Attribution therapy

1. Introduction

Version followed: Försterling's clinical attributional-retraining (AR) framework (1985 review; 1986 paper; 1988 book) as the conceptual base; the German school reattribution-training format (diagnosis, modification, effectiveness check), as summarised for practitioners by Dresel, for the feedback, modelling and reinforcement rules; Perry and colleagues' AR protocol (induction plus consolidation), including its health/ageing form, a single 18-minute modelling video (Parker et al. 2022); the eight-session attribution retraining group therapy (ARGT) of Wang et al. 2011 for clinical groups; Layden's attributional style therapy; Beck's reattribution technique within cognitive therapy; and the four-stage reattribution model for somatic presentations as used in the Morriss et al. 2007 trial (building on Goldberg, Gask & O'Dowd 1989). Misattribution therapy (Storms & Nisbett; Valins & Nisbett) is covered as a historical branch only.

Research note. Sources drawn on: the key source (Wang et al. 2011, ARGT pilot, full text), Parker et al. 2022 (AR video for older adults), Morriss et al. 2007 (reattribution training trial, BJPsych), Dresel's practitioner chapter on school reattribution training, and catalogue records for Perry et al. 2014 and two Manitoba AR theses. Blocks based on these carry [source: …] naming them. Not drawn on: Försterling 1985/1986/1988, Perry & Penner 1990, Haynes et al. 2009, Hall et al., Dweck 1975, Wilson & Linville, Layden 1982, Beck et al. 1979, Seligman 1990, Goldberg et al. 1989 (original paper), Salkovskis 1985. Content resting only on these stays [uncertain].

What it is. Attribution therapy is not a single branded therapy. It is a family of procedures that change the causal explanations people give for their successes, failures, symptoms and life events. It grew from Heider's and Kelley's attribution theory, Weiner's attributional theory of motivation and emotion, Seligman's learned helplessness (and its 1978 reformulation by Abramson, Seligman and Teasdale), and Bandura's self-efficacy theory. It survives mainly as (a) brief AR programmes in education, health and ageing, and (b) the "reattribution" techniques built into cognitive therapy, optimism/resilience training and somatic-symptom work. [source: Försterling 1985, 1988; Antaki & Brewin 1982; Weiner 1985, 1986] [uncertain: texts not drawn on]

How change happens. What a person does after an outcome depends less on the outcome itself than on why they think it happened. Weiner's three causal dimensions do the work:

Stability (fixed vs changeable) drives expectancy: failure blamed on a stable cause ("I'm stupid", "it's my age") predicts more failure, so the person gives up; an unstable cause ("I didn't prepare that way") keeps hope alive.
Locus (me vs outside me) drives self-esteem emotions: pride after success credited to self; lowered self-worth after failure credited to self.
Controllability drives specific emotions and action: guilt (and repair) when a failure was controllable; shame and withdrawal when it was not; anger at others when they are seen as in control.

The helplessness reformulation adds globality (affects everything vs one area): internal, stable, global explanations for bad events are the "depressive" or pessimistic explanatory style. AR shifts unhelpful explanations, usually from stable/uncontrollable causes (low ability, fixed traits, illness, age) towards unstable/controllable ones (effort, strategy, circumstances, learnable skills) for failures, and towards internal causes (ability, effort) for successes. Försterling's key addition: the target is a realistic attribution, not simply an "optimistic" one. When the evidence shows a cause really is uncontrollable, the work moves to other causes the person can act on, rather than encouraging false self-blame. [source: Dresel, Reattributionstraining (aim: realistic and favourable attributions; success to internal causes, failure to variable causes); Parker et al. 2022 (age and health de-emphasised in favour of strategy and effort)] [uncertain: Weiner 1985, Abramson, Seligman & Teasdale 1978 and Försterling 1985, 1988]

2. Framework

AR has no single phase manual. The phases below combine Perry's college protocol (pre-treatment, induction, consolidation, follow-up), Försterling's clinical sequence (diagnose attributions, check whether they are realistic, change them, test them in action) and the three-step structure of school reattribution training: diagnosis (questionnaires for groups; interview or working hypotheses for individuals), modification (commentary/feedback, modelling, reinforcement), and an effectiveness check by repeating the diagnostic measures. [source: Dresel, Reattributionstraining] [uncertain: Försterling 1985, 1988; Perry & Penner 1990; Haynes, Perry, Stupnisky & Daniels 2009; phase labels partly synthesis]

Phase 1: Attributional assessment.

Goals: identify the outcomes that matter (exam failure, rejection, symptoms, inactivity, a traumatic event); collect the client's spontaneous causal explanations; rate them on locus, stability, controllability and globality; link them to the emotion and behaviour that follow (giving up, avoidance, shame, rumination); and judge, with evidence, how realistic each one is (ability level, actual controllability).
Format: interview, plus a standard measure if wanted (see 4.2) and a one- to two-week attribution diary.
Move on when: you can name 1–3 recurring unhelpful attributions, their dimensional profile, the downstream emotion and behaviour, and an evidence-based alternative that would be both more realistic and more functional. [synthesis]

Phase 2: Attributional induction (presenting the alternative).

