Meta-analysis of the alliance–outcome relation in couple and family therapy
Effectiveness bank home page. Opens new windowReview analysis

This entry is our analysis of a review or synthesis of research findings added to the Effectiveness Bank. The original review was not published by Findings; click Title to order a copy. Free reprints may be available from the authors – click prepared e-mail. The summary conveys the findings and views expressed in the review. Below is a commentary from Drug and Alcohol Findings.

Links to other documents. Hover over for notes. Click to highlight passage referred to. Unfold extra text Unfold supplementary text
Copy title and link | Comment/query |

Meta-analysis of the alliance–outcome relation in couple and family therapy.

Friedlander M.L., Escudero V., Welmers-van de Poll M.J. et al.
Psychotherapy: 2018, 55(4), p. 356–371.
Unable to obtain a copy by clicking title? Try asking the author for a reprint by adapting this prepared e-mail or by writing to Dr Friedlander at

Amalgamation and review of research findings commissioned by the American Psychological Association reveals that working relationships in couple and family therapies are at least as important as in individual therapies. Practice recommendations will help therapists develop these relationships, augmenting the impacts of some of the most effective ways to treat substance use problems.

Summary [Though not specific to clients with drug and alcohol problems, the principles derived from this review of psychotherapy studies are likely to be applicable, partly because severe substance use problems generally form part of a complex of broader psychosocial problems. Addictions work may not necessarily best be conceptualised as psychotherapy, but there is a therapeutic element to it which makes these findings relevant to keyworkers and counsellors. This review updates an earlier version also in the Effectiveness Bank.]

Key points icon

Key points
From summary and commentary

Commissioned by a task force of the American Psychological Association, this review amalgamated findings relating outcomes to the strength of the collaborative working relationships in family and couples therapies (‘alliance’) between clients and between client and therapists.

The link between alliance and outcomes was moderate and statistically significant. Assuming a causal connection, relative to other influences it would be an important determinant of patient progress.

Though causality cannot be established by the types of studies included in the analysis, the safest stance is to presume that how the therapist is and behaves does influence how well couples and families do partly via the alliances they help generate with and between clients.

The featured review is one of several in a special issue of the journal Psychotherapy devoted to features of the therapist-client relationship related to effectiveness, based on the work of a task force established by the American Psychological Association. This particular review examined the links between outcomes of therapy conducted jointly with couples or families, and the working relationships or ‘alliances’ between clients and between clients and therapists. It complements a similar review of the alliance–outcome link in therapy for individual patients.

The concept of alliance was originally developed in individual psychotherapy. An influential formulation sees it as a collaborative stance composed of: agreement between patient and therapist on the goals of therapy; agreement on the tasks to be undertaken during therapy; and an emotional bond between patient and therapist. In concept and in practice, the working alliance in couple/family therapies is both similar to and different from that in individual psychotherapy. In both, therapeutic goals and tasks need to be discussed and agreed early and as therapy evolves, and in both the therapist needs to ‘click’ emotionally with the patient(s). Only in joint couple/family therapies, however, does the therapist need to develop and nurture multiple alliances simultaneously. Alliances develop on an individual level (patient-with-therapist) and a group level (group-with-therapist; within the patient group). These alliances interact, particularly when patients are in conflict, or the alliance of one is distinctly stronger than that of another. Family members who view their experience with the therapist very differently can end up polarised about the value of the therapy itself.

Measuring the alliance in family and couple therapy

The method most often used to assess family members’ individual alliances with the therapist is the Session Rating Scale – Version 3. The short version presents clients with four pairs of statements, one a negative, the other a positive, and asks them to rate their experience of today’s session by marking how close this was to either extreme. The positive end of each dimension is reproduced below:
• “I felt heard, understood, and respected.”
• “We worked on or talked about what I wanted to work on or talk about.”
• “The therapist’s approach is a good fit for me.”
• “Overall, today’s session was right for me.”

Couple versions of the Working Alliance Inventory offer examples of the questionnaires used to assess the multiple alliances in couple/family therapies. Sample statements below are taken from a short version completed by clients, who respond by choosing options ranging from “seldom” to “always”. In relation to thoughts or feelings about their counsellor, some of the questions are about the patient’s own perceptions, others about what their partner may think or feel, and others about what the couple think or feel.
• “The therapist and I trust one another.”
• “The therapist and I agree about how best to use the time in therapy.”
• “My partner and the therapist like each other.”
• “My partner and the therapist agree about the things we will need to do in therapy to help improve the situation.”
• “As a couple, we agree with the therapist about how best to use the time in therapy.”
• “The therapist and my partner and I (as a couple) are honest with each other.”

