Evidence-based relationships and responsiveness

1. Introduction

Version followed: the Third APA Interdivisional (Divisions 12 and 29) Task Force on Evidence-Based Relationships and Responsiveness, chaired by John C. Norcross (Norcross & Lambert, 2018, 2019, Vol. 1, therapist contributions; Norcross & Wampold, 2019, Vol. 2, therapist responsiveness). Practice detail is taken from Norcross's Task Force slides, the Norcross & Wampold (2013) Compendium of Treatment Adaptations (based on the second Task Force, 2011), a 2020 Spanish clinical synthesis of the 2019 volumes (Muela Aparicio & Sansinenea Méndez), and the Safran–Muran rupture-resolution model as published by the Center for Alliance-Focused Training. Where the editions disagree, this guide says so.

This is not a single therapy. It is a research-based framework that sits on top of whatever treatment you already deliver. It has two parts. (a) Evidence-based relationship elements: therapist behaviours and qualities, such as the alliance, empathy and collecting client feedback, that predict outcome across treatment types. (b) Responsiveness (treatment adaptation): tailoring the relationship and the treatment method to transdiagnostic patient characteristics beyond diagnosis, such as reactance, stage of change, preferences, culture, coping style, religion or spirituality, attachment and expectations. [source: Scranton Task Force page; Society for the Advancement of Psychotherapy (SAP) Task Force page; Norcross & Wampold, 2013]

How change happens, as the Task Force sees it: the relationship "makes substantial and consistent contributions to outcome" whatever the treatment type, and it works together with treatment methods, patient characteristics and clinician qualities. Adapting the relationship and the method to the whole person, not just the diagnosis, improves effectiveness. Muela & Sansinenea paraphrase the Task Force's position as: using the same treatment for every patient and every disorder is contraindicated. The Task Force says explicitly that its conclusions are not practice standards; they are to be combined with the other clinical evidence in each case. [source: Norcross & Lambert, 2018 abstract (Ovid); SAP page; Muela & Sansinenea, 2020]

2. Framework

The Task Force does not publish phases, a session count or a fixed format. It applies to individual, youth, couple/family and group therapy, and to any orientation (CBT, psychodynamic, humanistic, systemic). Its own practice recommendations form a loop that runs through the whole treatment [source: SAP page; Norcross slides]:

Make creating and cultivating the relationship a primary aim of treatment, using the elements rated demonstrably and probably effective.
Adapt the relationship and the treatment to patient characteristics where the evidence supports it.
Routinely monitor the patient's response to the relationship and to the treatment, so that you can repair collaboration, change technique and prevent premature termination.
Combine evidence-based relationships with evidence-based treatments, adapted to the patient.

Evidence ratings, used to set priorities (third Task Force, Vol. 1, as listed in Norcross's slides; the Vol. 1 abstract gives 9 demonstrably effective, 7 probably effective and 1 promising element, which matches the list) [source: Norcross slides; Norcross & Lambert, 2018 abstract]:

Demonstrably effective: alliance in adult individual therapy; alliance in child and adolescent therapy; alliance in couple and family therapy; cohesion in group therapy; empathy; collecting and delivering client feedback; goal consensus; collaboration; positive regard and affirmation.
Probably effective: real relationship; emotional expression; congruence and genuineness; repairing alliance ruptures; managing countertransference; promoting treatment credibility; cultivating positive expectations.
Promising, insufficient research: self-disclosure and immediacy.

Adaptation (responsiveness) ratings. The second Task Force (2011) rated reactance, preferences, culture and religion/spirituality as demonstrably effective, and stages of change and coping style as probably effective [source: SAP page]. The third Task Force (Vol. 2, 2019) has chapters on attachment style, cultural adaptation, cultural humility, coping style, preferences, reactance, religion/spirituality and stages of change [source: Muela & Sansinenea, 2020]. [uncertain] The Vol. 2 rating table is not drawn on, so the 2019 rating of each adaptation is not given here.

Decision procedure for the clinician

The stages below are [synthesis], organised from the Task Force recommendations and the Compendium.