Goals: give the client a credible alternative explanation through information, persuasion, consensus data and modelling. In Perry's version this follows soon after a real setback (e.g. first test feedback) so the message lands while it matters.
Format: educational and health AR is brief. In the Perry/Chipperfield version for older adults the whole induction was one 18-minute video of two older adults, watched once, with no consolidation task or booster [source: Parker et al. 2022]. College versions add a consolidation exercise after a handout or video (see 4.11, 4.12) [source: Hunter 1997 thesis record] [uncertain: Perry & Penner 1990]. Clinical group AR (ARGT) spreads the work over eight weekly two-hour sessions [source: Wang et al. 2011]. Dweck's classic child study used about 25 daily training sessions [uncertain: Dweck 1975].
Move on when: the client can state the alternative explanation in their own words and apply it to a recent event of their own. [synthesis]

Phase 3: Consolidation and practice.

Goals: make the new explanation the client's own through elaboration (writing, discussion), applying it to new events and graded tasks where effort or strategy changes really do change outcomes. The therapist then reinforces the new attribution at the moment of success or failure.
Format: in-session tasks plus homework; in Perry's protocol, a writing or group-discussion exercise straight after induction. [source: Hall et al. 2004, 2007; Haynes et al. 2009] [uncertain]
Move on when: attribution diaries show the alternative explanations appearing spontaneously, and behaviour has changed (persistence, re-engagement, help-seeking). [synthesis]

Phase 4: Generalisation and maintenance.

Goals: transfer to other life areas; plan for the next setback; spot early relapse signs (a return of "always/everything/me" language).
Format: spaced booster sessions; some educational protocols add a booster later in the term. [uncertain]
End when: the client can run a reattribution on a new setback without prompting. [synthesis]

Typical overall dose. Educational/health AR: a single induction (e.g. one 18-minute video) up to about three sessions [source: Parker et al. 2022; upper figure uncertain]. Clinical group AR: the ARGT pilot ran 8 weekly 2-hour sessions with outpatient groups mixed by diagnosis (major depression, generalised anxiety, OCD), one lead therapist (a registered clinical psychologist), a co-therapist who recorded whether each planned step happened and reviewed this with the lead before each session, and supervision every two weeks [source: Wang et al. 2011]. Clinical use within individual CBT: AR components spread over 6–20 sessions [synthesis].

3. Guiding principles

[source for 3.1–3.6: Försterling 1985, 1986, 1988; Weiner 1985; for 3.7: Janoff-Bulman 1979; others marked] [uncertain]

Target realistic attributions, not just positive ones. Before changing a client's explanation, check the evidence; you are correcting errors, not handing out comforting stories.
Functional causes for failure are unstable and controllable. For failures, steer towards effort, strategy, preparation or circumstances the client can change; away from fixed ability, character or fate.
Effort attributions only fit failures where effort really was low. Where effort was already high, or more effort will not close the gap, use strategy attributions ("a different method") so the client is not set up for more failure and self-blame. Use external causes (bad luck, a hard task) sparingly. [source: Dresel, Reattributionstraining] [uncertain: Försterling 1985; Anderson & Jennings 1980; Clifford 1984]
Credit successes internally. Help the client own the ability, skill and effort behind successes, so that success raises expectancy and self-concept rather than being written off as luck. Give ability feedback after success on tasks of medium difficulty; after very easy tasks it can backfire. [source: Dresel, Reattributionstraining] [uncertain: Layden 1982; Bandura 1977; Schunk 1983]
Use the person's own data. Kelley's covariation logic (consensus, consistency, distinctiveness) lets the client test explanations like a scientist rather than being argued out of them. [source: Kelley 1967; Försterling 1986, 1988]
Time the message. AR works best soon after a real, meaningful setback, while the person is actively searching for causes. [source: Perry & Penner 1990; Haynes et al. 2009]
Separate behaviour from character. If self-blame is warranted, aim it at specific, changeable actions (behavioural self-blame) rather than at the self (characterological self-blame), and never shift blame for genuinely uncontrollable harm (abuse, assault, illness) onto the client.
Attributions must be tested in action. A new explanation becomes believable when acting on it changes an outcome; pair talk with graded tasks. [source: Dweck 1975; Bandura 1977] [synthesis: emphasis]
Elaborate, don't lecture. The client should restate, apply and argue for the new explanation; passive information alone is weaker. [source: Hall et al. 2007] [uncertain]
No deception. Early misattribution studies misled participants about the cause of their arousal; clinical practice should use honest information only. [synthesis]
Respect culture and context. Fatalistic, religious or collectivist causal beliefs are not automatically errors; work with the client's framework and look for controllable elements within it. [synthesis]

4. Intervention inventory

The items are grouped by phase. Items 4.1–4.12a are core AR. Items 4.13–4.16 are reattribution techniques as used in cognitive therapy, optimism training and somatic work. Item 4.17 is historical. All example language is the writer's own illustrative wording. Status is marked per item.

4.1 Attributional assessment interview

Purpose / when: First sessions, and whenever a new setback arrives. Finds the explanations behind hopelessness, giving up, shame or avoidance. [source: Försterling 1988; Weiner 1985] [uncertain]
Steps:
1. Pick a specific, recent, emotionally live outcome (a failure, a rejection, a symptom flare, a success the client dismissed).
2. Ask open questions about why it happened and record the client's exact words. Ask for every cause they can think of, then which one matters most.
3. For the main cause, ask four rating questions (0–10 or yes/no): was it about you or something outside you? Will it still be there next time? Could you influence it? Does it affect this area only, or everything?
4. Ask what they felt and did next (gave up, avoided, tried harder, hid it). This makes the link between attribution, emotion and behaviour visible.
5. Check realism: what is the evidence for the cause; what do others in the same position experience; what happened on other occasions?
6. Write the chain down together, for example: event → "I'm hopeless at this" (internal, stable, uncontrollable, global) → shame → stopped trying.
Example language (illustrative): "When the result came back, what went through your mind about why it happened?" / "If you had to bet, will that cause still be there the next time you try?" / "How much could you have influenced it, 0 to 10?"
Watch for: Clients who give only socially acceptable causes; probe gently for the "real" explanation ("and the bit you'd not say out loud?"). It is working when the client starts naming causes and dimensions on their own and sees how a cause leads to a feeling. [synthesis]

