Table of Contents | Previous | Next |
Program Intensity: Number and Length of Sessions
Total length of program (hours in each session times the number of sessions)
Five studies evaluated a program that was fewer than 10 hours in length, 16 were between 10 and 15 total hours, and 15 were longer than 15 hours. There were 3 studies that did not provide information on the length of the sessions in their programs.
Number of Sessions
All 39 studies reported on the number of sessions in their interventions. Nine studies evaluated programs with fewer than 5 sessions, 20 were between 5 and 10 sessions, and 10 had 12 or more sessions.
Average Session Length
Thirty-five studies reported the length of the interventions. Fourteen studies evaluated programs with hour-long sessions, 19 had sessions between 2 and 4 hours, and 3 had sessions lasting 8 hours.
Number of Weeks in Program
Thirty-five studies reported how long their programs lasted. Four studies evaluated “weekend” programs, 17 evaluated programs lasting fewer than 10 weeks, and 12 evaluated programs lasting between 10 and 15 weeks.
Program intensity: Number of Couples in Each Session Group
Studies also differed in the number of couples included in each session group. It is important to note that not all self-defined therapy and counseling programs were one-on-one treatment programs, as one might expect. For example, nineteen studies evaluated programs with one couple per group (i.e. one-on-one therapy with the provider), and not all of these were called therapy or counseling programs. Two were communications skills programs, and one was a pre-marital preparation program. Fourteen studies contained 3 to 5 couples per group, and 3 of these were therapy programs. Two studies contained 10 to 12 couples per group, one of which was a counseling program. The other four studies did not present this information.
Sample Characteristics
Most of these studies had relatively small samples. The average sample size of combined treatment and control groups was 34 couples. Most studies presented data on age, years of education, and length or relationship or marriage. For most evaluations, the reviewers were unable to determine the income level for the sample. Information on the sample was available for most of the studies, and reviewers present it here as an “average of the average,” the studies’ average of the couples’ average in each study.
The average age of the couples was 34 and the average number of years of education was 14.6. Reviewers find that the average relationship length for the couples in these studies was 9.3 years. However, this number should be considered an underestimate because some study authors provided "relationship length" as the length of a couple’s marriage, while other study authors provided the total length of the relationship. Couples that were married actually had longer relationship lengths that were not captured by these studies.
Distress Level
To determine the distress level of couples in these interventions, reviewers took the evaluation authors’ description of the sample as distressed or non-distressed. If this information was not available, the reviewers took the average MAT (Marital Adjustment Test, Locke & Wallace, 1959) or DAS (Dyadic Adjustment Scale, Spanier, 1976) score for the group and used cutoff points for distressed couples used in the research field {MAT— 100 (out of a score range of 2 to 158) and DAS— 107} to classify the sample. The reviewers chose these scores because the field considers these score cutoffs to define “distressed” couples. Crane, et al. recommended that 107 was a comparable DAS cutoff score to the MAT’s 100.
It is important to note that some of these programs may have contained both distressed and non-distressed couples. Reviewers classified all pre-marital education programs as serving non-distressed couples. They made this decision because experts note that pre-marital programs traditionally focus on non-distressed couples and help couples maintain an already high level of functioning while preventing further problems (Carroll & Doherty, 2003).
From the sample of 39 evaluations, twenty studies assessed distressed couples. Fourteen studies contained couples that were non-distressed. Reviewers could not interpret this information from the reports of the other 5 studies.
Follow-up Assessment
All of the 39 evaluations included in this review included assessments of the sample at both pre-test (before intervention) and post-test (directly after intervention). Many researchers continued to follow their treatment group(s) after the post-test assessment to see if the initial impacts were sustained over time. However, the majority of studies did not assess the control group at these times because most control groups in the studies were on a wait-list for treatment. Control group participants often received the intervention directly after the post-test evaluation period.
However, eight studies did assess both treatment and control groups at follow-up periods after the post-test evaluation. The longest follow-up assessment periods for these studies ranged from 3 weeks to 12 months after post-test, and the median follow-up period was 2.5 months. The average effect size for relationship satisfaction taken at follow-up is .29 (95% confidence interval: .01, .58) and for relationship communication, it is .11, (95% confidence interval: -0.42, 0.65). This suggests that the impacts of these programs may deteriorate over time, however these estimates are based upon very small sample sizes (5 studies measuring follow-up for satisfaction and 2 studies measuring follow-up for communication).
Results Presented by Studies
The measures used to evaluate clients in these studies were either self-report or provider-report questionnaires on the topic of either relationship satisfaction or communication. The studies report the findings from the questionnaires in the form of an average “score” from each questionnaire. Studies derived all outcomes for relationship satisfaction from self-report questionnaires, while studies used a mixture of self-report and provider-report questionnaires to determine communication outcomes. Most studies reported mean scores and standard deviations for both the treatment and control groups at pre-test and post-test. Reviewers converted this information to create a standardized gain score difference effect size. Reviewers explain the derivation of this effect size later in the report.
The most common outcome measures used in these evaluations are the Dyadic Adjustment Scale (Spanier, 1976) and the Marital Adjustment Test (Locke and Wallace, 1959). Both of these are self-report questionnaires, as noted previously. The most common measures of communication are two observer-report questionnaires— the Marital Interaction Coding System (MICS) and Communication Skills Test (CST)— and one self-report questionnaire, the Marital Communication Inventory (MCI). These measures are widely used in this field of research, and there is a body of research documenting their validity. Other studies in this review used similar measures that are either specific to the study or less common in the field but have similar properties.
| Table of Contents | Previous | Next |