The review incorporated a meta-analysis which amalgamated results from relevant studies to provide estimates of the overall strength of the link between alliance and outcomes, and to be able to probe for influences on the strength of that link. The strength of the alliance–outcomes link was calculated as a correlation coefficient, an expression of the degree to which outcomes co-varied with the solidity of the alliance. The chosen metric ranges from -1 (perfect negative co-variation, meaning that as one side of the link gets larger the other diminishes) to +1 (perfect positive co-variation, meaning that as one side of the link gets larger so does the other). Correlation coefficients were also converted to effect sizes. Effectively these metrics indicate how influential the alliance had been if causally linked to outcomes.

Included were studies of patients receiving family or couple therapies which assessed alliance and related it to outcomes in a way which could be amalgamated with results from other studies. Searches uncovered 48 studies of 40 samples of patients (32 in family therapy and eight in couple therapy) totalling 2,568 families and 1,545 couples. Target problems ranged from parent–adolescent communication difficulties to substance use (10 of the 48 studies), child abuse or neglect, schizophrenia, or general family or couple dysfunction or distress. Most studies assessed alliance early in treatment. Outcomes included retention in treatment as well as mid-treatment improvement and client change at or after the end of treatment. Generally outcomes assessed the quality of family/couple relationships or psychological well-being and were assessed at the end of treatment, but in 12 studies retention was a component of a composite outcome or the sole outcome. Because alliance might be measured at several points in therapy and related to multiple outcomes, the 48 studies yielded 491 assessments of the strength of the link between alliance and outcomes. The analysis adjusted for the non-independence of assessments from within the same study.

Main findings

Across all the studies the strength of the link between alliance and outcomes equated to a statistically significant correlation of 0.30 and a moderate-strength effect size of 0.62, an association which could be used to predict about 8% of the differences in outcomes. [These figures were almost identical in studies of individual adult psychotherapy.] In other words, the more solid the working relationship or bond between therapist and clients and within the family or couple, on average the greater the progress clients make and/or the longer they are retained in therapy, though this tendency is not found in every study or every family/couple–therapy pairing. However, this finding seemed to have been affected by a disproportionate failure to include studies which found non-significant or smaller alliance–outcomes links. When this possibility was adjusted for, the correlation fell substantially to 0.18. Though the strength of the alliance–outcomes link varied substantially across the studies, in only four of the 48 was it negative, and then usually very slightly.

Seven studies reported on the link between outcomes and the disparity in the strength of the alliances reported by different members of the couple or family. A statistically significant correlation of 0.32 (equating to a moderate effect size of 0.67) indicated that on average the greater the disparity, the worse the outcomes. Again, the strength of the disparity–outcomes link varied substantially across the studies, but adjustment for possibly missing studies did not significantly affect the finding.

Next the analysts looked for factors reported by the studies which might have accounted for differences in the strength of the alliance–outcomes link. Correlations were stronger when the targeted child in family therapy was relatively younger or adults (whose ages were averaged) relatively older, when patients were either specifically recruited for the study or seeking help rather than ordered or legally coerced into treatment, and when the sample included relatively more fathers/male caregivers or male partners. At a substantial 0.53, of the therapy models the correlation was much stronger for cognitive-behavioural than attachment/emotion-focused or integrated or mixed therapies (correlations 0.2 to 0.3) or (at 0.12 the weakest correlation) structural/functional and multisystemic models. Despite these differences, the significant correlations indicate that alliance was associated with outcomes within each of these major approaches. Its correlation with outcomes was strongest when alliance with the therapist was assessed at the level of the family/couple system rather than based on the perceptions of individual patients, either averaged or (lowest correlation) that of one individual. At 0.40, the correlation was strongest when a scale which taps into family dynamics (and uniquely, assesses how safe clients feel ‘opening up’ in therapy; eg, “There are some topics I am afraid to discuss in therapy”) was used to measure the alliance rather than other methods. Averaging alliance measures over time produced the strongest link to outcomes (correlation 0.42), but at 0.24 the correlation remained statistically significant when assessed early in treatment. The link with alliance was also strongest when outcomes related to parental skills or family/couple functioning rather than symptom reduction at an individual level, the attainment of therapeutic goals, or retention in treatment.