Stage	Goal	What you do	Move on when
A. Intake and assessment (sessions 1–2)	Know the person beyond the diagnosis	Assess preferences, stage of change, reactance, coping style, cultural identity and preferred language, religious or spiritual importance, outcome expectations and attachment style. Start a baseline outcome measure and an alliance measure.	You have a working profile and a baseline score.
B. Contracting and alliance formation (sessions 1–3)	Agree goals and tasks and form a bond	Negotiate goals together; give a credible, culturally fitted rationale; accommodate strong preferences or openly discuss disagreement; set directiveness and structure to match reactance, coping style and stage.	The client endorses the goals and the method, and alliance scores are adequate.
C. Working phase (every session)	Deliver the chosen treatment responsively	Use empathy, positive regard and congruence; facilitate emotional expression; collect feedback every session; watch for rupture markers; manage countertransference.	Not applicable: this stage continues throughout.
D. Course correction (whenever triggered)	Prevent deterioration and dropout	Triggers are a not-on-track signal, a drop in alliance scores or a rupture marker. Repair the rupture, review the stage of change and reactance, change the method or format, or review the case with the client.	Scores are back on the expected track, or the plan has been jointly revised.
E. Ending	Consolidate gains	Review progress against baseline with the client; process the relationship; agree a plan for maintenance.	Not applicable.

3. Guiding principles

[source: Task Force conclusions and recommendations on the SAP page and in Norcross's slides; adaptation principles from Norcross & Wampold, 2013; Muela & Sansinenea, 2020]

The relationship is a treatment, not a preliminary. Plan and monitor the relationship as deliberately as you plan techniques.
The relationship and the method act together. Choose techniques with an eye to what they do to the alliance, and vice versa.
Treat the person, not just the disorder. Diagnosis alone is an incomplete basis for choosing treatment, so assess the transdiagnostic characteristics.
Adapt, do not standardise. Change directiveness, structure, focus, format and stance to fit each client, and change them again as the client changes.
Monitor routinely and formally. Measure progress and alliance every session; do not rely on your own impression, which tends to miss deterioration [synthesis from the feedback rationale in Norcross's slides].
The client's view counts most. Client-rated alliance, empathy, regard and cultural competence predict outcome better than therapist ratings; therapist self-rated cultural competence did not predict outcome (Muela & Sansinenea). In practice, ask the client.
Resistance is information about fit. Treat resistance or reactance as a possible sign that the method or level of directiveness does not match the client; it is the therapist's problem to solve.
Collaborate and be transparent. Share decisions, including your disagreement with a preference.
Combine evidence-based relationships with evidence-based treatments. Do not drop specific methods; deliver them inside a responsive relationship.
Responsiveness is clinical judgement, not an algorithm. Muela & Sansinenea stress that personalisation cannot be automated and should not be applied mechanically.
Avoid discredited relationship behaviours. These are a confrontational style, frequent interpretations, hostile or blaming processes, assumptions, therapist-centredness and ignoring early ruptures (Norcross's slides, from reviews and Delphi polls).
The findings are not practice standards. Weigh them alongside all the other evidence about this client.
Watch your own reactions. Countertransference management, and not matching directiveness to your own reactance, are part of the method (Compendium; Norcross's slides).

4. Intervention inventory

The framework has two kinds of intervention. Part A covers the relationship elements, which are therapist behaviours. Part B covers the adaptation methods, which are ways of tailoring treatment. The Task Force defines and rates the elements, and its slides and the Compendium give some practice pointers. The Task Force sources give few step-by-step procedures. Steps are therefore marked [synthesis] unless a source is named, and all example wording is illustrative.

Part A. Relationship elements

A1. Building the therapeutic alliance (individual adult therapy)

Purpose: The alliance is agreement on goals, consensus on tasks and an emotional bond. It is rated demonstrably effective in adult, youth and couple/family therapy. Build it in every case, from session 1. [source: UCLA info sheet (Norcross & Hill, 2004); Norcross slides]
Steps [synthesis]: (1) In session 1, ask what the client wants to be different, and reflect it back in their words. (2) Explain how you propose to work and check whether this makes sense to them (tasks). (3) Convey interest, warmth and reliability (bond). (4) Give a brief alliance measure at the end of sessions, or ask directly. (5) Respond to any low or falling score within the same or the next session (see A14).
Example language (illustrative): "Before we finish, how did today feel? Did we work on what matters most to you, in a way that suits you?"
Watch for: Client-rated alliance predicts outcome better than your own rating. Signs that it is working are agreed goals, active participation and attendance. The Task Force found weaker alliance–outcome links in substance use and eating disorders [source: Norcross slides], so expect to have to work harder on the alliance there.

