Australian Couples in Millenium Three 

In order to analyze the available evidence on the effects of marriage and relationship education in this chapter I first review the range of available approaches to relationship education. This is followed by an examination of the methodological challenges in evaluating the effects of relationship education. Then there is an analysis of the research evidence itself.

Range of approaches to relationship education

General approaches to relationship education

There is a huge diversity of approaches to marriage and relationship education. One dimension of this diversity is the degree of specificity and documentation provided for particular approaches. Some relationship education programs, such as the Prevention and Relationship Enhancement Program developed by Markman and colleagues (Markman, Stanley, & Blumberg, 1995), and the Premarital Personal and Relationship Evaluation Program (PREPARE) of Olsen and colleagues (Olsen, Fournier, & Druckman, 1996) have extensive documentation about their procedures, and there is a well developed training system for people who provide marriage and relationship education within these program structures. Most of these highly structured programs have been developed within research settings, and are based on psychological research on the influences and determinants of relationship satisfaction. At the other end of the continuum are a series of marriage and relationship education programs that have been developed by practitioners. These programs often have been developed locally, and the procedures that are followed often are not extensively documented, but rather depend upon the particular experiences of the educators concerned (Harris, Simons, Willis, & Barrie, 1992). Often the content of these programs reflects the feedback from participants in marriage and relationship education programs over a number of years, rather than being developed from concepts derived from psychological research.

A second dimension of marriage and relationship education programs is the mode of delivery. Many marriage and relationship education programs are delivered in groups, but others are delivered to individual couples, and some may be delivered to individual partners (Harris et al. 1992). Almost all programs as they are described in reviews of marriage and relationship education in Australia (Harris et al. 1992), focus on face-to-face interaction between marriage educators and participants. There are a small number of programs which involve telephonebased services, or self-directed materials. For example, a flexible relationship education program entitled the Relationship Education and EnhancEment program (REDEE) has been developed which involves self-directed learning, without faceto- face contact between participants and educators. The REDEE program consists of a video tape, guide book and telephone education service (Halford, Moore, Wilson, Burrows, & Farrugia, 1998).

A third dimension in the variety of relationship education programs is the intensity of those programs. Some programs are single sessions of one or two hours, whilst other programs involve 15 or even 20 hours of contact (To have and to hold, 1998). There are also variations in format. For example, some more intensive programs are run across a weekend in a single block, and other programs are run in multiple sessions over a number of weeks.

A fourth dimension in marriage and relationship education programs is the educational goal(s) of the program. Some programs focus primarily on the transmission of information, the clarification of expectations about relationships, and raising of awareness. Typical formats of such programs involve some didactic information presentation, plus discussion between the partners or within a group context. A somewhat different focus for relationship education places emphasis upon the acquisition of skills. For example, within the PREP program of Markman et al. (1994) they emphasize development of positive communication skills and a reduction of negative communication and negative conflict management as core educational goals for the program.

[ top ]

Information and awareness approaches

One useful way to conceptualize and organize the great diversity of marriage and relationship education programs is that they can be seen as falling into three broad categories. These categories, and representative examples of programs of each type, are presented in Table 2. The first general category could be described as information and awareness approaches. In this approach the emphasis is upon transmission of information, clarification of expectations, and increasing awareness of key relationship processes that influence relationship outcomes. Some programs include demonstration of relevant relationship skills such as communication, but there generally is not active training in these skills. From the available surveys of marriage and relationship education, it would seem that the majority of couples who receive marriage and relationship education in Australia receive this form of education (Harris et al. 1992; To Have and to Hold, 1998).

A limitation of the information and awareness programs is that many have been developed locally by practitioners and the exact content and process of the programs often is not well documented. The approaches often have grown from practical experience of delivering marriage and relationship education, and do not draw upon conceptual models or research available in the relevant literature. The lack of standardization of these programs means that they cannot readily be evaluated in scientific research.

[ top ]

Inventory approaches

A second category of marriage and relationship education programs is inventorybased programs. There are numerous inventories available. The most widely used in Australia are PREPARE (Olsen et al. 1996) and the Facilitating Open Couple Communication Understanding and Study (FOCCUS) (Larsen & Holman, 1994; Williams & Jurich, 1995). In these programs each partner completes an inventory which assesses a broad range of couple functioning dimensions, and the couple are provided with systematic feedback about the results of that assessment. The use of these inventories varies significantly between educators. Personal reports by educators with substantive experience with these programs indicates that many educators present the results of the feedback, and facilitate couples' discussion of their responses to that feedback. The goal seems to be to provide awareness and facilitate goal setting for positive change. Other marriage educators report that they use the inventories to diagnose couples' needs, negotiate with the couple to define particular learning goals that they want to achieve, and may supplement completion of the inventory with a variety of experiential exercises to achieve the negotiated goals. For example, a series of structured exercises have been developed that can be used with PREPARE (Olsen, Dyer & Dyer, 1997). However, the structured exercises still focus largely on promotion of awareness, and do not involve specific skills training.

