy; it is simply more efficient to teach couples these skills in groups rather than in a typical therapy session.

Working with individuals

GR: In my practice. many individuals (both adults and older adolescents) who come for therapy present with relationship issues, but may also have specific problems in other areas (e.g., depression, addictions). Research suggests that relationship problems and many other mental health problems are comorbid, and in many cases relationship issues are part of the cause as well the treatment of the individual concern. Thus, discussing principles from individual-oriented relationship education curricula (like Within My Reach) with clients and teaching them communication skills for their romantic and family relationships has been very valuable not only for their relationship issues but also for their individual issues.

HM: We have found in our research that relationship education delivered to individuals can be successful in providing a positive impact on the attending partner as well as the non-attending partner (Markman et al., 2009). This finding counters the prevailing wisdom in the field that individual therapy for relationship issues is detrimental. This perspective in the field formed in part from an early study by Gurman and Kniskern (1977) which found that seeing one partner when there is a relationship problem may increase risks for relationship deterioration.

Benefits of Becoming Trained in Relationship Education

GR: As a clinician, I have found that knowing relationship education curricula has also enhanced my ease of transitioning from teaching skills to working directly on content in therapy sessions. Having the training and skills to lead workshops in relationship education has raised my awareness of appropriate times to teach skills formally or informally in therapy. Depending on the clients’ needs, I sometimes teach skills from PREP or Within My Reach in a highly structured format and assign homework directly from the curriculum. Other times, I teach the skills less directly and provide coaching for the clients on using skills while they are grappling with an issue in session. A background in teaching these skills to a wide range of audiences has helped me to decide which approach to use. More generally, knowledge of relationship education curricula extends a clinician’s toolbox, providing a wider repertoire of skills, activities, and resources with which to provide clients. Many of the activities that attendees complete, and most of the techniques or concepts they learn during relationship education workshops, can easily translate into therapy. Thus, training and experience in relationship education can increase flexibility as a therapist.

Training in relationship education can also lead to new avenues for providing services. In my own consulting work, I have recently extended relationship education to non-romantic relationships. For over a year, my colleagues and I have worked with the Colorado Department of Transportation on a project teaching relationship education to their maintenance workers. Over 1500 snow plowers and road crew workers are receiving relationship education that is aimed both at improving communication and relationships among co-workers, as well as, relationships with family members at home. This eight-hour course has been very well received by the administration, as well as by the employees themselves.

HM: I hope that service providers and researchers in the couples field will consider seeking training in one of the research-based relationship education programs. You will have the potential to 1) reach couples who really want our services, but have been uncomfortable seeking them, 2) teach tools that people can learn relatively easily, and 3) offer services that have the potential to achieve positive impacts not just on relationships, but on children, individual well-being, and one’s community.

While relationship education services do not generally bring in a lot of income to the service provider, social policy makers now recognize the benefits of increasing healthy relationships and marriages in the community. Thus, providers may be able receive reimbursement through existing or new grants or institute funding, even if they provide services for free. To the extent funds continue to be available for such community-based relationship education services, positions will be open for marriage and family therapy practitioners to be involved at many levels, from workshop leaders, to trainers of other leaders, to supervisors.

GR: In general, relationship education, when it is well developed and well delivered, often is a portal for individuals and couples to seek more intensive services for individual, couple, or family problems. Being in a safe and supportive educational setting that provides valuable skill training often leads attendees to consider (and often seek) opportunities for getting more help. In addition to participants receiving a general referral list (see “Getting More Help” document, please email the authors for a generic copy), relationship education service providers can give targeted referrals to individuals who otherwise may not receive them. Thus, these practices increase access to all forms of individual, couples, and family interventions.

HM: Finally, you can amplify your impact on couples by blending relationship education in your work. Let’s say a couples therapist is seeing 20 couples a week and couples stay in treatment an average of 4 months. Thus, the therapist is impacting about 60 couples a year. However, if the therapist was also to implement a relationship education event (e.g., a one day workshop) every month with 20 couples in the event, he or she would be impacting an additional 240 couples, some of whom may seek therapy services with that therapist after the relationship education event.