A2. Alliance with children and adolescents, and with their parents

Purpose: The alliance with the young person and the alliance with the parent were both associated with outcome, at similar levels. Build both. [source: Norcross slides]
Steps [synthesis]: (1) Hold separate time with the young person and with the parent or carer. (2) Agree goals that each party can endorse, and name any differences openly. (3) Explain confidentiality limits clearly to both. (4) Check the alliance with each party separately.
Example language (illustrative): "Your mum wants fewer arguments. What would you like to be different, even if it's different from what she wants?"
Watch for: The young person seeing you as "on the parents' side".

A3. Alliance in couple and family therapy

Purpose: The alliance is with each member and with the system as a whole. Couple and family therapy showed similar effects. [source: Norcross slides]
Steps [synthesis]: (1) Give each member speaking time and empathic reflection. (2) Frame goals as shared by the system. (3) Watch for split alliances, where one member is strongly allied and another weakly. (4) Get alliance ratings from every member, not only from the most verbal.
Example language (illustrative): "I want to make sure this works for both of you. Does either of you feel I've been seeing it more from the other's side?"
Watch for: One member withdrawing or dropping out. This is an early marker of a split alliance.

A4. Group cohesion

Purpose: Group cohesion is the group-therapy equivalent of the alliance, and is rated demonstrably effective. Interpersonally oriented group leaders showed especially strong cohesion–outcome links. [source: Norcross slides]
Steps [synthesis]: (1) Prepare members before the first session (purpose, norms, what to expect). (2) Early on, link members' experiences to each other. (3) Model and encourage member-to-member feedback, not only leader-to-member. (4) Address conflict openly and safely. (5) Monitor cohesion and alliance with brief measures.
Example language (illustrative): "Sam, you nodded when Priya described that. Does it connect with something for you?"
Watch for: Subgrouping, scapegoating and silent members.

A5. Goal consensus

Purpose: Therapist and client agree on treatment goals and expectations. Rated demonstrably effective. [source: UCLA info sheet; Norcross slides]
Steps [synthesis]: (1) Elicit the client's goals in concrete, observable terms. (2) Share your formulation and the goals you would suggest. (3) Negotiate a shared list. (4) Write the goals down and review them at set intervals, for example every 4–6 sessions. (5) Renegotiate them when the client's priorities shift.
Example language (illustrative): "If therapy went well, what would you be doing in three months that you're not doing now?"
Watch for: A client agreeing to goals that are really yours. Look for flat agreement and little follow-through.

A6. Collaboration

Purpose: The client is an active partner in the therapy. Rated demonstrably effective. [source: UCLA info sheet; Norcross slides]
Steps [synthesis]: (1) Set session agendas together. (2) Offer choices between methods. (3) Design homework with the client and review it. (4) Ask for the client's view before you give yours.
Example language (illustrative): "Here are two ways we could tackle this. Which feels more like you?"
Watch for: Homework that is never done. This may mean the task was not jointly owned.

A7. Empathy

Purpose: Understanding the client's thoughts and feelings from the client's point of view, and communicating that understanding. Rated demonstrably effective. Norcross's slides note that the empathy–outcome link was stronger in CBT than in experiential, humanistic or psychodynamic therapy. Empathy therefore matters inside structured treatments too. [source: UCLA info sheet; Norcross slides]
Steps [synthesis]: (1) Listen for the client's felt meaning, not just the content. (2) Reflect it in tentative language. (3) Check the reflection and accept correction. (4) Adjust the depth and frequency of empathic responses to the client: some clients find intense empathy intrusive.
Example language (illustrative): "It sounds like what hurt most wasn't the criticism itself, but that it came from him. Is that close?"
Watch for: The client correcting or elaborating, which shows engagement. Rely on the client's sense of being understood, not on your own belief that you understood.

A8. Positive regard and affirmation

Purpose: Warm, unconditional acceptance of the client's experience. Rated demonstrably effective. Client ratings of regard predicted outcome best, and effects were larger in mood and anxiety disorders than in severe mental illness. [source: UCLA info sheet; Norcross slides]
Steps [synthesis]: (1) Convey respect and warmth in tone and attention. (2) Affirm efforts and strengths specifically. (3) Separate acceptance of the person from agreement with every behaviour.
Example language (illustrative): "Coming in today, when everything in you wanted to cancel, took real effort."
Watch for: Hollow praise. Affirmation works when it is specific and believed.