In contrast, the strength of the alliance–outcomes link did not significantly vary depending on whether therapists in a study tended to be male or female, nor whether the therapy was for families versus couples. Also the link remained significant whether alliance was assessed on the basis of the clients’ responses or those of the therapist or an observer, and also remained significant whether outcomes were assessed not just at the end of treatment but in a later follow-up.

Patients contribute substantially to the formation of alliances in couple/family therapy. Though these influences were not checked by amalgamating research findings, the reviewers found evidence that the male partner’s perceptions of the alliance are more influential those of the female, that alliances are difficult to sustain when clients’ problems are severe, and that when the adolescent child is the focal client, their perceptions may be most influential. Clients tend to do best when they feel comfortable, have a trusting emotional bond with their therapists, and stay engaged in the negotiation and renegotiation of therapy goals and tasks. Family members who seem to feel safe with one another in the therapeutic context tend to be emotionally expressive and vulnerable; they ask each other for feedback, encourage openness, and disclose thoughts, feelings and memories that may never before have been shared. On the other hand, hostility, sarcasm and prolonged cross-blaming tend to signal a troubled within-family alliance.

Practice recommendations

Studies found by the review shed little light on whether the association between alliance and outcomes is due to stronger alliances actually causing better outcomes. Nevertheless, the reviewers’ assessment was that strong, balanced therapeutic alliances improve outcomes of couple and family therapies both during and after the end of therapy, and that this is the case across different theoretical approaches guided by treatment manuals and couple and family therapies as normally practised. On that basis they offered practice recommendations for therapists including the following:
• Each person’s alliance matters. Therapists are strongly advised to recognise that balanced alliances facilitate therapy, and that continual monitoring of the strength of the alliance with each individual and within the family/couple unit is essential for therapeutic success. Alliances interact, and patients closely observe how other participants are relating to the therapist. It is particularly important to ‘pull in’ quiet or reluctant family members.
• By definition, the therapeutic alliance is the result of patient–therapist reciprocity. It is essential for therapists to identify markers of patients’ receptivity to therapeutic change attempts.
• Be particularly alert to the strength of the alliance within the couple or family unit, as this seems the most crucial for engagement, retention and ultimate treatment success. Couples and families who enter therapy with a strong shared sense of purpose seem to have the greatest chance of success. Research has shown that even highly experienced therapists tend to try to engage and connect with individual clients rather than the family/couple system as a whole, overlooking the quintessentially systemic feature of couple and family work.
• Use safety- and connection-enhancing interventions to strengthen the within-family/couple alliance and engagement in the joint treatment process. When engagement is weak, the therapist can non-defensively explore the reasons behind a patient’s resistance.
• Identifying clients’ shared feelings and experiences and validating their common struggle can strengthen the within-family/couple alliance. Then the therapist can suggest overarching goals shared by all the clients.
• ‘Split’ alliances (when the alliance with the therapist is stronger for one client than for another) are common, and the more disaffected client may actually be more negative than they disclose in therapy. Therapists can take steps to repair the alliance and prevent drop-out; focusing on the emotional bond with the disaffected patient may prove most helpful.
• Trying not to weaken the alliance is not enough; drop-out can occur if therapists fail to use alliance-enhancing responses when a ‘rupture’ (such as when a patient questions the value of treatment or responds to another family member defensively or sarcastically) is evident. Examples include indicating that some positive change has already taken place, expressing interest in the patient’s life apart from therapeutic concerns, acknowledging that psychotherapy involves taking risks, or emphasising family members’ commonalities or shared experiences.
• Even patients ordered or coerced into treatment can form strong working alliances and once meaningfully engaged, can benefit considerably. Therapists can enhance these patients’ engagement by requesting, rather than imposing, in-session and homework tasks.
• Be aware that parents closely observe their children’s reactions and tend to evaluate improvements based on their assessments of the child’s alliance with the therapist, but that adolescents tend to be more attuned to their own reactions to the therapist than those of their parents.
• A poor alliance with adolescents can be improved by adopting the stance of the less powerful partner in the relationship, avoiding domineering or authoritarian responses – though aligning too strongly with an adolescent may harm alliances with the parents, particularly if they see the treatment as solely about changing the child, not challenging their own behaviour.
• Engaging reluctant adolescents is promoted by helping them define personal treatment goals, presenting as their ally, not forcefully challenging their resistance, and by empathically and non-defensively encouraging parents to support the adolescent’s involvement in treatment.
• A key (but not one universally applicable) to success with heterosexual couples may be working early in treatment to create a particularly strong alliance with the man, particularly if the woman initiated help-seeking. Later it seems important to ensure the female partner continues to be invested in therapy.
• Faced with high emotional reactivity and conflict, possibly the most important safety intervention is either to ask one (or more) patients to step out of the room for a brief period or to conduct alternating sessions with different family ‘sub-systems’. Drop-out is likely when parents/couples feel highly unsafe.
• Asking patients after each session to complete a brief questionnaire assessing the alliance can enhance it. When in this way family members disclose their private experience of the joint context, the therapist is better prepared to directly address alliance strains or ruptures.