4.2 Standardised measurement of attributional/explanatory style

Purpose / when: Baseline and outcome; useful in groups and research-informed practice. The ARGT pilot gave the Attributional Style Questionnaire (internality, stability, globality and a derived hopelessness score) before and after the 8-week group, alongside self-esteem, well-being and social-functioning measures. [source: Wang et al. 2011] [uncertain: Peterson et al. 1982; Peterson & Seligman 1984]
Steps:
1. Choose a licensed measure that fits the population: the Attributional Style Questionnaire (ASQ) for adults (hypothetical good and bad events, the respondent names the main cause and rates it on internality, stability, globality); the Children's Attributional Style Questionnaire (CASQ) for children; academic-specific versions for students. Do not improvise items. Obtain the instrument from its authors or publisher.
2. Alternatively, use CAVE (Content Analysis of Verbatim Explanations): extract causal statements from session transcripts, diaries or letters and rate them on the same dimensions.
3. Feed back the profile in plain terms ("for bad events you tend to pick causes that are about you, permanent and everywhere").
Example language (illustrative): "This isn't a test you pass or fail; it shows the kinds of explanations your mind reaches for first."
Watch for: Composite scores hide domain differences; pair them with the interview. A shift in the bad-event composite between baseline and end is one marker of change. [synthesis]

4.3 Attribution diary (self-monitoring of successes and failures)

Purpose / when: From Phase 1 onwards. Collects real-time data and builds the habit of noticing explanations. In Layden's attributional style therapy, clients recorded successes as well as failures so that their own part in positive outcomes became visible. [source: Layden 1982] [uncertain: exact format not verified]
Steps:
1. Give a simple sheet with columns: date, event, success or setback, my explanation, rating on me/not-me, lasting/temporary, controllable/not, everything/just this, feeling (0–100), what I did next.
2. Ask for at least one success and one setback per day. Low-mood clients will "find no successes", so define success broadly (completed tasks, small kindnesses, coping).
3. Review in session: look for patterns (successes credited to luck or others; failures credited to the self in stable, global terms).
4. Later, add a column for an alternative, realistic explanation.
Example language (illustrative): "Write down the small wins too: getting out of bed when you didn't want to counts. Then the key column: why did it go that way?"
Watch for: Diaries that become self-criticism logs; keep reviews curious and non-judgemental. It is working when the success column fills and the client begins writing alternative causes without being asked. [synthesis]

4.4 Psychoeducation on the attribution model

Purpose / when: Start of Phase 2. Gives a rationale: explanations shape feelings and persistence. In Perry's protocol this is the induction, a short handout or video describing how changing explanations for poor performance (from ability to effort and strategy) helps students improve. [source: Perry & Penner 1990; Haynes et al. 2009; Weiner 1985] [uncertain]
Steps:
1. Draw the three dimensions (me/not me, lasting/temporary, controllable/not) and add everything vs just this.
2. Use a neutral example (two students fail the same test; one blames being "not a maths person", the other blames revision method) and ask the client to predict each student's feelings and next steps.
3. Point out that the explanation, not the event, predicted giving up versus trying again.
4. Add the realism caveat: we keep explanations that fit the facts.
5. Link to the client's own chain from 4.1.
6. For groups or classes, a short video or written account of a credible peer describing this shift can deliver the same content (see 4.7).
Example language (illustrative): "Same exam, same mark. One student decides 'I'm just not clever enough' and drops the course; the other decides 'I revised the wrong way' and changes method. What made the difference?"
Watch for: Hearing it as "it's all in your head" or "just try harder". Clarify that we're looking for the accurate and usable cause. It is working when the client applies the model to their own example. [synthesis]

4.5 Covariation analysis (testing explanations with consensus, consistency and distinctiveness)

Purpose / when: Phases 1–2, when an attribution may be inaccurate. Försterling proposed using Kelley's covariation principle in therapy: the client checks whether the cause they chose fits the pattern of evidence. [source: Kelley 1967; Försterling 1986, 1988] [uncertain]
Steps:
1. State the client's attribution as a hypothesis ("I was rejected because I'm unlikeable").
2. Consensus: do other people get the same outcome in the same situation? (High consensus points to the situation.)
3. Consistency: does this outcome happen to you every time in this situation, or only sometimes? (Low consistency points to unstable causes.)
4. Distinctiveness: does it happen to you across many situations, or only this one? (High distinctiveness points away from a global personal cause.)
5. Gather the data: memories, others' reports, small surveys, behavioural records.
6. Draw a conclusion that fits all three; rate belief in the old and new explanations.
Example language (illustrative): "If 'unlikeable' were the cause, what would we expect: rejected by everyone, everywhere, every time? Let's check what actually happens."
Watch for: Data that confirms an internal stable cause (e.g. a genuine skill deficit). Then accept it and move to what is controllable (skill training, strategy, environment). It is working when belief in the new explanation rises because of evidence, not deference to the therapist. [synthesis]