A9. Congruence and genuineness

Purpose: The therapist is personally integrated in the relationship and conveys their own personhood. Rated probably effective. Effects were larger for older, licensed and more experienced therapists. [source: UCLA info sheet; Norcross slides]
Steps [synthesis]: (1) Let your words, tone and body language match. (2) Do not feign feelings. (3) When your reaction is relevant, share it briefly and in the service of the client (see A16).
Example language (illustrative): "I notice I'm feeling a bit lost in what we're doing. Can we step back?"
Watch for: Genuineness turning into self-expression for your own needs.

A10. Real relationship

Purpose: The personal relationship beyond transference and technique: genuineness and realistic perceptions between client and therapist. Rated probably effective. [source: Norcross slides; definition [uncertain]: paraphrased from the general literature]
Steps [synthesis]: Be a real person within the professional frame. Acknowledge real events, such as your lateness or a holiday, honestly. Respond to the client as an adult with their own perceptions.
Example language (illustrative): "You're right, I was distracted at the start today. I'm sorry; I'm with you now."
Watch for: Boundary blurring.

A11. Facilitating emotional expression

Purpose: Encourage the client to experience and express emotion in session. Rated probably effective. Norcross's slides caution that these findings are associations, not proof of cause. [source: Norcross slides]
Steps [synthesis]: (1) Notice emotion as it appears (voice, eyes, posture). (2) Slow down and invite the client to stay with it. (3) Help put it into words. (4) Help the client regulate it if it escalates, then reflect on what it means.
Example language (illustrative): "Your voice changed just then. Can we pause there? What are you feeling right now?"
Watch for: Overwhelm. Keep emotional work within the client's window of tolerance [synthesis], and use more structure with anxiously attached clients (B7).

A12. Cultivating positive outcome expectations

Purpose: The client's expectations of gain, of procedures, of the therapist's role and of length of treatment are related to outcome. "Cultivating positive expectations" is rated probably effective. [source: UCLA info sheet; Norcross slides; Compendium]
Steps: (1) Assess the client's expectations of outcome, process, role and duration (Compendium; UCLA). (2) Correct mistaken role or length expectations early [synthesis]. (3) Build realistic hope, for example by linking the client's past successes to the plan [synthesis]. (4) Treat changes in expectations during therapy as outcomes too, since they can change (Compendium).
Example language (illustrative): "Many people with what you're describing get real relief. Let's set a first marker so we can both see whether it's working."
Watch for: False promises. Expectations should be optimistic but credible.

A13. Promoting treatment credibility

Purpose: The treatment rationale makes sense to the client. Rated probably effective. Few studies looked at therapist credibility itself. [source: Norcross slides]
Steps [synthesis]: (1) Explain how the method is expected to help, in the client's terms. (2) Link it to the client's own account of the problem; the Compendium advises adapting the explanation of the problem to the client's cultural beliefs. (3) Ask how convincing the client finds it, and adjust.
Example language (illustrative): "Does this way of understanding what keeps the panic going fit your experience, or does part of it not ring true?"
Watch for: Polite agreement without belief.

A14. Collecting and delivering client feedback (routine outcome monitoring)

Purpose: Detect clients who are not progressing or are deteriorating, and act before they drop out. Rated demonstrably effective, and one of the two elements where the Task Force considers a causal claim reasonable. The two most-studied systems are the Outcome Questionnaire system (OQ-45, Lambert) and the Partners for Change Outcome Management System (PCOMS; Miller and Duncan). Feedback reduced deterioration and roughly doubled reliable or clinically significant change in clients predicted to do poorly. [source: Norcross slides; Lambert, Whipple & Kleinstäuber, 2018 abstract (USU repository)]
Steps: The process in Norcross's slides runs as follows. (1) Measure progress formally and regularly, usually every session. (2) Compare the client's scores with an expected-recovery benchmark. (3) Feed the result back to the therapist and to the client. (4) Discuss it explicitly in session. (5) Optionally use Clinical Support Tools, which are structured prompts for not-on-track cases covering alliance, motivation, social support and the need for referral. Practical setup [synthesis]: use a licensed or freely licensed measure through its official channel, give it at the start of the session, look at the score before you begin, and save the last minutes for a brief alliance check.
Example language (illustrative): "Your score has gone up a bit for three weeks, which tells me what we're doing isn't helping yet. What's your sense of it? Should we change something?"
Watch for: Measures filled in but never discussed, which forfeits the benefit [synthesis]. Signs it is working: not-on-track clients return to track after a plan change, and fewer clients deteriorate. Do not reproduce measure items; obtain the instruments from their owners. [uncertain] The USU repository record for the 2018 meta-analysis is marked "withdrawn"; The article's current status is [uncertain].