Findings logo commentary Though research is not definitive, for reasons explained below the safest stance for trainers, supervisors, therapists, counsellors and clients, is to presume that good working relationships are important determinants of success in couple and family therapies, and that nurturing, maintaining, and as needed, re-establishing these relationships, are core tasks. The recommendations in the featured review ( above) aim to aid therapists in those tasks.

The reviewers’ practice recommendations are based on the likelihood of a causal link ( below) between alliance and patients’ progress, which can be leveraged by the therapist to augment that progress. In other words, that how the therapist is and behaves affects how well their patients do, and does so partly via the collaborative bond they help form between themselves and the clients and between the joint clients. Client–therapist bonds can be seen as the convergence or emergent result of the components also addressed by reviews (listed at the end of this analysis) commissioned by the same American Psychological Association task force, including empathy, establishing collaboration and agreement on goals, repairing ruptures in the client–therapist relationship, demonstrating positive regard for the patient, and conveying the credibility of the therapy.

In family and couple therapies, the therapist faces the additional and crucial task of establishing a productive working relationship between the clients themselves and engaging each individual without alienating another. How tricky that can be is instanced by the recommendation that “a key to success may involve working early on to create a particularly strong alliance with the male partner”, while later ensuring “that the female partner continues to be invested in therapy” – a difficult balancing act which risks alienating one or even both of the dyad. Expressing this recommendation in gender-neutral terms more clearly exposes its complexity and generalises it to other couples. Reading only slightly between the lines, it is being said that when one of the partners feels themselves ‘dragged into’ therapy, it is important to rapidly engage them in the process before they totally disengage – a priority which the partner who initiated therapy-seeking might well appreciate. However, this process should not go so far or last so long that the initiating partner feels neglected, that the therapist in not on their side, or that there has been no pay-off from first focusing on their partner.

Family and couples therapies for substance use problems

In the substance use sector, that the extra investment and skill needed for family/couple therapies is worthwhile is indicated by the greater success of couple therapies than individual therapies for adults. Among the minority of patients for whom working in couples is suitable, acceptable and safe, the advantages can extend beyond substance use to the family and the children. Behavioural couples therapy was one of only two psychosocial therapies recommended by Britain’s National Institute for Health and Clinical Excellence (NICE) for the treatment of problems related to illicit drug use. Among other therapies, NICE guidance on the treatment of alcohol problems also recommends the same approach. For adolescents too, multi-prong therapies centred on the family have emerged (1 2) as probably the most effective approach. Britain’s National Institute for Health and Care Excellence has recommended multi-prong programmes centred on the family for problem-drinking children with other major problems or limited social support, signalling their particular suitability for the most severely affected and multiply problematic youngsters.

In the featured review, across all 10 studies of patients being treated for substance use problems the alliance–outcomes link accumulated to statistically significant correlation of 0.19, considerably lower than the 0.33 among samples seeking help with their couple or family relationships. However, whether the substance use focus was the reason for a weaker link is unclear, because it overlapped with other factors associated with weaker alliance–outcomes links. Three of the four studies which recorded a negative correlation between alliance and outcomes were among the 10 substance use studies. These also shared another feature associated with a weaker link – that the therapy was based on structural/functional or multisystemic models – as did nine of all 10 substance use studies. All the substance use samples included people who had been coerced into treatment and all but one were of family therapy, suggesting that an older child who did not seek the help they were led in to was often the locus of the presenting problem, factors associated with weaker links between alliance and outcomes.

Despite their advantages, formal couple or family therapies for substance use problems seem rare in the UK, partly because of the training, supervision and competencies needed to safely and effectively handle the complexities described in the featured review.

Probably but not necessarily causal

Given the nature of the studies which supported the reviewers’ recommendations, causality cannot be considered proven, but for at least two reasons it seems likely. First is the consistency of the association between the strength of the alliance and outcomes. Though sometimes very small and non-significant, in only four of the 48 studies amalgamated by the review was this relationship negative, and in three only marginally. Second is the strength of the association. Across all the studies, assuming a causal connection and in the context of other influences, it would be a relatively important determinant of patient progress, often exceeding many others including the type of therapy. Last is the plausibility of the proposition that establishing a good working relationship will help keep patients in therapy and actively working with the therapist and with each other towards agreed therapeutic goals, and that this greater opportunity for therapy to work will often translate into it actually working better. Additionally, there seems little or nothing to lose and possibly much to gain from promoting good working relationships with and between clients, but nothing to gain and possibly much to lose from failing to do so.

However, causality could not be established by the types of studies included in the featured analysis. Generally these documented the development of client perceptions of the alliance during therapy and related these perceptions to outcomes. Such studies are generally unable to eliminate the possibility that (for example) families and couples who were going to do well in any event were more likely to relate well to each other and to their therapists, and that therapists too would be more able to work well with them. The review itself offers examples of how this might happen, noting that patients with less severe problems and who are more comfortable in therapy, more able to focus on problems other than their own, and more trusting, form stronger alliances. Conceivably, such attributes mean they will do better in therapy, regardless of the relationships they form with the therapist. Perhaps too, a part of the link between alliance and outcomes is due to patients who were already doing well in therapy feeling appreciative of their therapists (though this is argued against by the link remaining even when alliance is assessed early), or therapists more capable of generating these feelings being more competent in other ways. In these scenarios, alliance would remain associated with better outcomes, but not because it helped cause them. As causality theorists have illustrated, “Thunder correlates with power outages, but thunder does not cause power outages. To distinguish causal from noncausal correlations, it is important to control for alternative causes.” Without effectively random (and almost certainly unethical) allocation of patients to high- and low-alliance therapies, these ‘alternative causes’ cannot completely be eliminated.

Such considerations are common to individual psychotherapy too, but in family and couple therapies there is the added obscurant that the outcomes are often framed in terms of the relationships between the couple or the family members and the adequacy of their joint functioning, yet in microcosm this is partly what is also assessed by measures of the alliance which tap family dynamics, the measures most closely related to outcomes. This comes closer than is comfortable to a meaningless tautology – better relationships and functioning assessed in one way merely meaning better relationships and functioning assessed in another way – or in the case of ‘split’ alliances, disagreements over relationships assessed in one way meaning disagreement and conflict assessed in another way.

Listed below are analyses of the other reviews commissioned by the American Psychological Association task force.
Overall conclusions based on the other 16 reviews
The ‘real’, person-to-person relationship
Therapist–client alliance
Alliance in child and adolescent therapy
Cohesion in group therapy
Therapist empathy
Goal consensus and collaboration
Positive regard
Therapist congruence/genuineness
Therapist self-disclosure and ‘immediacy’
Therapist and client emotional expression
Repairing ruptured alliances between therapists and clients
Managing ‘countertransference’
Treatment credibility
Treatment outcome expectations
Feeding back client progress data therapists

Last revised 28 March 2019. First uploaded 12 December 2018

Open Effectiveness Bank home page

Top 10 most closely related documents on this site. For more try a subject or free text search

REVIEW 2011 Evidence-based psychotherapy relationships: Alliance in couple and family therapy

REVIEW 2011 Evidence-based psychotherapy relationships: The alliance in child and adolescent psychotherapy

REVIEW 2018 Meta-analysis of the prospective relation between alliance and outcome in child and adolescent psychotherapy

ABSTRACT 2011 Evidence-based therapy relationships: research conclusions and clinical practices

REVIEW 2018 Positive regard and psychotherapy outcome: a meta-analytic review

REVIEW 2018 Congruence/genuineness: a meta-analysis

REVIEW 2018 The alliance in adult psychotherapy: a meta-analytic synthesis

REVIEW 2018 Cohesion in group therapy: a meta-analysis

REVIEW 2018 Therapist empathy and client outcome: an updated meta-analysis

REVIEW 2018 Meta-analyses of the relation of goal consensus and collaboration to psychotherapy outcome