Marital therapy effects: little or nothing with high relapseChapter 1, Myths and Mistakes of Marital Therapy
Excerpts from: The Marriage Clinic: A Scientifically Based Marital Therapy (Norton Professional Books) 1st Edition
by John M. Gottman Ph.D.
W. W. Norton & Company; 1 edition (August 17, 1999)
ISBN-13: 978-0393702828Pages 4-5More people seek therapy for marital problems than a any other type of problem (Veroff, Kulka, & Douvan, 1981). But how effective is the most sought after form of help? The largest effect sizes found in meta-analysis were those for behavioral marital therapy (for a review of those meta-analyses, see Bray and Gerald 1995). There are two ways to get a large effect in such a meta-analysis. One is by having a large experimental group effects. The other is by having a large deterioration in no-treatment control groups.
Indeed, when Jacobson (1984) reanalyzed data from four of his behavioral marital studies (which had among largest defects), he reported that, although 55% of the couples improved after treatment, only 35% were in the nondistressed range at the end of therapy. He concluded the significant effect sizes in controlled marital therapy studies may exist largely because people tend to deteriorate in the waiting list control groups. Jacobson Addis (1993) wrote, "success that investigators have had establishing these effects for their preferred treatments is not impressive as first thought. Improvement rate in the absence of treatment is so low that even small changes in the experimental treatment are likely to be statistically significant.”
Furthermore, a pervasive problem exists for almost all marital therapies that have been systematically evaluated using a long-term follow-up (as a minority of studies): a ubiquitous relapse effect. Of the couples that make some initial gains in therapy, a sizable percentage of these couples, about 30 to 50%, relapse in two years (Jacobson and Addis, 1993). Something like the second law of thermodynamics seems to function in marriage - that is, marital distress exist, things usually deteriorate (entropy increases). Therapeutic effects of the therapies that have been scientifically evaluated are generally weak, and there is a very high relapse rate.
What affects, in general, does marital therapy have? In our longitudinal study we have typically found a strong positive correlation (about .50) between having been in marital therapy and getting divorced. Is our result representative? The best study we have available on what therapy is "out there" that is, non-university based therapy, was done by Cookerly in 1980. Cookerly conducted a five year follow-up of 326 clients treated by a wide variety of marriage therapies in the US. The separation/divorce rate was 43.6% after five years. The highest separation/divorce rate occurred in the first year after therapy. These rates are considerably higher than base rates would lead one to predict. So marital therapy appears to be a reliable vehicle towards divorce!
To summarize, after taking a hard look at relapse rates, our current best estimate is that for about 35% of couples, marital therapy is effective in terms of clinically significant, immediate changes, but that after year about 30 to 50% of lucky couples who made the initial games relapse. This means that we all can claim is that in the best studies, conducted in universities with careful supervision, only between 11-18% of couples maintain clinically meaningful initial gains when treated with our best marital therapies. What must the actual success rate be in real clinical practice, with standards of training, treatment adherence, and supervision are generally lower, if not absent altogether? I think that we must conclude that it is likely that we have intervention method all to that nets relatively small effects, and we have a huge relapse problem.
Pages 8-9Perhaps the most influential theory of marital therapy is the "active listing model". This model forms the basis most marital treatments (see Gottman, Coan, Carrere, and Swanson, 1998; Jacobson and Margolin, 1979). Hypothesis is that stable, happy marriages are characterized by active, empathetic listening during conflict resolution, and that Italy marriages are characterized by the absence of this quality. That is, in order to have a happy, lasting marriage, partners need to be able to be nondefensive and empathetic listeners, even when they feel they are being attacked by their spouse. In most marital therapies this assumption is translated into some form of listener-speaker exercise…
There is a superbly conducted study by the Munich group, Hahlweg, Schindler, Ravensdorf, and Brengelmann (1984). They followed gurneys method precisely, comparing his active listening with the behavioral treatment that combined behavior exchange plus problem solving training. To their credit they also used observational methods, and they had follow-ups at six months and a year. They reported that in the short term:
- Active listening show decreases negative interactions with no increases in positive interactions
- Behavioral intervention showed both decreases negativity and increases in positivity.
In the long term, Hahlweg et al wrote, couples in the behavioral group reported significant and stable increases of quarreling behavior compared to couples in the active listening group, who return to pretreatment levels. On their communication self-report scale couples in a group remained stable, whereas couples in the active listening group relapsed to pretreatment levels. In assessing the clinics sick significance (not just statistical significant, of results, the typical couple in the behavior group scored within the “happy” ranges of marital quality whereas a typical couple in the active listening group was within the “unhappy” range.
Poll