A15. Repairing alliance ruptures (Safran–Muran model)

Purpose: A rupture is a breakdown in collaboration or a strain in the bond. Repairing ruptures is rated probably effective, and training therapists in rupture resolution showed a small benefit. [source: Norcross slides; Center for Alliance-Focused Training (CAFT) pages]
Recognise the type [source: CAFT; Eubanks 2026 workshop description]:
- Withdrawal ruptures ("movements away"): for example, vague or over-compliant answers, changing the topic, minimal responses, or agreeing without engaging. These example markers are a paraphrase of the general literature; the sources do not list them [uncertain].
- Confrontation ruptures ("movements against"): complaints, criticism of the therapist or of therapy, pushing for control.
Steps: (1) Notice the rupture and the shift in your own feelings [source: CAFT]. (2) Respond non-defensively and attend to the relationship directly; adjust your own behaviour and collect feedback [source: Norcross slides]. (3) Immediate strategies: explain the rationale for a task or goal again, change the task or goal to one the client finds more agreeable or relevant, or clarify a misunderstanding [source: CAFT]. (4) Expressive or exploratory strategy, metacommunication, meaning communication about the communication process: name what seems to be happening between you, invite the client to explore it with you, empathise with their negative feelings, and take responsibility for your part [source: CAFT]. (5) Follow the resolution path. With withdrawal, help the client move toward clearer statements of discontent and toward self-assertion. With confrontation, help them move from anger, to disappointment and hurt, to contact with vulnerability and the wish to be cared for. Explore avoidant shifts away from these, which are driven by fear of being too aggressive or too vulnerable [source: CAFT]. (6) Muela & Sansinenea add a choice point: when the bond is strong, acknowledge the rupture directly; when it is weak, start with indirect strategies such as changing the task or goal [source: Muela & Sansinenea, 2020].
Example language (illustrative): "I have the sense I've been pushing this exercise and you've been going along with it. Am I right? I'd rather hear if it isn't working for you."
Watch for: Ignoring early ruptures ("ostrich behaviour") is listed among the discredited behaviours [source: Norcross slides]. Signs it is working: the client voices dissatisfaction more openly, renewed engagement, and a recovered alliance score.

A16. Managing countertransference

Purpose: Countertransference here means the therapist's reactions shaped by their own unresolved issues. Acting it out is associated with poorer outcome, and managing it with better outcome. Managing countertransference is rated probably effective. [source: Norcross slides; UCLA info sheet]
Steps: Build the five capacities the sources name: self-insight, self-integration, anxiety management, empathy and conceptualising ability [source: Norcross slides; UCLA]. In practice [synthesis]: (1) after sessions, note strong pulls such as rescuing, irritation, boredom or attraction; (2) ask which reaction belongs to you and which says something about the client's relational pattern; (3) take it to supervision or personal therapy; (4) use it to understand the client, not to act on it.
Example language (illustrative, internal or supervision): "I keep extending this client's sessions. What am I avoiding?"
Watch for: Repeated boundary drift with one client.

A17. Self-disclosure and immediacy

Purpose: Self-disclosure means sharing something personal: to validate reality, normalise, strengthen the alliance or model an alternative [source: UCLA]. Immediacy means discussing the here-and-now relationship. Rated promising. Clients reported benefits such as a closer relationship and new insight, and negative effects were rare. [source: Norcross slides]
Steps: Disclose judiciously, to meet the client's needs and not yours. Disclosure helps most when the client feels alone or vulnerable; immediacy helps most when relational problems arise [source: Norcross slides]. Keep disclosures brief, then turn the focus back to the client and check its impact [synthesis].
Example language (illustrative): "I've had times of grief too, so I know a little of how disorienting it can be. What's it been like for you this week?"
Watch for: The focus moving onto you, and disclosures the client then feels they must take care of.

A18. Avoiding discredited relationship behaviours

Purpose: Prevent harm and dropout. [source: Norcross slides, from literature reviews and three Delphi polls]
Steps: Audit your own practice, for example on recordings, for: (1) a confrontational style (the issue is style, not content); (2) frequent interpretations; (3) hostile, blaming, pejorative or rejecting processes; (4) assumptions about the client's experience; (5) therapist-centredness; (6) ignoring early ruptures. Norcross's slides also suggest reversing the effective elements to find ineffective behaviours: authoritarian, unempathic or unsupportive conduct.
Example language: Not applicable.
Watch for: Patterns that show up across several clients, rather than one-offs.