4.6 Consensus information and normalising (the Wilson & Linville procedure)

Purpose / when: Early setbacks in a transition (first year at university, new job, new diagnosis, early therapy). People who think they are uniquely failing infer stable personal causes. Wilson and Linville gave worried first-year students statistics and short interviews with older students showing that grades typically improve after the first year, encouraging an unstable explanation. [source: Wilson & Linville 1982, 1985] [uncertain: details not verified]
Steps:
1. Identify the transition and the client's belief that their difficulty is unusual and permanent.
2. Provide accurate information on how common the difficulty is and how it usually changes with time and practice (data, testimony from credible peers, group members' experiences).
3. Ask the client to say what this implies about the cause of their own difficulty.
4. Agree concrete actions consistent with the temporary explanation.
Example language (illustrative): "Most people starting this job find the first months rough, and most find it gets easier. What does that suggest about why it's hard for you right now?"
Watch for: Only use true information; if the client's difficulty is genuinely atypical, say so and switch to 4.5/4.9. Working when hopelessness and avoidance fall and engagement continues. [synthesis]

4.7 Attributional modelling

Purpose / when: Phase 2, especially in groups, classrooms and with clients who distrust a therapist's "spin". A credible model (peer, former client, video) describes a setback, their first stable or uncontrollable explanation, the switch to an effort or strategy explanation, and the improvement that followed. Modelling (a credible model voicing favourable attributions, based on social learning theory) is one of three standard reattribution techniques, with commentary/feedback and reinforcement (4.8) [source: Dresel, Reattributionstraining]. Verified example: in Parker et al. 2022 (older adults at a day hospital) an 18-minute video shows one older man explaining his inactivity by age and poor health, and a peer coaching him towards controllable causes (poor strategy, low effort) and changes of approach such as going slower, using a walker or taking breaks; by the end he voices the new explanations himself [source: Parker et al. 2022]. [uncertain: Försterling 1985, Perry & Penner 1990 and Schunk 1983]
Steps:
1. Pick or prepare a model similar to the client (age, situation, starting difficulty); coping models (who struggle first) generally beat mastery models. [source: Schunk & Hanson 1985] [uncertain]
2. The model narrates out loud: the setback → the initial helpless explanation → a different, controllable explanation → a specific change in action → the result.
3. Debrief: what explanation did the model start with and switch to; what did they do differently; where does this apply to you?
4. In live work, the therapist can model by thinking aloud about their own minor setbacks (keep self-disclosure brief and relevant).
Example language (illustrative, therapist self-modelling): "When my first draft got torn apart, my first thought was 'I can't write'. Then I noticed I'd skipped the outline. Next draft I outlined first, and it went much better."
Watch for: Models who are too polished; "they're not like me" reactions. It is working when the client identifies with the model and names their own equivalent switch. [synthesis]

4.8 Operant reinforcement of effort and strategy attributions (in-task attributional feedback)

Purpose / when: During graded tasks (academic work, rehabilitation exercises, social or activity tasks). Two linked techniques from school reattribution training: attributional commentary (spoken or written feedback on a result that names a favourable cause, such as effort or strategy) and operant reinforcement (praise the favourable explanations the client voices spontaneously; question unfavourable ones mildly, never punish). Commentary can also carry consistency and consensus information (compared with earlier results and with peers), which supports variable or external explanations for failure [source: Dresel, Reattributionstraining]. Classic origins: Dweck's 1975 study of helpless children, in which some arranged failures were attributed to insufficient effort, and later verbal reinforcement of children's own effort statements (Andrews & Debus 1978) [uncertain: Dweck 1975; Andrews & Debus 1978; Försterling 1985].
Steps:
1. Set up a graded task where effort or strategy really affects success.
2. Include occasional manageable failures; success-only programmes do not teach the client how to handle failure.
3. After a failure, state, or prompt the client to state, an unstable, controllable cause (effort, method), then retry with that change.
4. When the client spontaneously gives an effort or strategy explanation, reinforce it immediately (acknowledge, agree, link to the next try).
5. After a success, link it to the client's effort and, as skill builds, to their growing ability (see 4.10).
6. Fade your prompts so the client produces the attributions on their own.
Example language (illustrative): "That one didn't work. What could you change on the next try?" ... "Right, you slowed down and checked each step, and it came out."
Watch for: Effort feedback when effort was already high (it implies low ability; switch to strategy feedback); non-specific praise; switching feedback type too often. A tested sequence: effort feedback early in a new topic, ability feedback once skill has visibly grown, with few switches [source: Dresel, Reattributionstraining] [uncertain: Schunk 1983]. It is working when persistence after failure increases and self-generated controllable attributions appear. [synthesis]

4.9 Strategy attribution and strategy training

Purpose / when: When effort alone will not close the gap, or the client is already working very hard. Attributing failure to the method used is unstable, controllable and doesn't imply lazy or stupid. It must be paired with real strategy teaching. [source: Anderson & Jennings 1980; Clifford 1984; Försterling 1985] [uncertain]
Steps:
1. Review the failed attempt in detail: what exactly did the client do, step by step?
2. Identify at least one alternative strategy (preparation, pacing, asking for help, breaking the task down, environmental change).
3. Reframe the failure as information about the strategy, not the person.
4. Plan a test of the new strategy; review results with an attribution question ("what made the difference?").
5. Build a personal "strategy menu" for future setbacks.
Example language (illustrative): "You worked hard; effort wasn't the problem. Let's look at how you went about it and find one thing to do differently."
Watch for: Strategy talk without real skill gain leads to repeated failure and a slide back to ability attributions. Make sure the strategy is learnable and likely to help. Working when the client generates alternative strategies after setbacks. [synthesis]

4.10 Crediting success (internal attributions for positive outcomes)

Purpose / when: Depression, low self-esteem, impostor beliefs: successes dismissed as luck, others' help or task ease. Success credited to ability and effort builds expectancy and self-concept [source: Dresel, Reattributionstraining]; session 7 of ARGT is given to self-esteem and attributions for positive events [source: Wang et al. 2011]. Layden's attributional style therapy targeted this side of the depressive style [uncertain: Layden 1982; Bandura 1977; Abramson et al. 1978].
Steps:
1. From the diary, choose a success the client wrote off.
2. List every contributing cause together, including outside ones (to keep it realistic).
3. Ask, for each personal contribution: what skill, quality or choice did that take? Would it be there next time? Does it show up elsewhere?
4. Estimate the client's share of the outcome realistically; contrast with the initial "luck" estimate.
5. Write a one-line ownership statement and ask the client to repeat this process for successes as homework.
Example language (illustrative): "Luck may have played a part. What did you bring to it? What would have happened if you hadn't prepared?"
Watch for: Grandiosity is rarely the risk; discomfort and rejection of praise are common. Keep it evidence-based rather than complimentary. It is working when the client credits themselves without prompting and success raises expectations for the next task. [synthesis]

4.11 Consolidation by elaborative writing

Purpose / when: Immediately after induction (4.4/4.7), to make the message personal and memorable. Hall, Perry and colleagues found AR with an elaborative writing task outperformed AR alone for some students. [source: Hall et al. 2004, 2007] [uncertain: task wording not verified; the steps below are a paraphrased reconstruction]
Steps:
1. Ask the client to write, in their own words, the main idea of what they just heard or watched (how explanations affect performance and feelings).
2. Ask them to describe a past setback of their own, how they explained it at the time, and how they would explain it now using controllable causes.
3. Ask how they will apply this to an upcoming challenge (specific plan).
4. Review the writing together, or with a peer in group formats.
Example language (illustrative): "Take ten minutes. Write how you explained the last exam that went badly, then how you'd explain it now, and what you'll do differently before the next one."
Watch for: Rote copying of the handout. Ask for personal detail. It is working when the writing contains specific, controllable causes linked to concrete plans. [synthesis]

4.12 Consolidation by group discussion

Purpose / when: Group, classroom and workshop AR. Peers rehearse the new explanations aloud and supply consensus information to each other. In Perry's Manitoba programme, discussion of the causes of success and failure has been tested as a follow-up to the standard AR video (Hunter 1997 compared video alone with video plus discussion, an aptitude test or a lecture) [source: Hunter 1997 thesis record]. For a full clinical group programme see 4.12a. [uncertain: Perry & Penner 1990; Haynes et al. 2009]
Steps:
1. After induction, set small groups (3–6) a structured prompt: each member shares a recent setback and their first explanation.
2. The group helps generate alternative controllable and realistic explanations and one action step for each member.
3. Groups report back; the facilitator highlights common patterns (normalising) and corrects any slide into blame.
4. Close with each member stating their own plan.
Example language (illustrative): "Notice how many of you said 'I'm just not good at this'. Let's hear one alternative explanation for each story that you could actually do something about."
Watch for: Dominant members; collusion in hopelessness; peers offering unrealistic reassurance. It is working when members challenge each other's stable explanations with evidence. [synthesis]

4.12a Attribution retraining group therapy (ARGT): eight-session clinical group

Purpose / when: Outpatient groups with depression, generalised anxiety or OCD where a negative attributional style and low self-esteem feed the symptoms. The rationale is the hopelessness/self-esteem account of depression (attributional diathesis plus low self-esteem plus failure events). Groups are mixed by diagnosis. [source: Wang et al. 2011]
Steps: session themes as reported in the pilot (one 2-hour session a week; the full manual is a Chinese-language book, Wang & Zhang 2008) [source: Wang et al. 2011]:
1. Introductions and building mutual support.
2. What symptoms mean and how thinking affects them.
3. The role of attributions (teach the dimensions, as in 4.4).
4. Members' upbringing and basic beliefs, and how these shaped their habitual explanations.
5. Rebuilding attributional style: identify automatic thoughts rooted in unhealthy attributions and offer healthier alternative explanations; practise new behaviour.
6. Consolidating the new attributions and behaviours.
7. Self-esteem, personality and attributions for positive events (see 4.10).
8. Future plans and ending the group.
- Staffing: a lead therapist plus a co-therapist who records whether each planned step happened and reviews this with the lead before each session; regular supervision. [source: Wang et al. 2011]
- Within-session methods (4.1, 4.3, 4.5, 4.12) fit these themes but are not described in the paper. [synthesis]
Example language (illustrative): "Take the thought 'I'll never get better'. What cause is it pointing to? Is there an explanation that fits the facts better and leaves you something to do?"
Watch for: Mixed-diagnosis groups need examples from each condition. The pilot measured attributional style (ASQ), self-esteem, well-being and social functioning before and after; similar markers show whether it is working [source: Wang et al. 2011]. Homework and detailed technique were not reported. [uncertain]

4.13 Reattribution of responsibility: reviewing the facts (cognitive therapy)

Purpose / when: Excessive self-blame for negative events in depression, guilt and grief. Beck and colleagues described "reattribution" as going back over the events with the patient and applying logic to share out responsibility fairly. The aim is not to absolve the patient of all responsibility but to remove unfair self-blame. [source: Beck, Rush, Shaw & Emery 1979, ch. on cognitive techniques] [uncertain: paraphrase, page not verified]
Steps:
1. Get a detailed, time-ordered account of the event.
2. Identify what the client knew at the time (not hindsight), what options they actually had and what other factors and people contributed.
3. Examine the standard the client applies to themselves; ask whether they would apply it to a friend (double standard).
4. Reach a revised, proportionate statement of responsibility; identify any genuine, specific, behavioural part they can learn from or repair.
Example language (illustrative): "Let's slow it down. At that moment, what did you actually know? What else was going on that you didn't control?"
Watch for: Hindsight bias; the client hearing "it wasn't your fault" as dismissive of real mistakes. Working when guilt intensity drops and any real responsibility turns into repair rather than rumination. [synthesis]

4.14 Responsibility pie chart

Purpose / when: Inflated responsibility (OCD, guilt after accidents, parents blaming themselves). A visual method for sharing out causes; widely used in CBT for OCD (Salkovskis' responsibility model) and in general CBT texts. [source: Salkovskis 1985; van Oppen & Arntz 1994; Greenberger & Padesky 1995] [uncertain]
Steps:
1. Ask the client to rate their own responsibility for the outcome (0–100%). Write it down but do not draw it yet.
2. List every other possible contributing factor and person, as exhaustively as possible.
3. Draw a circle and allocate slices to each other factor first.
4. Give the client's own slice last, with what is left.
5. Compare with the initial rating; discuss what changed.
Example language (illustrative): "Before you give yourself a slice, let's give everything else its share: the weather, the other driver, the timing..."
Watch for: The client inflating each other factor to "prove" a point, or dismissing the exercise as a trick. Use real reasoning for each slice. Working when the re-rating falls and the client can hold a proportionate view outside the session. [synthesis]

4.15 Explanatory-style disputation (optimism training lineage)

Purpose / when: Pessimistic explanatory style (internal, stable, global for bad events) in depression prevention and resilience work. Seligman's learned optimism model uses an ABC record (Adversity, Beliefs, Consequences) followed by Disputation and Energisation; disputation checks evidence, generates alternatives, decatastrophises implications and weighs usefulness. School resilience programmes (Penn Resiliency Program) teach the same dimensions to young people in child-friendly terms. [source: Seligman 1990; Gillham, Reivich, Jaycox & Seligman 1995] [uncertain]
Steps:
1. Record an adversity, the in-the-moment explanation and the emotional or behavioural consequence.
2. Classify the explanation on personal/impersonal, permanent/temporary, pervasive/specific.
3. Dispute: evidence for and against; alternative causes (especially changeable, specific ones); if the explanation is true, what are the realistic implications; is it useful to dwell on it?
4. Note the shift in feeling and energy; plan an action.
5. Practise daily for several weeks; in groups, use role-play and age-appropriate examples.
Example language (illustrative): "Is that 'always', or 'this time'? Is it 'everything', or 'this one thing'? Is it all you, or are there other pieces?"
Watch for: Mechanical positive thinking that ignores evidence; return to realism (principle 1). Working when disputation becomes quick and automatic and mood recovers faster after adversity. [synthesis]

4.16 Reattribution of physical symptoms (Goldberg reattribution model)

Purpose / when: Primary care and liaison settings for patients attributing distress-related bodily symptoms only to physical disease. Goldberg, Gask and O'Dowd's original model (1989) is usually described in three stages; the version used in the Morriss et al. 2007 trial, followed here, has four: feeling understood, broadening the agenda, making the link, negotiating further treatment. Its aim is a negotiated explanation that links the symptom, a psychosocial problem and a physiological or time-based mechanism. "Reattribution" here means broadening the explanation from disease alone, not denying symptoms. [source: Morriss et al. 2007] [uncertain: Goldberg et al. 1989]
Steps:
1. Feeling understood: take a full history of the physical symptoms; explore psychosocial problems, mood and the patient's own beliefs and fears about the cause; carry out a relevant examination and any needed investigations. [source: Morriss et al. 2007]
2. Broadening the agenda ("changing the agenda" in older accounts): summarise the physical and psychosocial findings and negotiate them with the patient; acknowledge that the symptoms are real. [source: Morriss et al. 2007]
3. Making the link: explain how the physical symptom connects with life problems or stress, through a time link (symptoms worsened when the problem began) or a physiological one (e.g. anxiety → muscle tension → pain; poor sleep → fatigue) [source: Morriss et al. 2007]; demonstrate where possible (e.g. sustained muscle tension producing discomfort) [uncertain]. Negotiating further treatment: agree follow-up or treatment for the symptoms, the psychosocial problems or any mental disorder [source: Morriss et al. 2007]. In that trial GPs learned the model in 6 hours (three 2-hour practice sessions) with video material, role-play and video feedback.
Example language (illustrative): "The pain is real. I've also noticed it got much worse around the time things at home became difficult. Shall we look at how stress and tension can feed into pain like this?"
Watch for: Linking too early, before the patient feels heard, triggers "you think it's all in my head". Don't skip appropriate medical assessment. Working when the patient accepts a both/and explanation and engages with psychological management. [synthesis]

4.17 Misattribution procedures (historical; do not use as described)

Purpose / when: Historical only. Early attribution therapy (Ross, Rodin & Zimbardo 1969; Storms & Nisbett 1970; Valins & Nisbett 1971) tried to reduce anxiety or insomnia by leading people to attribute their arousal to a neutral outside source (e.g. a placebo pill or noise) instead of to their own fear. Results replicated poorly and the method relied on deception. [source: Valins & Nisbett 1971; Försterling 1988 review] [uncertain]
Steps (for understanding, not delivery): a neutral external cause for arousal was presented; the person then reinterpreted their arousal as not caused by fear or personal weakness.
Legitimate modern equivalent [synthesis]: honest psychoeducation that arousal has ordinary, benign causes (caffeine, exertion, adrenaline, sleep loss) combined with behavioural testing, as in CBT for panic and health anxiety. No deception, no inert pills presented as active.
Watch for: Any "placebo explanation" approach that misleads clients breaches informed consent. [synthesis]

5. Handling problems

The primary AR literature is mainly experimental and educational and says little about clinical difficulties. Except where a source is given, the guidance below is [synthesis], drawn from Försterling's realism principle and general CBT practice.

Self-blame produced by AR (the main known risk). Sign: after a shift to effort or controllable explanations, the client concludes "so it is my fault", with more guilt or shame. Response: separate responsibility for the past from influence over the future; aim controllable explanations at specific behaviours and strategies (behavioural, not characterological, self-blame) [source: Janoff-Bulman 1979] [uncertain]; use 4.13/4.14 to keep the share realistic; add self-compassionate framing. Never use AR to suggest a client could have prevented abuse, assault, discrimination or illness.
The cause really is uncontrollable or stable (progressive illness, disability, structural disadvantage, real ability limits). Response: don't argue it away (Försterling's realism principle) [source: Försterling 1985, 1988] [uncertain]. Validate it, then look for other causes in the same outcome that are controllable (strategy, support, environment, goal choice). Change the goal if needed, and turn to acceptance-based work for the part that can't change.
Stuck: client agrees intellectually but doesn't believe it ("head vs heart"). Response: stop verbal persuasion; set behavioural tests where acting on the new explanation can change the outcome (4.8/4.9); collect diary evidence; use covariation data (4.5); rate belief before and after each test [synthesis]. If motivation stays low because the client's goal is to avoid looking incompetent, reattribution alone is not enough: add work on goals and on a growth view of ability [source: Dresel, Reattributionstraining].
Resistance: "you're telling me to just try harder". Response: acknowledge the history of such advice; shift from effort to strategy and circumstances (4.9); show that the aim is accuracy. Ask the client to help design the test.
No successes to work with (severe depression). Response: make tasks much smaller so that some success is guaranteed (activity scheduling); start with crediting the effort of trying. Delay failure-attribution work until some activation has returned. Assess risk as standard.
Overwhelm or flooding of shame when reviewing failures. Response: choose less loaded events first; slow down; ground; switch to a success or a third-person example (4.4), then return gradually. For trauma-related self-blame, use a trauma-focused protocol (e.g. cognitive processing therapy or trauma-focused CBT) rather than ad hoc AR.
Non-adherence with diaries or homework. Response: ask why, attributionally ("what did you tell yourself when you didn't do it?"); this is often the same helpless explanation and makes good material. Simplify to one entry a day, use phone notes or session-start recall; praise partial completion with a strategy, not character, framing.
Ruptures: client feels blamed or patronised. Response: name it, take your share ("I may have pushed the 'what could you change' question too soon"), return to validation before any reattribution. Rupture repair follows general alliance practice (e.g. Safran & Muran) [uncertain].
Over-correction: externalising everything. Sign: the client starts blaming others for all failures, with anger and passivity. Response: return to covariation (4.5); balance aims at realistic internal-controllable explanations, not external ones.
Group-specific: peers reinforce hopelessness or offer false reassurance. Response: structure turns; ask the group for evidence-based alternatives; ensure the facilitator presents consensus data.
Cultural or religious causal frameworks (fate, God's will, karma, family honour). Response: explore respectfully; don't treat these as distortions; look for controllable actions consistent with the client's beliefs (e.g. "what part is given to you to do?"); consult a cultural adviser if needed.
Children and adolescents. Response: use concrete examples, games and coping models; involve teachers and parents in effort- and strategy-based feedback (4.8); avoid fixed-ability praise. [source: Dweck 1975; Gillham et al. 1995] [uncertain]
Medically unexplained symptoms: patient hears "it's all in your head". Response: go back to stage 1 of 4.16 (feeling understood); use both/and language; keep appropriate medical follow-up. [source for the stage model: Morriss et al. 2007; response is synthesis]
Risk. If attributional work uncovers hopelessness ("nothing will ever change") with suicidal thoughts, carry out a standard risk assessment and safety plan, including lethal-means counselling (safe storage and removal), before going on with AR. Hopelessness is a recognised risk marker; treat its appearance as a cue to assess, not just to reattribute.

Sources

Sources: the first seven items below (the last three as catalogue records with abstracts only). The remaining items are not drawn on.

Books and papers cited (drawn on: Wang et al. 2011; Parker et al. 2022; Morriss et al. 2007; Dresel. All others from the writer's knowledge, not drawn on):

Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. D. (1978). Learned helplessness in humans: Critique and reformulation. Journal of Abnormal Psychology.
Anderson, C. A., & Jennings, D. L. (1980). When experiences of failure promote expectations of success: The impact of attributing failure to ineffective strategies. Journal of Personality.
Andrews, G. R., & Debus, R. L. (1978). Persistence and the causal perception of failure: Modifying cognitive attributions. Journal of Educational Psychology.
Antaki, C., & Brewin, C. (Eds.) (1982). Attributions and Psychological Change: Applications of Attributional Theories to Clinical and Educational Practice. Academic Press. (Includes Layden, M. A., attributional style therapy.)
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review.
Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive Therapy of Depression. Guilford.
Clifford, M. M. (1984). Thoughts on a theory of constructive failure. Educational Psychologist.
Dweck, C. S. (1975). The role of expectations and attributions in the alleviation of learned helplessness. Journal of Personality and Social Psychology.
Försterling, F. (1985). Attributional retraining: A review. Psychological Bulletin.
Försterling, F. (1986). Attributional conceptions in clinical psychology. American Psychologist.
Försterling, F. (1988). Attribution Theory in Clinical Psychology. Wiley.
Gillham, J. E., Reivich, K. J., Jaycox, L. H., & Seligman, M. E. P. (1995). Prevention of depressive symptoms in schoolchildren: Two-year follow-up. Psychological Science.
Goldberg, D., Gask, L., & O'Dowd, T. (1989). The treatment of somatization: Teaching techniques of reattribution. Journal of Psychosomatic Research.
Greenberger, D., & Padesky, C. A. (1995). Mind Over Mood. Guilford.
Hall, N. C., Perry, R. P., et al. (2004; 2007). Attributional retraining and elaborative learning studies. Journal of Social and Clinical Psychology; Journal of Applied Social Psychology. [uncertain: exact titles not verified]
Haynes, T. L., Perry, R. P., Stupnisky, R. H., & Daniels, L. M. (2009). A review of attributional retraining treatments: Fostering engagement and persistence in vulnerable college students. In Higher Education: Handbook of Theory and Research. Springer.
Janoff-Bulman, R. (1979). Characterological versus behavioral self-blame. Journal of Personality and Social Psychology.
Kelley, H. H. (1967). Attribution theory in social psychology. Nebraska Symposium on Motivation.
Perry, R. P., & Penner, K. S. (1990). Enhancing academic achievement in college students through attributional retraining and instruction. Journal of Educational Psychology.
Peterson, C., Semmel, A., von Baeyer, C., Abramson, L. Y., Metalsky, G. I., & Seligman, M. E. P. (1982). The Attributional Style Questionnaire. Cognitive Therapy and Research.
Peterson, C., & Seligman, M. E. P. (1984). Causal explanations as a risk factor for depression. Psychological Review.
Ross, L., Rodin, J., & Zimbardo, P. G. (1969). Toward an attribution therapy. Journal of Personality and Social Psychology.
Salkovskis, P. M. (1985). Obsessional-compulsive problems: A cognitive-behavioural analysis. Behaviour Research and Therapy.
Schunk, D. H. (1983). Ability versus effort attributional feedback. Journal of Educational Psychology.
Schunk, D. H., & Hanson, A. R. (1985). Peer models: Influence on children's self-efficacy and achievement. Journal of Educational Psychology.
Seligman, M. E. P. (1990). Learned Optimism. Knopf.
Storms, M. D., & Nisbett, R. E. (1970). Insomnia and the attribution process. Journal of Personality and Social Psychology.
Valins, S., & Nisbett, R. E. (1971). Attribution Processes in the Development and Treatment of Emotional Disorders. General Learning Press.
van Oppen, P., & Arntz, A. (1994). Cognitive therapy for obsessive-compulsive disorder. Behaviour Research and Therapy.
Weiner, B. (1985). An attributional theory of achievement motivation and emotion. Psychological Review.
Weiner, B. (1986). An Attributional Theory of Motivation and Emotion. Springer.
Wilson, T. D., & Linville, P. W. (1982; 1985). Improving the academic performance of college freshmen: Attribution therapy revisited / Improving the performance of college freshmen with attributional techniques. Journal of Personality and Social Psychology.

Added in the verification pass:

Dresel, M. (n.d.). Reattributionstraining REA. In Handwörterbuch Schulleitung (40th supplement). Cites Försterling & Stiensmeier-Pelster (1994) and Ziegler & Schober (2001).
Morriss, R., Dowrick, C., Salmon, P., et al. (2007). Cluster randomised controlled trial of training practices in reattribution for medically unexplained symptoms. British Journal of Psychiatry, 191, 536–542.
Parker, P. C., Chipperfield, J. G., Hamm, J. M., Perry, R. P., Krylova, A., Chuchmach, L., & Hladkyj, S. (2022). Attributional retraining: Promoting psychological wellbeing in older adults with compromised health. Frontiers in Sports and Active Living, 4, 949501.
Perry, R. P., Chipperfield, J. G., Hladkyj, S., Pekrun, R., & Hamm, J. M. (2014). Attribution-based treatment interventions in some achievement settings. Advances in Motivation and Achievement, 18, 1–35.
Wang, C., Zhang, J., Li, J., Zhang, N., & Zhang, Y.-L. (2011). Attribution retraining group therapy for outpatients with major depression disorder, generalized anxiety disorder, and obsessive-compulsive disorder: a pilot study. Journal of Biomedical Research, 25(4). doi:10.1016/S1674-8301(11)60046-8
Wang, C., & Zhang, Y.-L. (2008). Get back the peace of mind [ARGT case-analysis book used as the treatment manual; Chinese; cited in Wang et al. 2011].

← Previous · II-078Cognitive restructuring and Socratic questioning Next · II-080 →Stress inoculation training and self-instructional training