The most extensively researched inventory is PREPARE, and this instrument has been through a number of revisions. The latest version is PREPARE 2000, which can be scored by computer. There also are variants of PREPARE such as ENRICH, which is designed for couples who have been living together in a committed relationship, and MATE which is for couples who have been together for extended periods of time. PREPARE was developed from a conceptual base of systems theory, and derives many of its components from psychological research. PREPARE involves comprehensive assessment across a wide range of relationship areas including couple communication, conflict resolution, parenting, religion, closeness, and flexibility. In two studies scores on PREPARE were shown to predict relationship satisfaction, and separation in couples across the first four to five years of marriage (Larson & Olsen, 1989; Fowers & Olsen, 1986).

FOCCUS is a 156-item instrument which assesses couples functioning across a wide range of areas including lifestyle expectations, friends and interests, personality, communication, religion and values, sexuality, readiness to marry, beliefs about the marriage covenant, and identification of key problem areas. FOCCUS was originally developed by Markey, Micheletto and Becker in 1985, and the latest revision is described in a publication by Markey and Micheletto (1997). The original FOCCUS program was developed for use by couples within the Catholic church, and the content has significantly more focus on spirituality and religious aspects of marriage than many other programs. Parallel versions have been developed for couples who are entering inter-faith marriages, second marriages, or couples who are cohabiting. In one study FOCCUS scores were shown to predict couples' relationship satisfaction across the early years of marriage (Williams & Jurich, 1995).

The inventory programs like PREPARE and FOCCUS have the advantage of being clearly structured, and hence amenable to scientific evaluation. Furthermore their development from psychological research and theory means that they seem likely to have positive benefits for couples, as they do target factors shown to predict relationship outcomes. However, a presumption underlying these approaches is that awareness will promote better couple coping. As Silliman, Stanley, Coffin, Markman, & Jordan (in press) point out, identification of partner differences or relationship weaknesses may be counterproductive unless couples are helped to deal effectively with the differences or issues identified. For example, feedback on divergent expectations may lead to profitable discussion, but couples lacking conflict management skills may be unable to resolve these differences.

[ top ]

Skills training programs.

The third broad category of relationship education programs is skills training programs. There are a number of examples of such programs including Guerney and colleagues' (Guerney, 1977, 1987, Guerney & Maxson, 1990) Relationship Enhancement program, Markman and colleagues' (Markman et al. 1994; Stanley, Blumberg, & Markman, in press) PREP program, and the Couples Communication program (Miller, Wackman, & Nunally, 1976). In each of these programs couples receive instruction on use of key relationship skills, which can be a mixture of lectures, demonstrations, and audio-visual presentations. Couples also receive opportunities to practice these new skills, and receive feedback from educators about their use of skills. Moreover, most programs involve some structured assignments couples are asked to undertake between sessions to practice applying skills within their own relationship.

The content of the various skills training programs has a number of common elements. For example, positive communication, conflict management, and positive expressions of affection are included in Relationship Education, PREP and Couple Communication (Guerney, 1977; Markman et al. 1994; Miller et al. 1976). There also are significant variations. For example, PREP places the most emphasis in content on the prevention of negative conflict, as this is argued to be central to the prevention of relationship problems (Markman et al. 1994). In Relationship Education the development of partner empathy for each other receives very strong emphasis (Guerney, 1977), whilst this has less emphasis in PREP.

The differences in skills focused on in skills training programs reflects, at least in part, a lack of consensus about which relationship skills are central to relationship satisfaction and stability. For example, recently Gottman et al. (1998) argued that active listening skills, which often are taught as part of communication skills within relationship education programs, really are not predictive of relationship satisfaction in the early years of marriage. Gottman argued that husbands' responsiveness to their wives attempts to influence the relationship was a better predictor of relationship satisfaction, and this responsiveness should be a focus of skills training in relationship education. However, Stanley, Bradbury, and Markman (in press) argued the presented data did not support Gottman et al.'s view, and that current communication skills training practices should not be abandoned. Examining the measures used by Gottman et al. (1998) it is not at all clear that they really measured influence responsiveness. Moreover, the previous longitudinal research showing communication predicts relationship satisfaction was ignored by Gottman in reaching his conclusions. Whilst I feel Gottman et al. (1998) were not justified in reaching the conclusions they drew, the specific communication skills that enhance relationship satisfaction need further research.

A variant on the skills training programs is the SELF-PREP program (Halford & Behrens, 1996; Halford et al. 1999). Like the original PREP program, this program is an active skills training program that targets a wide range of relationship skills such as communication, conflict management, intimacy enhancement, and sexuality. In contrast to the original PREP, rather than just focusing on the teaching of specific skills, a specific objective of the SELF-PREP program is to teach selfdirected learning skills, or meta competencies. In this approach couples are taught to evaluate their own skills within particular relationship domains, to set goals for self-directed change, are assisted to develop self-change plans, and their skills in evaluating the effectiveness of self-change efforts are also developed. An advantage of this approach is that it does not presuppose specific skills will be universally helpful to couple relationships. Rather, in SELF-PREP couples are taught to evaluate the effects of particular behaviors within their relationship and to self-direct change to enhance their relationship.

The assessment of relationship meta competencies has been a subject of recent research. Wilson, Halford, Lizzio, and Kimlin (1999) have developed a self-report measure of relationship meta competencies. Respondents rate the extent to which they agree with a series of items such as "I can identify things that I do which strengthen our relationship", "I am clear about the type of relationship I want to have with my partner", and "I have lots of good intentions about improving my behaviour in our relationship, but I don't seem to follow these changes through". Whilst research on this construct is limited there is evidence that it can be reliably measured, and that it might be an important construct that influences long term relationship satisfaction.

Given that there is not broad agreement about exactly which skills may be adaptive for particular couples, and that the adaptive skills may vary by culture and the particular challenges that couples are confronting, the idea of self-regulation of relationship meta competencies is something that warrants further investigation. Already one program which utilised this construct has been the subject of a controlled trial, and positive long-term effects of a marriage education program using SELF-PREP were reported at four year follow-up (Halford et al. 1999).

[ top ]

Additional Approaches to Relationship Education

In addition to the widely used approaches to relationship and marriage education described above, there are a number of new directions and possible innovations for providing relationship education. There is a huge diversity of self-help books and other materials available to couples providing advice on relationship education. Sayers, Colombo, Pirolli, Brigidi & Kohn (1996) identified 98 books currently available for the lay public on marriage and couple relationships. Little to nothing is known about the extent to which couples apply ideas gained from such reading, or the effectiveness of such materials. Christensen and Jacobson (1994) reviewed the use of self-help books in self-directed learning for a variety of personal growth areas. They concluded that while there was limited research on the effectiveness of self-help books or other self-directed learning resources, self-help books that taught skills that were not too complex could be quite effective. They did not make specific conclusions about the effectiveness of advice on marriage and relationships in printed or audio-visual form, as there was insufficient evidence to reach conclusions on this. However, they did conclude at the end of their extensive and scholarly review that the possible benefits of selfdirected change through provision of brief advice, written and audiovisual materials had not been taken sufficiently seriously by professionals. They argued that the available evidence suggested that often programs were effective, and that there was a preference by many people in the community to access information in these alternative ways.

There are a number of developed audio-visual programs that can be offered in self-directed learning mode. For example, PREP Educational Videos Incorporated have produced a series of videotapes entitled "Fighting for Your Marriage" used quite widely in the United States. Whilst there is no research currently evaluating the effectiveness of watching these videotapes, the content of the tapes is based on extensive psychological research and it is possible that some couples do benefit considerably from exposure to such materials.

Several web sites exist which provide couples with access to a variety of educational materials about marriage and relationships. For example the Relate web site set up by the Department of Family and Community Services has a series of tips on it about how to enhance the quality of relationships. It is unclear whether simple guidance in this form, or more elaborate information that might be made available in a similar form, is helpful in the relationship education area. However, there is extensive research that has been done on providing a variety of health education materials via web sites. In a recent review conducted for the United States Department of Health and Human Services, Eng and Gustafson (1999) concluded that many health education programs can be delivered effectively via the web. They reported on numerous web sites that had been subjected to rigorous evaluation, and where the access of potential consumers of information and ideas was much greater than could be expected by offering face to face programs. Moreover, they found that people who had accessed such web sites often reported them as being extremely helpful, and as allowing people to then select what sort of further information they might find most useful or helpful.

Elliot (1999) reviewed web based education programs on family life. Relatively few of these web sites focused on relationship education for couples. Moreover, Elliot noted that few sites provided the opportunity for interaction, thus limiting the chance of skills acquisition. The opportunities for use of new technology to allow people to access information on couple relationships has not been explored sufficiently.

A potential advantage of self-directed learning materials is that they may engage people who would otherwise be resistant to marriage and relationship education. Such engagement might produce significant positive relationship benefits. Even if there are no immediate benefits from using such materials alone, they might encourage people to present for more traditional face-to-face forms of marriage and relationship education.

[ top ]

Methodological issues in evaluation of relationship education

Outcome measures

Satisfaction with the content and process of relationship education should be, and often is assessed routinely in relationship education course delivery. However, ratings of satisfaction with course content are influenced by a wide range of factors such as the price paid for the program, the marketing of the program, and the quality and comfort of the venue (Morris, Cooper & Gross, 1999). Reported satisfaction with courses can guide relationship educators as to aspects of the program that were not well received by participants, and this is helpful. However, reported satisfaction does not necessarily reflect the value of course content, and it probably does not reflect the long-term benefit participants may gain from course participation. In other words, reported course satisfaction is important as part of evaluation of relationship education. But rated satisfaction alone is inadequate to evaluate programs.

As distinct from satisfaction with the relationship education course, it is important to assess partners' satisfaction with their relationship. Relationship satisfaction has the highest face validity of possible measures of relationship education outcomes. However, there are methodological problems with measuring the construct of relationship satisfaction. Reports of relationship satisfaction seem to reflect a broad general sentiment about the relationship, regardless of the specific questions posed (Heyman, Sayers, & Bellack, 1994). For example, whether partners are asked about satisfaction with resolving conflict, sharing time together, or sense of intimacy, partners tend to give either uniformly positive or negative responses across items. This has lead most researchers to operationalize relationship satisfaction as a single score. While the collapsing of seemingly different dimensions of relationship satisfaction into a single satisfaction score may seem conceptually unsound, evaluation of reports of relationship satisfaction consistently show partners often make uni-dimensional, global evaluations of their relationships (Eddy, Heyman & Weiss, 1991; Heyman et al. 1994).

Partners who are satisfied with their relationships have a positive bias in their reports about their relationships (Fowers et al. 1996) and tend to report that just about anything that could be positive about their relationship, is positive. Sometimes this bias sees very unrealistic. For example, some satisfied couples endorse descriptions of their relationship such as "My partner meets my every need" or "I have never ever thought about the possibility of not being married to my partner" (Fowers et al. 1996). Moreover, most satisfied married couples estimate the chance of them ever divorcing their partner as extremely low, despite the well-documented high rates of relationship break up. Fowers labels the phenomenon of making seemingly unrealistically positive reports on a relationship as idealistic distortion. Idealistic distortion is correlated with relationship satisfaction and it has been suggested that such a distortion may help maintain relationship satisfaction (Fowers et al. 1996).

Idealistic distortion in reports of relationship satisfaction may make such reports insensitive to positive effects of marriage or relationship education. Two approaches to measurement of relationship satisfaction may overcome these problems. First, Fowers et al. (1996) utilize a measure of idealistic distortion and correct self-reported relationship satisfaction for the effects of idealistic distortion. In other words, very high relationship satisfaction scores that result from idealistic distortion can be corrected to a more realistic satisfaction score. It is possible that this method may allow detection of real increases on relationship satisfaction.

Fincham et al. (1997) developed a two dimensional scale of relationship satisfaction called the Positive and Negative Marital Qualities Scales (PNMQS). This scale separates partner rating of satisfaction with positive aspects of the relationship from ratings of dissatisfaction with negative aspects of the relationship. Fincham et al. showed that the ratings of positive and negative aspects of the relationship are relatively independent, and this structure may overcome the idealistic distortion effect.

The recent developments in measurement of relationship satisfaction may overcome the limitations of previous scales, and could allow better assessment of the immediate effects of marriage and relationship education on relationship satisfaction. However, it still is likely that many couples presenting for marriage and relationship education will have high levels of initial relationship satisfaction, and this may impose ceiling effects on changes in relationship satisfaction. However, given that relationship satisfaction often erodes over time (Karney & Bradbury, 1995), if relationship education prevents that erosion, then the effects of relationship education on relationship satisfaction may be evident some years after relationship education.

Couple interaction is an important index of relationship education outcome. As noted in the chapter on the nature of strong couple relationships, particular characteristics of couple interaction can define strong relationships. Couple interaction often is assessed by observation in research, but this is very time consuming and unrealistic for routine program evaluation. Self-report measures of communication, conflict management, affection, and other important aspects of couple interaction can be administered at the beginning and end of programs to establish if couples have achieved gains in this crucial area. Moreover, such evaluation can be an important part of the educational experience. Pre-program administration of assessment measures can be used to assess couples' educational needs, as is routinely done in the use of standardized inventories. Post-program evaluations can be used to identify areas that have been strengthened across the program, and to help couples identify future learning needs. Greater routine evaluation of programs in this form should be encouraged.

Research design challenges

There are significant methodological problems associated with evaluating the effects of relationship education programs. First, it is important to know how couples would have gone without the intervention of relationship education. This can only be done in the context of having some sort of comparison or control group. Furthermore, it is necessary to ensure that the couples who are receiving the relationship education program are comparable to the couples who are in the comparison condition. The only real way to be certain that there is comparability is to randomly assign couples to receive the relationship education program or be in the comparison condition. However, if couples are randomly assigned to receive no relationship education when they desire relationship education, they may seek education outside the study. This obviously violates the research integrity of the study.

Different researchers have come up with different responses to this challenge. Markman, Floyd, Stanley & Storaasli, 1988 and Markman, Renick, Floyd, Stanley & Clements, 1993 initially recruited engaged couples and then randomly assigned them to be either offered relationship education or not. Unfortunately only a third of the couples who were offered relationship education agreed to participate. These agreeing couples were then compared with couples who had not been offered relationship education. A difficulty with this approach is that the selfselection into relationship education confounds the actual delivery of relationship education. In other words, any observed effects may be due to couples who are keen to do relationship education being the ones who would have done better anyway. Hahlweg and colleagues (1998) invited couples who were being married in the Catholic church in Germany to participate in either a skills based relationship education program, or a standard program provided by their local priest. Couples could choose which of the programs they preferred. Again, the difficulty is that couples who were choosing a skills based program might, by this very act of self-selection, be couples who were going to do better than the other couples. Halford and colleagues (1999) randomly assigned couples to receive either their SELF-PREP program or a minimal information and awareness program. The advantage of this strategy was that it ensured that the couples were comparable prior to receiving relationship education, and all couples who expressed interest in relationship education received a program. Moreover, the results of their research showed that couples were satisfied with both programs and rated them highly. Thus couples were receiving something which they had positive expectations about, and it was possible to establish if the relationship education program had specific beneficial effects.

A second serious flaw with most existing research concerns the lack of adequate follow-up. Most studies have evaluated only immediate or short-term effects of relationship education programs. A survey of 85 couple and family prevention and enrichment studies found that only 40% included follow-up measures at some point after post-test (mean = 12 weeks), with the longest follow-up assessment occurring at 12 months (Giblin, Sprenkle, & Sheehan, 1985). There is a need to carry out long-term controlled trials and to randomly assign participants to relationship education or alternative conditions in order to establish if marriage and relationship education does have a beneficial effect. Retention of couples within the study for long-term follow-up can be problematic. In much clinical and relationship education research there are substantial attrition rates from couples sampled, which seriously compromises the conclusions that can be drawn about long-term effects. Strategies such as keeping regular contact with couples, ensuring that couples provide a series of contact names and addresses so that they can be recontacted if they move, and paying couples for follow-up assessments in longterm research, are all important possibilities to enhance retention in studies.

[ top ]

Empirical evidence on the effects of relationship education

General findings

There are a very large number of research studies evaluating marriage education and enrichment, and there have been numerous reviews and meta-analyses of that evidence (Bagarozzi & Rauen 1981; Bradbury & Fincham, 1990; Christensen & Heavey, 1999; Dyer & Halford, 1998; Giblin et al. 1985; Guerney & Maxson 1990; Hahlweg & Markman, 1988; Sayers, Kohn & Heavey, 1998; Van Widenfelt, Markman, Guerney, Behrens, & Hosman, 1997). The conclusions drawn by reviewers of research on the effects of marriage and relationship education diverge quite markedly, even when examining the same evidence. For example, Guerney & Maxson (1990) commenting on the meta-analysis of outcome studies undertaken by Giblen et al. (1985) concluded "there is no doubt that, on the whole, enrichment programs work and the field is an entirely legitimate one" (p. 1133). In contrast, Bradbury & Fincham, (1990) concluded from the results of the same meta-analyses: "prevention programs have not yet been shown to produce lasting changes in relationships" (p.397).

Given the diversity of conclusions drawn by previous reviewers of the evidence, it is important to analyze very carefully the available evidence, in order to establish exactly what effects have been demonstrated with which programs for which couples. I begin by looking at generic marriage and relationship education programs that target all couples, and then focus on programs for couples during the transition to parenthood, after divorce, and entering step families.

First, there is a general finding that most couples who complete competently run pre-marriage education programs generally report high satisfaction with the programs (Harris et al. 1992). This high satisfaction is evident across information and awareness programs, inventories and skills training programs (Halford & Behrens, 1996). Reported satisfaction generally is lower amongst couples who are younger, who have fewer years of formal education, who have to travel further to attend sessions, and who perceive the programs as intrusive into their privacy (Russell & Lyster, 1992). Couples who complete relationship education within a month or two of their marriage also tend to report less satisfaction than couples who had more time between doing the program and marriage (Russell & Lyster, 1992). Whilst high consumer satisfaction is desirable, this does not address the effects of education on relationship outcomes.

A 1986 meta-analysis of 85 relationship education and enhancement programs found an average effect size of .44 across all education programs and relationship outcome measures (Giblin et al. 1985). In meta-analyses of outcome studies an effect size of 0.3 is usually seen as small, 0.5 as moderate and 0.8 as large (Cohen, 1997). Thus an effect of 0.44 is close to moderate effect size, which is potentially an important effect if it occurs across all programs. However, the length and content of programs included in the meta-analyses was highly variable, most studies lacked any sort of control group, and only a very small number of the studies included any follow-up results. Moreover, the Giblin et al. meta-analysis included all locatable studies, whether or not they were published, and no studies were excluded on methodological grounds. Taken together, these results suggest that most studies included in this meta-analysis failed to meet usual scientific standards of evaluation.

Hahlweg & Markman (1988) also conducted a meta-analysis, and they included only seven studies, all of which were published studies that were controlled trials. Moreover, they focused their review on programs that included relationship skills training. They found a mean effect size of 0.79 for education programs relative to controls. This large effect size was found in the highest quality studies.

Both the Giblin et al. (1985) and Hahlweg and Markman (1988) meta-analyses found differences in effect sizes as a function of type of measure used to assess change. Observational measures of relationship skills showed substantially larger differences between groups than self-report measures of relationship satisfaction [0.76 versus 0.35 in Giblin et al. (1985) and 1.51 versus 0.52 in Hahlweg & Markman (1988)]. Furthermore, Giblin et al. (1985) found greater effect sizes for self-report measures classified as assessing relationship skills (0.63) than for measures classified as assessing relationship satisfaction (0.34). Thus, since the mid- to late 1980s it has been well established that relationship education produces large improvements in relationship skills in the short term, and that there are small short-term increases in relationship satisfaction.

Giblin et al. (1985) found a larger effect size on relationship satisfaction for studies which included more distressed couples (0.51) than for studies with fewer distressed couples (0.27). The short-term effects of relationship education on relationship satisfaction are likely to be small when participating couples are mostly satisfied at the beginning of the program. Given that marriage and relationship education is supposed to prevent relationship problems developing, it is important to focus on studies that evaluate the effects of education on relationship satisfaction and stability over periods of years, focusing particularly on the effects on couples who initially are satisfied in their relationships.

There are only seven controlled trials evaluating marriage or relationship education programs for currently satisfied couples entering committed relationships that include follow-up assessments of six months or more. These studies are summarized in Table 3. All these programs have targeted engaged, dating, or recently married couples, and have consisted of between four to eight face-to-face group sessions of two to three hours duration. All of these evaluations were of skills based programs, with almost all of them focusing upon PREP or a variant of PREP.

Across studies there is a consistent finding that, relative to no intervention or minimal intervention controls, couples acquire the targeted skills (Avery, Ridley, Leslie, & Milholland, 1980; Renick, Blumberg, & Markman, 1992; Markman & Hahlweg, 1993; Markman et al. 1988; Miller, Nunnally, & Wackman, 1975; Wampler & Sprenkle, 1980). Long term follow-up of the maintenance of acquired skills is less well investigated, but three recent findings show maintenance of acquired skills over a period of some years (Hahlweg, Thurmair, Eckert, & Engel, 1996; Halford, Sanders, & Behrens, 1996, 1999; Markman et al. 1993). However, attenuation of training effects was reported to occur over a five to ten year period in the only study to have follow-up over that period of time (Stanley, Markman, St. Peters, & Leber, 1995)

[ top ]

1. Not from original source. Most detailed published report.
   PREP = Premarital Relationship Enhancement Program

The long-term effects of relationship education on relationship satisfaction and risk of divorce are not well documented. Short-term increases in satisfaction from relationship education are modest in some studies (Renick et al. 1992; Ridley, Jorgensen, Morgan, & Avery, 1982) and not evident at all in many studies (e.g., Markman et al. 1988; Markman & Hahlweg, 1993; Van Widenfelt, Hosman, Schaap, & Van der Staak, 1996). This lack of effect on relationship satisfaction in the short term may be a function of the fact that relationship satisfaction is already high in the targeted populations. In addition, there may be a ceiling effect operating in the measures currently used to measure relationship satisfaction.

The most meaningful index of the efficacy of relationship education is its longterm effects; unfortunately only four studies have follow-ups of more than 12 months. Markman and colleagues have found in two studies that skills based relationship education was associated with enhanced relationship satisfaction or functioning 2 and 5 years after marriage (Hahlweg et al. 1996; Markman et al. 1993). The Markman et al. study also found that across the 3, 4, and 5-year followups, the intervention couples reported significantly fewer instances of spousal physical violence than control couples. A third study using an almost identical education program did not replicate these results (van Widenfelt et al. 1996). The van Widenfelt et al. study differed from the Markman studies in that high risk couples were targeted. This finding may indicate limitations in developing relationship education programs solely through research at the level of universal populations.

The fourth study was a randomized controlled trial of a skills based relationship education program, with collection of relationship satisfaction and stability data at four year follow-up (Halford et al. 1999). A unique aspect of this study was that couples were stratified into high and low risk for relationship problems on the basis of negative family-of-origin experiences (parental divorce or inter-parental violence). Couples completing PREP were found to have significantly higher relationship satisfaction at 4-year follow-up than couples in a control condition, but this effect was only evident for couples at high risk of relationship problems. The possibility that marriage and relationship education may have differential effects for different couples needs replication, but does suggest that some couples may benefit more from relationship education that other couples.

In summary, the effects of skills based marriage and relationship education programs on relationships skill is well established: the programs produce increases in skills that are sustained for at least the first few years of a committed relationship. There is some evidence that PREP and its variants prevent the erosion of relationship satisfaction over time, but these effects may be limited to couples at high risk for relationship distress. Replication of the long-term effects of PREP in a randomized controlled trial is highly desirable.

[ top ]

Relationship education Across the Transition to Parenthood

There are a number of well-documented programs assisting with the transition to parenthood. Antenatal childbirth classes are extremely common, but tend to focus predominantly on childbirth itself, with some lesser coverage of early parenting experiences. There are some programs that extend into the period after childbirth, though most often these programs are targeted at mothers seen as being at high risk for problems. For example, there are several programs for single mothers, mothers with social disadvantage, premature birth children, or for mothers suffering from post-natal depression (e.g., Barnard, Booth, Mitchell, & Telzrow, 1988; Barnard, Morisset, & Spieker, 1993; Egeland & Erickson, 1993; Erickson, Korfmacher, & Egeland, 1994; Lieberman & Pawl, 1993; Heinicke, 1995; Olds & Kitzman, 1993; Olds, Kitzman, & Cole, 1986; Meisels, Dichtelmillar, & Liaw, 1993; Pope et al. 1997). Almost all of these programs focus on the parent-child relationship, and on changes that often involve mothers but not fathers. A number of these programs have been shown to reduce stress in mothers and problems in children (Cowan & Cowan, 1995). However, couple relationship functioning has not been a primary focus of most programs, and the impact on supporting couples relationships has not been assessed.

Table 4 is a summary of reports on couple relationship education programs for the transition to parenthood that have been developed and evaluated. The Becoming a Family Project (Cowan & Cowan, 1992) is the only randomized clinical trial of a couple-focused transition to parenthood program. The intervention was quite intensive, consisting of 24 weekly small group meetings. The content of the program included leader input and group discussion on the transition to parenthood, effects on the couple relationship, and how particular couples were coping. The results were striking: declines in marital satisfaction were less severe in intervention than control couples, and 18 months postpartum none of the 24 couples from the intervention group had divorced while 10 couples from the control groups had separated or divorced.

Two additional pilot studies of structured couple-based primary prevention transition to parenthood interventions have recently been completed. Heavey (1995; cited in Silliman et al. in press) and Jordan (1995; cited in Silliman et al. in press). In both these studies modified versions of the Premarital Relationship Enhancement Program (PREP, Markman et al. 1994) were used to provide relationship skills training to married couples expecting their first child. Recruitment, attendance, retention, and evaluation data support the fact that couples expecting the birth of a first child are very open to such programs, found them useful, and evaluated them positively. Given these results and the Cowan and Cowan (1995) findings further research on couples-based interventions to assist couples with the transition to parenthood are needed.

[ top ]

Adjustment to separation and re-partnering

Despite the number of relationships that break down and the severe consequences that can result, there are few controlled outcome studies of relationship education for adults adjusting to relationship breakdown. In a recent meta-analysis only 7 studies were found that included both a treatment and a control group (Lee, Picard & Blain, 1994). Of these seven studies only three had a written program manual, and only three reported any follow-up assessment. The reported length of programs varied widely from Woody, Colley, Schlegelmilch & Maginin (1985) who conducted a week-end workshop of about 10 hours of faceto- face contact to Lee and Hett (1990) and Salts and Zongker (1983) who provided approximately 24 hours. All the programs reported including content such as individual adjustment, communication skills, financial planning, parenting issues, social support, and dating. All the programs can best be described as information and awareness programs, rather than skills training programs.

Three studies found that treatment helped to reduce depressive symptoms (Bloom, Hodges & Caldwell, 1982; Lee & Hett, 1990 and Malouff, Lanyon & Schutte, 1988) while only one study found improvements in anxiety (Lee & Hett, 1990). In terms of overall distress, Bloom et al. (1982) found significant improvements for the treatment group. In terms of divorce-specific variables only one study (Thiessen, Avery & Harvey, 1980) reported improvements after the program. No evidence was found that treatment reduced single parenting problems, influenced attitudes towards the ex-spouse, or improved communication or type of contact with the ex-spouse.

Lee et al. (1994) concluded from their meta-analysis that existing relationship education programs for divorced adults have only moderate beneficial effects. The programs do help participants reduce depression and distress. Unfortunately it has not yet been demonstrated whether divorce interventions are useful with regard to divorce-related issues including single-parenting, relationship with the former partner, social support and everyday practical issues such as finances, employment and homemaking. No examination has been made of the effects of such programs in reducing relationship problems in subsequent relationships. Future research would benefit from greater attention to teaching skills that will help people adjust to separation, and to prepare them better for re-partnering successfully.

[ top ]

Stepfamilies

There have been a large number of relationship and family education programs described for stepfamilies. Of 23 programs described by Hughes & Schroeder (1997) 17 were for adult partners only, 2 for children only, and 4 for whole stepfamilies. Many of these programs address issues particularly relevant for stepfamilies. For example, many describe content focused on good relationships between stepparent-stepchild, biological parent-child, the couple, siblingstepsibling, non-residential parent - parent/stepparent, development of family roles, rules, boundaries and hierarchies, resolving grief over various kinds of losses by stepfamily members, and education about differences between first marriage families and stepfamilies. More than half of the stepfamily programs included at least some elements of what would typically be included in couple relationship education programs, such as communication skills, problem solving and conflict resolution. However, almost no attention was paid in any of the programs to the effects of the non-residential parents on stepfamilies, sibling/stepsibling rivalry, or legal and financial issues (Hughes & Schroeder, 1997). This reflects a significant gap between the research on influences on stepfamily functioning and the content of education programs for stepfamilies.

There have been three controlled trials of the effectiveness of relationship education programs for stepfamilies (Brady & Ambler, 1982; Nelson & Levant, 1991; Nicholsen, Halford, & Sanders, 1996). Brady and Ambler (1982) randomly assigned 33 couples to either an education group program or a wait list control. In the stepfamily relationship education program the primary focus was on parent/stepparent - child relationships, discipline, communication skills, problem solving and increasing knowledge about stepfamily life. Participants completed the Family Environment Scale and Belief About Step-parenting Role questionnaire (designed by the researchers), at pre- and post-intervention. There was no significant difference between the conditions on parent's perception of the family environment on 9 of the 10 Family Environment Scales, and no change was found for either stepparent or parent on the Belief About Step-parenting Role measure.

Nelson and Levant (1991) randomly assigned parents in stepfamilies to either a family education program or a wait-list control condition. The topics covered in the program were similar to Brady and Ambler's (1982) program: parent/stepparent - child relationships, discipline, communication skills and problem solving. Unlike Brady and Ambler's (1982) program their course was run with either the parent or the stepparent alone. Pre- and post-intervention participants completed a battery of self-report measures on communication, family environment, and couple relationship satisfaction. The program produced modest improvements in parents self-reported communication skill, but no change was found for parents' perceptions of stepfamily adaptability, cohesion or relationship satisfaction.

A third program developed by Nicholsen et al. (1996) randomly assigned 43 couples to either a group stepfamily intervention or a minimal intervention, selfhelp control group. Couples in the self-help condition worked through Ruth Webber's "Living in a Stepfamily" book covering issues such as understanding couples, relationships with children, developing a positive stepfamily, ghosts from the past, and the role of the stepparent. In the group program couples covered the topics of stepfamily relationships, enhancing relationships with children, discipline and partner support, conflict management, communication skills, problem solving and family activities. Active relationship skills training was incorporated into this program. All participants were assessed at pre- and postintervention, and at 12-month follow-up on behavioural measures of communication between a child and the partners, and between the partners. Participants also completed a battery of self-report measures on child and couple adjustment, parenting practices and confidence, and couple relationship satisfaction. Post intervention across both conditions the number of issues causing conflict between couples had significantly decreased, children reported lower levels of depression, and parents reported fewer negative life events. Ongoing analyses will provide evidence of the longer term effects of the program, but the initial results look promising.

Other programs

There are a number of descriptions of programs that target couples in other transitions. For example, there have been a number of descriptions in the literature of programs to help people cope with major life crises such as unemployment, retirement, and other major transitions. However, there are no reported controlled trials on any of these programs. There is a major need for further research to look at ways in which couples can be supported and helped to sustain mutually satisfying relationships across different life challenges.

One exception is the recent work by Halford, Scott, et al. (in press) who targeted couples in which the woman had been recently diagnosed with breast or gynaecological cancer. These couples were recruited at the time of diagnosis and were provided with a couples-based coping program. The coping program included training in communication and mutual support skills, as well as education about the nature of cancer and its treatment. The program was developed collaboratively with medical specialists, and a key focus was to assist the couple to support each other through the difficult processes of diagnosis and treatment. In a controlled trial outcome it was found that this program was significantly better than standard care in that women who received the program were substantially less depressed, and reported higher relationship satisfaction, than couples who were in the control group. Importantly, the women's husbands also reported significantly less psychological adjustment difficulties, and better coping as a result of taking part in the couple coping program.

[ top ]

Summary and Conclusions

The vast majority of available research on marriage and relationship education has been completed on pre-marriage programs. There is a well replicated finding that most couples who complete competently run pre-marriage relationship education programs report high satisfaction with the program.

The effects of information and awareness approaches to relationship education have not been adequately evaluated. At this time there is no scientific evidence that this general approach enhances couple relationships. Moreover, given that many programs offered have been developed at a local level and the content is not clearly documented, it is hard to imagine conducting systematic evaluation of such programs. Moreover, many information and awareness programs have been developed with limited attention to existing knowledge on the nature and influences on strong couple relationships.

Inventory based programs such as PREPARE and FOCCUS also lack adequate evaluation of their effects. These programs are well documented, do have a coherent theoretical base, and do derive much of their content from established knowledge about relationship influences. However, there is a need to evaluate the long-term effects of such programs. In particular the presumption that identifying and discussing areas of relationship strength and weakness in itself allows couples to enhance their relationship needs to be evaluated.

There is evidence that skills based training programs produce significant increases in both observed and self-reported couple communication and conflict management skills, and that these increases are of moderate to high effect sizes. Positive effects of skills training programs have been observed with several skills based programs. In the short term there are modest increases in relationship satisfaction, but these tend only to occur in couples with mild relationship problems. The long-term effects of skills based relationship education on relationship satisfaction and stability are not as well established. There is some evidence that the PREP skills training program enhances relationship satisfaction maintenance across the first 4 to 5 years of marriage, and may reduce rates of relationship break up and relationship aggression. However, the methodological problems with existing studies limit the confidence that can be placed in these conclusions. The benefits of programs may be restricted to couples at high risk for relationship problems.

Relationship education at time points other than at pre-marriage has received little research attention. Available programs for the transition to parenthood, formation of step- families and coping with severe illness all show promise and warrant further development and evaluation. Programs on coping with separation do have some benefits in reducing severe distress, but their effects on relationship satisfaction after re-partnering has not been investigated. The most promising programs all seem to utilize skills training, and this approach may be the best general approach to pursue.