Part B. Adaptation (responsiveness) methods

For each adaptation, the general procedure is: assess the characteristic at intake, tailor the treatment to it, reassess as it changes, and discuss decisions openly with the client. [source: Norcross & Wampold, 2013]

B1. Matching directiveness to reactance

Purpose: Reactance is a tendency to be easily provoked by external demands and to resist perceived threats to autonomy. Matching the therapist's directiveness to the client's reactance improves outcome and reduces dropout. [source: Compendium; UCLA; Muela & Sansinenea, 2020]
Steps: (1) Assess the client's trait reactance, and track resistant behaviour in session as a state [source: Compendium; Muela & Sansinenea]. A self-report reactance measure may help; obtain it through its authors [synthesis]. (2) High reactance: use self-directed methods, less structure, fewer directives, and choices. Avoid provoking reactance [source: Compendium]. (3) Low reactance: use explicit guidance and more structure [source: Compendium]. (4) Treat resistance as a sign of mismatch and vary your directiveness systematically [source: Compendium]. (5) Do not set directiveness by your own reactance level, a common novice error [source: Compendium]. (6) Use feedback, especially with reactant clients, to prevent dropout [source: Muela & Sansinenea].
Example language (illustrative): High reactance: "You know your life better than I do. Here are a few things others have found useful; you decide what, if anything, to try." Low reactance: "This week, I'd like you to do the breathing practice twice a day and note your anxiety before and after."
Watch for: "Yes, but" responses, missed homework and arguing. These are cues to step back from directing.

B2. Tailoring to stage of change

Purpose: Match methods and stance to the client's readiness for change. Precontemplators do worse when pushed into action. [source: Compendium; Muela & Sansinenea]
Steps: (1) Assess the stage (precontemplation, contemplation, preparation, action or maintenance) [source: Compendium; UCLA]. (2) Do not treat every client as being in action [source: Compendium]. (3) In early stages, use consciousness-raising and emotion-generating or emotional-relief methods; in action, use skills training, behavioural methods, counterconditioning, stimulus control and reinforcement [source: Compendium; Muela & Sansinenea]. (4) Adjust your stance: a caring, empathic presence in precontemplation, a Socratic teacher in contemplation, an experienced coach in action, a consultant in maintenance [source: Muela & Sansinenea]. (5) Set realistic goals, such as moving the client one stage on, for example from precontemplation to contemplation; motivational interviewing suits this [source: Muela & Sansinenea]. (6) Avoid stage mismatch, where therapist and client are working on different goals [source: Compendium].
Example language (illustrative): For contemplation: "Part of you wants to cut down and part of you isn't sure it's worth it. Can we look at both sides?"
Watch for: Dropout after action-oriented homework is set. This suggests the stage was misjudged.

B3. Assessing and accommodating preferences

Purpose: Clients who receive their preferred treatment do somewhat better and are much less likely to drop out. [source: Compendium; Muela & Sansinenea]
Steps: (1) Ask about preferences at the start of treatment: modality (therapy or medication), method or orientation, format (individual, couple, family or group), therapist characteristics (such as gender, age or religion), role preferences (a listener or an active advisor, homework or none), and length [source: Compendium; Muela & Sansinenea]. (2) Lower barriers to stating preferences, such as lack of information or a belief that expressing them is rude: give brief information and invite them [source: Compendium]. (3) Accommodate strong preferences when ethical and feasible [source: Compendium]. (4) If you think a preference will not serve the client, say so openly and decide together [source: Compendium].
Example language (illustrative): "Some people want mostly to talk things through; others want practical tools and homework. What would suit you? You won't offend me either way."
Watch for: Preferences that conflict with the evidence for the problem, such as avoiding all exposure in a phobia (see Section 5). Flexibility is needed [source: Muela & Sansinenea].

