Table of Contents | Previous | Next |
Conclusions
The current review supports evidence from previous narrative reviews and meta-analyses that marriage and relationship programs provide benefits for the couples they serve. Average effect sizes of .68 for relationship satisfaction and .26 for relationship communication indicate that in these programs evaluated, the couples that received treatment made significant gains in satisfaction when compared to couples that did not receive the treatment.
Still, the results presented in this review should be interpreted with caution for several reasons. First, despite the rigorous inclusion criteria applied to this group of studies, there are numerous quality concerns— all of which were identified in the previous paragraphs. Second, this review was not able to examine thoroughly the types of raw data that comprised effect size statistics. Most measures used by these studies were well-documented indicators of satisfaction and communication, but it is still unclear exactly what each questionnaire asked and how different topics within the questionnaires were weighted versus others. Authors of the studies do not provide this information. Thus, reviewers were cautious to include only measures that are recognized in this field of research, but specific information about them is still unknown. Third, a sample size of 39 studies is very small. A small sample of studies inhibits the reviewers from thoroughly examining the correlation between different characteristics of studies, programs, and samples and the studies’ impacts. Fourth, most— 32— of these studies assessed both the treatment and control groups only at pre- and post-intervention measurement periods. They do not take follow-up measurements to assess impacts over time. Therefore, for the majority of the studies, reviewers are unable to ascertain how the couples’ outcomes would change over time. Thus, while this group of 39 studies represents the highest quality evidence available in the field, there are still many concerns that should caution the reader of this review.
Answering the policy question
As explained previously, this review is designed to provide answers to a general policy question posed by the U.S. Department of Health and Human Services. This is the question of how programs may help improve the relationships of couples (specifically low-income couples) and how their improvement may affect the well-being of children. This review informs policymakers on this topic in several ways. First, it presents the current state of high-quality evidence in the field of marriage program evaluation. Second, it reports an average impact of the evaluations as an average effect size and examines how this effect size changes for different types of relationship programs, clients, and studies. The review brings good news, as it indicates that evaluations of marriage programs show significant positive effects on average. Third, it provides a thorough examination of the quality of these studies, thus flagging general quality issues that face the field of program evaluation, and specifically marriage program evaluation.
There are several pieces of the greater policy question that this review is not able to answer, simply because the answers do not yet exist in this field of research. As there are no studies that include “low-income” couples, the review cannot determine the impact of marriage programs on low-income populations. Also, because very few of these studies follow an “intent-to-treat” model, the review cannot measure the impact of these interventions on couples that drop out of the program, nor can it predict how likely couples would be to want to participate in such a program in the first place. Another concern is that while this review does report on the effectiveness of programs evaluated by the field, it is not able to speak to the effectiveness of the current programs in operation at the moment. Because this review examines only the impacts of evaluated programs, one must consider whether or not evaluated programs would be different from non-evaluated programs. Lastly, these studies do not evaluate the impact of relationship programs on clients’ children. Some studies do mention whether or not the clients have children, but there are no measures taken on their well-being. Thus the review is not able to assess the programs’ effectiveness on children of clients. In the future, it will be necessary for researchers to examine programs that serve low-income clients and assess the well-being of clients’ children in order to inform policymakers on questions of greater interest.
| Table of Contents | Previous | Next |