B4. Cultural adaptation and cultural responsiveness

Purpose: Align the treatment with the client's culture. Culturally adapted treatments outperform unadapted ones for ethnic and racial minority clients. [source: Compendium; Muela & Sansinenea]
Steps: (1) Conduct therapy in the client's preferred language, one of the strongest adaptations; use interpreters when the client's proficiency is very low [source: Compendium; Muela & Sansinenea]. (2) Weave cultural content, values and metaphors into the treatment [source: Compendium]. (3) Adapt the explanation of the problem to the client's illness beliefs [source: Compendium]. (4) Offer an ethnically or racially similar therapist when the client wants one and it is possible [source: Compendium]. (5) The more adaptation components, and the more specific the cultural group, the better [source: Compendium; Muela & Sansinenea]. (6) Explore how the client sees you, and routinely ask about cultural fit, since client-rated (not self-rated) cultural competence predicts outcome [source: Muela & Sansinenea]. (7) Make visible, genuine efforts to fit the culture [source: Compendium]. Cultural humility (an open, other-oriented, non-superior stance toward the client's cultural identity) has its own chapter in Vol. 2 [source: Muela & Sansinenea]; its specific practices are [uncertain].
Example language (illustrative): "I'd like to understand how people in your family and community make sense of what you're going through. That will help us find an approach that fits."
Watch for: Stereotyping, which means applying group-level adaptations to an individual without asking them [synthesis].

B5. Matching to coping style

Purpose: Match the treatment focus to the client's habitual coping style. [source: Compendium; Muela & Sansinenea; UCLA]
Steps: (1) From the client's history, assess whether they mainly internalise (self-blame, withdrawal, inhibition), externalise (acting out, blaming others, impulsivity) or are balanced [source: Compendium; Muela & Sansinenea]. (2) For externalisers, use symptom-focused, skill-building methods such as CBT techniques. (3) For internalisers, use insight-oriented, interpersonal or experiential methods. (4) With internalisers, the Compendium advises starting with symptom-reducing, stabilising methods, then moving toward insight work. (5) Choose among several methods that fit the client's personality and preferences [source: Compendium].
Example language (illustrative): For an internaliser: "Alongside the coping skills, I'd like us to look at where this harsh voice toward yourself came from."
Watch for: Externalising clients disengaging from open-ended exploration, and internalising clients feeling skills-only work is superficial.

B6. Religion and spirituality

Purpose: Offer religion- or spirituality-accommodative therapy to clients who want it. It is as effective as secular therapy on mental-health outcomes and better on spiritual outcomes. [source: Compendium; Muela & Sansinenea]
Steps: (1) Ask about the importance of religion or spirituality and whether the client wants it included [synthesis; implied by the sources]. (2) For clients who want it, consider integrating religious or spiritual content, for example their values, practices or texts, into an established method. This is the treatment of choice when spiritual outcomes matter highly, especially for very religious clients [source: Compendium]. (3) Do not expect added religious content to improve psychological outcomes beyond the secular version [source: Compendium]. (4) Be careful when religion is part of the problem, where including it may be counterproductive [source: Muela & Sansinenea]. (5) Reflect on your own beliefs and train in religious and spiritual competence [source: Muela & Sansinenea].
Example language (illustrative): "Your faith sounds central to how you cope. Would you like it to be part of how we work, or would you rather keep it separate?"
Watch for: Imposing your own beliefs or values, in either direction.

B7. Adapting to attachment style

Purpose: More securely attached clients benefit more from therapy, and adapting to insecure attachment is an emerging area. [source: Compendium; Muela & Sansinenea]
Steps: (1) Assess attachment style before or early in treatment [source: Compendium; Muela & Sansinenea]. (2) Anxious attachment: keep a clear, structured frame and avoid overwhelming experiential techniques [source: Muela & Sansinenea]. (3) Avoidant attachment: be active and engaged without pressing the client; committed but not over-involved [source: Muela & Sansinenea]. (4) Procedures focused on interpersonal and intimate relationships have preliminary support for insecure attachment [source: Muela & Sansinenea]. (5) Treat movement toward more secure attachment as an outcome [source: Compendium].
Example language (illustrative): For avoidant attachment: "We can go at whatever pace works. I'll keep checking in, and you can tell me if that feels like too much."
Watch for: Anxious clients escalating contact between sessions, and avoidant clients cancelling after a close session.

B8. Other adaptations listed by the first Task Force (2002, via Norcross & Hill, 2004)

These adaptations appear in the 2004 summary of the first Task Force. Their rating in the 2011 or 2019 lists is [uncertain]. [source: UCLA info sheet]

Functional impairment: use longer, more intensive intervention for more impaired clients.
Anaclitic (sociotropic) versus introjective (autonomous) style: relationship-focused clients may value connection-oriented work; self-definition-focused clients may value autonomy and achievement themes [synthesis on the practical meaning].
Assimilation of problematic experiences: track where the client is on an eight-stage sequence, from warding off or dissociating from a problematic experience to integrating and mastering it, and choose interventions suited to that level [synthesis on the practical use].

5. Handling problems

The Task Force sources name the core responses: monitor routinely, repair ruptures, treat resistance as mismatch, adapt to stage and preferences, and manage countertransference. Most of the specific responses below are my own synthesis and are marked accordingly.

Stuck client or not on track (scores flat or worsening on ROM): (1) Share the data with the client and ask what they make of it [source: Norcross slides on feedback]. (2) Check the alliance and the goals; renegotiate them [source: Task Force recommendation to monitor and repair]. (3) Re-assess stage of change, reactance and coping style, and switch method or format if they are mismatched [source: Compendium]. (4) Run a structured review of alliance, motivation, social support and diagnosis, in the style of Clinical Support Tools; consider referral or more intensive care [source: Norcross slides; specifics are synthesis]. (5) Seek consultation after two or three off-track sessions [synthesis].
Resistance or reactance: Reduce directiveness, offer choices and hand control back to the client. Explore what drives the reactance. Give the client more control over the treatment contract [source: Compendium; Muela & Sansinenea]. Do not argue for change [synthesis consistent with B2].
Overwhelm during emotional work: Slow down, shift to grounding or a cognitive or structured task, and then debrief together on the pacing [synthesis]. With anxiously attached clients, keep more structure from the start [source: Muela & Sansinenea]. For internalising clients, stabilise before deeper insight work [source: Compendium]. In trauma work, Norcross & Wampold (2019) argue that the relationship and adaptations must accompany guideline treatments [source: SAP abstract]. Pacing, readiness and grounding should follow the trauma protocol you are using [synthesis].
Non-adherence (homework, attendance): Check for a stage mismatch, such as action tasks for a contemplator [source: Compendium]. Check whether the task was set collaboratively and fits the client's preferences [synthesis]. With highly reactant clients, offer tasks as experiments the client chooses [synthesis from B1]. Ask about practical and cultural barriers [synthesis].
Ruptures: Follow A15. Use immediate strategies such as re-explaining the rationale, changing the task or clarifying first. Use metacommunication when the bond can hold it. Take responsibility for your part [source: CAFT; Muela & Sansinenea].
Client always rates the alliance as perfect: High scores can hide withdrawal [synthesis]. Make negative feedback explicitly welcome, thank the client for any criticism, and ask one specific improvement question, such as "What one thing could I do differently?" [synthesis].
Therapist discouraged or defensive about feedback: Feedback flags risk early, when change is still possible. Use supervision and treat a not-on-track signal as information, not judgement [synthesis].
Preference conflicts with the evidence (for example, a client with OCD who refuses any exposure): State your concern openly and collaboratively [source: Compendium]. Negotiate a graded or modified version, or a time-limited trial of the preferred approach with agreed review points [synthesis]. Document the shared decision [synthesis].
Cultural mismatch or microaggression: Acknowledge it non-defensively, take responsibility and invite the client's perspective. This is a form of rupture repair [synthesis combining A15 and B4]. Offer a referral to a culturally or linguistically matched therapist if the client prefers [source: Compendium on matching].
Split alliance in couple or family work: Spend balancing time with the less-allied member, reframe goals as shared, and if needed see members individually for a session [synthesis].
Low group cohesion or scapegoating: Name the process, protect the member being scapegoated, and link members' shared concerns [synthesis].
Countertransference enactments: Pause, consult, and if the client was affected, repair with appropriate immediacy (A17) [source: Norcross slides on CT management; repair step synthesis].
Risk (suicidality or self-harm) emerging: Do a structured risk assessment and safety planning according to local standards, including lethal-means counselling (safe storage or removal). A strong alliance supports disclosure, but it does not replace risk procedures [synthesis].
Modality-specific pitfall, adaptation overload: Trying to tailor to every characteristic at once can paralyse decisions. Prioritise the one or two characteristics most relevant to this client's engagement, such as reactance in an argumentative client or stage in an ambivalent one [synthesis; consistent with the Compendium's "one or more" characteristics].
Modality-specific pitfall, treating the framework as a standard: The Task Force states its findings are not practice standards. Integrate them with your diagnosis-specific treatment and clinical judgement [source: SAP page].

Sources

Sources: