Cognitive restructuring and Socratic questioning

1. Introduction

Version followed: Beckian cognitive therapy as set out in Beck, Rush, Shaw & Emery (1979) and Judith S. Beck, Cognitive Behavior Therapy: Basics and Beyond (3rd ed., 2020), with Socratic questioning structured by Padesky's four-stage model in its 1993 form and her 2019 update, which renames it "Socratic Dialogue" (expanded in Padesky & Kennerley, Dialogues for Discovery, 2023), core-belief methods as in Padesky (1994), and thought records in the Greenberger & Padesky Mind Over Mood (2nd ed., 2016) format. Ellis's REBT disputing is described as a named alternative variant where it differs.

Source coverage (read first). Sources: Padesky's four stages of Socratic questioning and the changing-minds vs guiding-discovery distinction (her 1993 paper and 2019 update); the core-belief methods in 4C (prejudice metaphor, building the new belief, positive data log, continua, historical test, early-scene role-play, core belief worksheet; Padesky 1994); the order of skills in Mind Over Mood (publisher contents page); the J. Beck question set (through a secondary summary of the book and a Beck Institute training deck, not the book itself); seven-column thought records and the Socratic vs didactic balance (Wills 2008 extract); REBT's ABC model, the three disputing types, rational-belief components and healthy vs unhealthy emotions (Wikipedia and Turner 2016). Left [uncertain]: session counts from Beck et al. (1979) and J. Beck (2020), the five-column layout of Beck's 1979 record, the name "intellectual–emotional role-play", the origin of rational-emotive imagery, REBT disputing styles, the responsibility-pie attribution, and the Salkovskis, Safran & Muran, CPT, CBTp and safety-planning points. No manual wording is reproduced beyond short cited phrases; all example language is illustrative.

What it is. Cognitive restructuring is not a stand-alone therapy. It is the core component procedure of cognitive therapy and most CBT protocols: the client learns to catch the thoughts that come up in specific upsetting situations, treat them as hypotheses rather than facts, test them against evidence and experience, and build more accurate, more useful alternatives. A common summary gives four steps: identify the automatic thoughts, identify the distortions in them, dispute them with Socratic questioning, and develop a rational response [source: Wikipedia: Cognitive restructuring]. Socratic questioning (Socratic dialogue, guided discovery) is the main way the therapist conducts that testing. The therapist asks questions the client can answer that bring relevant information they are not attending to into focus, rather than persuading them [source: Padesky 1993; Beck et al. 1979; J. Beck 2020].

How change happens. In the cognitive model, emotional and behavioural reactions follow from how a situation is interpreted. Interpretations come from deeper beliefs, which run from automatic thoughts to intermediate beliefs (rules, attitudes, assumptions) to core beliefs about self, others and the world. Distress persists because biased appraisals are accepted unexamined and are kept going by avoidance and selective attention. Change comes when appraisals are examined and revised, and when the revised beliefs are tested in action so they become believed emotionally as well as intellectually [source: Beck 1976; Beck et al. 1979; J. Beck 2020]. Ellis's ABC model locates disturbance in rigid, absolutistic demands ("musts") rather than in distorted inferences, so REBT disputes the demand itself [source: Ellis 1962; Ellis & Dryden 1997].

2. Framework

Restructuring is delivered inside a structured CBT course: individual sessions of about 45–50 minutes with mood check, agenda, homework review, work on one or two problems, new homework and feedback. Beck et al. (1979) describe roughly 15–20 sessions for unipolar depression, often twice weekly at first. J. Beck (2020) says many straightforward cases need about 6–14 sessions, with more for complex or personality-disorder presentations [uncertain: exact figures vary by edition]. It is also used in groups and in guided self-help such as Mind Over Mood and Burns's Feeling Good [source: Greenberger & Padesky 2016; Burns 1980].

The levels are worked in roughly this order. The phases overlap; the therapist moves on when the criteria are met, not by session number [source: J. Beck 2020; Beck et al. 1979; criteria partly synthesis]. The same bottom-up order is the chapter sequence of Mind Over Mood (2nd ed.): moods → situations, moods and thoughts → automatic thoughts → evidence → alternative or balanced thinking → new thoughts, action plans and acceptance → underlying assumptions and behavioural experiments → new core beliefs, gratitude and acts of kindness → maintaining gains [source: Mind Over Mood publisher contents page].

Phase 0: Conceptualisation and socialisation (sessions 1–2).

Goals: build a cognitive conceptualisation of the presenting problem; teach the cognitive model using the client's own recent example; agree goals; set up the session structure.
Severe depression: Beck et al. (1979) put behavioural methods (activity scheduling, graded tasks) before heavy cognitive work, because concentration and hopelessness limit verbal work [source: Beck et al. 1979].
Move on when the client can describe one of their own situation → thought → emotion → behaviour sequences and says the model fits at least partly.

Phase 1: Identifying automatic thoughts and emotions (sessions 1–4).

Goals: the client catches automatic thoughts and images in specific situations; tells thoughts apart from emotions; rates emotion intensity (0–100) and belief in each thought (0–100%); finds the "hot" thought most linked to the emotion [source: Greenberger & Padesky 2016; J. Beck 2020].
Format: done in session first, then as homework using a three-column record (situation, emotion, thought).
Move on when the client can reliably record situation, emotion and thought separately, with ratings, between sessions.

Phase 2: Evaluating and responding to automatic thoughts (sessions ~3–10).

Goals: Socratic examination of hot thoughts; spotting thinking errors; building balanced or alternative responses; re-rating belief and emotion; turning the remaining predictions into behavioural experiments; action plans when a thought is accurate and acceptance where the situation cannot be changed [source: J. Beck 2020; Beck et al. 1979; Mind Over Mood ch. 10 title "New Thoughts, Action Plans, and Acceptance", contents page].
Format: full thought record (seven columns in the Mind Over Mood version; five in Beck's 1979 dysfunctional thought record), coping cards, experiments.
Move on when the client completes records on their own with some drop in emotion; when recurring themes across situations point to underlying assumptions; and when symptoms have eased enough to allow deeper work.

Phase 3: Intermediate beliefs (rules, attitudes, assumptions; mid-therapy).

Goals: identify the conditional assumptions ("If I…, then…") and rules behind recurring automatic thoughts, using the downward arrow and themes; examine their advantages and disadvantages; test them with experiments; build more flexible alternatives [source: J. Beck 2020; Beck et al. 1979].
Move on when assumptions are named and partly tested, and when the core belief behind them is clear enough to state.

Phase 4: Core beliefs (later therapy; only if needed).

Goals: identify the core belief (J. Beck groups negative self-beliefs as helpless, unlovable or worthless [source: J. Beck 2020, via secondary summary]); construct the new belief the client would like to hold; collect evidence over months with the core belief worksheet, positive data log, continua, historical test and early-scene role-play [source: Padesky 1994; J. Beck 2020]. Imagery restructuring of early memories is used in some protocols but is not described in Padesky 1994 [uncertain].
Format: this work is slow. Padesky (1994) reports new-belief ratings often only around 5–10% after 4–8 weeks of logging and 20–40% over the following month or two, and recommends keeping the log for at least six months [source: Padesky 1994]. Many depression and anxiety clients recover without extensive core-belief work [uncertain].

Phase 5: Consolidation and relapse prevention (last 2–4 sessions, sessions spaced out).

Goals: the client acts as their own therapist; lists the skills that helped; plans for setbacks; writes a coping card or blueprint; booster sessions [source: J. Beck 2020; Beck et al. 1979].
Finish when the client can, without prompting, catch and evaluate a hot thought and plan an experiment.

REBT variant. REBT follows ABCDE: Activating event or Adversity → Beliefs (rigid vs flexible) → Consequences (emotions and behaviour) → Disputing → Effective new belief; some accounts add F, the new feelings and behaviour [source: Wikipedia: Rational emotive behavior therapy; Turner 2016]. It goes to the core demands earlier and more directly than Beck's bottom-up order, and disputing begins in the first sessions [source: Ellis & Dryden 1997; Dryden & Neenan 2004 — uncertain].

3. Guiding principles

[source: Padesky 1993 and 2019 for the collaborative stance, guided discovery and summaries; Wills 2008 extract for persuasion and the didactic balance; Beck et al. 1979; J. Beck 2020; Kazantzis, Dattilio & Dobson 2017. Practical glosses are synthesis.]

Collaborative empiricism. Client and therapist are co-investigators who test the client's beliefs against data. Practically, ask "Shall we look at how true this is?" rather than telling the client they are wrong.
Guided discovery, not changing minds. Padesky (1993) contrasts questioning aimed at changing minds (the therapist already has the answer and leads the client to it step by step, which makes the therapist the author of the conclusion) with questioning aimed at guiding discovery, where the therapist is genuinely curious and the conclusion belongs to the client. In her words, "If you are not truly curious to know the answer, don't ask the question" [source: Padesky 1993]. Practically, ask questions you do not know the answer to, and be willing to find that the thought is accurate.
Thoughts are hypotheses, not facts. Practically, rate belief as a percentage and treat each thought as something to check.
Work on specific situations. Restructuring works on concrete moments (when, where, what went through your mind), not general complaints. Practically, ask "Can you tell me about the most recent time?"
Target the hot thought. Work on the thought or image carrying the most emotion and most relevant to the client's goals, not every thought [source: Greenberger & Padesky 2016]. Practically, ask "Which of these thoughts makes you feel the worst?"
Accuracy and usefulness, not positive thinking. The aim is a balanced, realistic, workable view. Rosy reframes do not last. Practically, the alternative must be one the client believes at least somewhat.
Empathy and validation first. Emotions are understandable given the interpretation. Questioning without empathy feels like cross-examination. Practically, reflect the feeling before asking about evidence.
Some thoughts are true. If the evidence supports the thought, move to problem-solving, acceptance, or the meaning drawn from it ("and if that's true, what does it mean?") [source: J. Beck 2020].
The client reaches the conclusion. The therapist summarises, and the client synthesises and states the conclusion in their own words. Padesky notes that therapists often skip this synthesising step [source: Padesky 1993]. Practically, ask "How does this fit with your original thought?" instead of stating the moral.
Change belief through action. Insight is tested in behaviour, because experiments produce the strongest belief change [source: Bennett-Levy et al. 2004]. Practically, end most restructuring with "How could we test this this week?"
Conceptualisation drives choices. Choose which thoughts and beliefs to target from the case formulation. Practically, link each target back to the client's goals and maintaining cycle.
Ask answerable questions, and balance Socratic and didactic. Socratic questions should be ones the client has the knowledge to answer, should bring relevant but unattended information into focus, and usually move from concrete to abstract and back to concrete action [source: Padesky 1993]. Questioning cannot draw out knowledge the client does not have, so teach directly when the client lacks information (e.g. panic physiology) [synthesis]. Wills (2008) recommends a more educational style when clients cannot yet separate feelings from thoughts or cannot identify any thought, and warns that clients resist therapists they see as too persuasive [source: Wills 2008 extract]. Padesky (1993) gives no explicit criteria for switching to teaching.
Structure, feedback and summaries. Summarise every few minutes during questioning. Padesky (1993) names too few summaries as a common error, and in 2019 strongly prefers written summaries in the client's own words, ideally copied out by the client [source: Padesky 1993, 2019]. Get feedback at the end of each session [source: J. Beck 2020]. Practically, ask "What did you think of today? Anything I said that bothered you?"
Measure change. Re-rate belief and emotion after each intervention. A shift of even 10–20 points counts.
Teach the client to do it themselves. The goal is self-therapy. Practically, gradually hand over the questions, using written lists and coping cards.
REBT-specific principles [source: Ellis & Dryden 1997]: treat rigid demands ("must", "should", "have to") as the core target; encourage unconditional self-acceptance rather than self-rating; dispute actively and with force; aim for a lasting philosophical shift, not only symptom relief. Practically, address "I must not fail" before "I'll probably fail."

4. Intervention inventory

The interventions are grouped in the order they are usually introduced. All example language is illustrative, written for this guide, not quoted from manuals. [source for 4A–4B: Beck et al. 1979; J. Beck 2020; Greenberger & Padesky 2016; Padesky 1993; Burns 1980. Drawn on: Padesky 1993 and 2019, the Mind Over Mood contents page, the Wills 2008 extract and a Beck Institute training deck; the J. Beck question set comes through a secondary summary. Other attributions are from the books — see Source coverage.]

4A. Identifying thoughts

1. Socialising the client to the cognitive model

Purpose / when: first session or two, so the client sees why you will look at thoughts. Repeat whenever the client returns to "the situation made me feel it."
Steps: (1) Ask for a recent, specific moment when their mood changed. (2) Draw three boxes on paper: situation, thoughts, emotions/body/behaviour. (3) Fill them in with the client's own words. (4) Ask how someone with a different thought would feel in the same situation. (5) Ask the client to say the link in their own words. (6) Give a short reading or diagram as homework.
Example language (illustrative): "When your manager didn't reply, what went through your mind?… And when you thought that, how did you feel?… If a colleague had thought 'she's just swamped', how might they have felt?"
Watch for: the client hearing "it's all in your head." Normalise automatic thoughts as quick and believable, and as something everyone has. It is working when the client brings examples of the link without prompting.

2. Eliciting automatic thoughts and images

Purpose / when: whenever there is a change in emotion, in or between sessions. This is the basic skill everything else depends on.
Steps: (1) Pick a specific situation. (2) Ask what was going through their mind just then. (3) If they can't say: (a) catch an emotion shift happening in the session ("What went through your mind just now?"); (b) have them picture the scene in present tense, with sensory detail, until the feeling returns, then ask again; (c) role-play an interpersonal moment with you as the other person; (d) ask about images or pictures in the mind; (e) offer a guess, or the opposite of the likely thought, so they can correct you; (f) ask what the situation meant about them, their future or others [source: J. Beck 2020; Beck et al. 1979]. (4) Write the thought down in the client's own words. Turn questions into statements ("What if I fail?" becomes "I'll fail") [source: J. Beck 2020].
Example language (illustrative): "Let's slow it down. You're sitting in the car outside the house. What do you see? … What's going through your mind as you look at the front door?" / "Were you thinking something like 'they'll be glad I'm late'?" (offering an opposite)
Watch for: the client reporting interpretations of feelings ("I felt that it was unfair") or descriptions instead of thoughts. Images and memories that carry the meaning. It is working when the elicited thought matches the intensity of the emotion. If it doesn't, keep looking for the hotter thought.

3. Distinguishing and rating emotions and thoughts

Purpose / when: early, so thoughts can be tested and change can be measured.
Steps: (1) Teach that emotions are usually one word (sad, anxious, angry, ashamed). (2) Ask the client to sort a list of their statements into thoughts and feelings. (3) Rate each emotion 0–100 for intensity. (4) Rate belief in each thought 0–100%. (5) Use a feelings list if the client struggles to name emotions [source: Greenberger & Padesky 2016].
Example language (illustrative): "'I feel like a failure' is really a thought. The feeling underneath might be sad or ashamed. How strong is it, 0 to 100?"
Watch for: alexithymia or emotional numbing. If so, track body sensations first. It is working when ratings are differentiated rather than all 100.

4. Identifying the hot thought

Purpose / when: before evaluation, so the work goes to the thought that drives the emotion.
Steps: (1) List all the thoughts and images from the situation. (2) For each, ask what it says about them, their life or their future, to reach the meaning. (3) Ask which one goes best with the strongest emotion, or which one hurts most [source: Greenberger & Padesky 2016]. (4) Circle it and evaluate only that one.
Example language (illustrative): "Of everything you wrote, which thought, if it were true, would hurt the most?"
Watch for: the client picking a safe, peripheral thought. It is working when evaluating the chosen thought shifts the emotion. If the emotion doesn't move, you probably have the wrong thought.

5. Three-column (situation–emotion–thought) record

Purpose / when: first between-session monitoring task (phase 1).
Steps: (1) Fill in one example together in the session. (2) Agree on a trigger to fill it in, which is a noticeable change in mood. (3) Ask for a few entries, not perfection. (4) Review them at the start of the next session.
Example language (illustrative): "This week, any time you notice your mood dip, jot the situation, the feeling and its strength, and whatever went through your mind. Three entries would be great."
Watch for: blank forms (see Section 5, non-adherence) and entries written long after the event. It is working when entries are specific and timely.

6. Recognising thinking errors (cognitive distortions)

Purpose / when: once thoughts are being caught. Naming a pattern creates distance and points to which questions to use.
Steps: (1) Give a list of common thinking errors, e.g. all-or-nothing thinking, overgeneralisation, mental filter, disqualifying the positive, mind-reading, fortune-telling, magnification and minimisation, emotional reasoning, "should" statements, labelling, personalisation [source: Burns 1980; Beck et al. 1979 lists an overlapping set, e.g. arbitrary inference, selective abstraction]. (2) Ask the client to match their recorded thoughts to the categories. (3) Use the category to choose the next questions (e.g. mind-reading leads to evidence about what the other person actually thinks).
Example language (illustrative): "Looking at the list, does 'she thinks I'm useless' fit any of these patterns?"
Watch for: labelling used as self-criticism ("I'm catastrophising again, I'm hopeless") or as a way to dismiss the thought without examining it. Labelling alone rarely changes belief, so follow it with evaluation [synthesis].

4B. Evaluating and responding to thoughts

7. Socratic dialogue: Padesky's four-stage guided discovery

Purpose / when: the main way to examine any thought, assumption or belief, and to draw conclusions from experiments.
Steps [source: Padesky 1993 and Padesky 2019; the 2019 version calls the process "Socratic Dialogue"]: (1) Informational questions: concrete questions the client can answer that bring relevant information into awareness, starting from a specific example (who was there, what was said, how do you know). The 2019 version adds informational experiences, e.g. having the client show how they looked, or comparing a photo or video with their mental image. (2) Empathic listening: listen for the client's own words, metaphors and emotions, and for what is left out; stay open to surprising answers. In 2019 Padesky adds listening for the triggers, beliefs, behaviours, emotions and body reactions that the relevant cognitive model predicts, and checking whether missing pieces are absent or just outside awareness. Do not fill silences with your own answer. (3) Summaries: every few minutes, in the client's own words, metaphors and images rather than your paraphrase (Padesky's "humble parrot"). In the 2019 version they are written down, ideally copied by the client in their own handwriting, with time left at the end of the session for this. (4) Analytical and synthesising questions: what the client makes of the information, and how it fits with the original belief. In 2019 she adds change-oriented questions: which ideas would help most, how the client could use them this week, and what next step they could take.
Example language (illustrative): "So far you've told me three things: your sister called twice last week, she asked you to her birthday, and she was short with you yesterday when her baby was ill. How do those fit with the thought 'nobody cares about me'?"
Watch for: leading questions with an obvious answer ("changing minds"). Rapid-fire questions that feel like a test. Skipping the summary, so the client loses the threads. Stopping before the synthesising question, or the therapist stating the conclusion; Padesky (1993) names too few summaries and skipped synthesis as common errors [source: Padesky 1993]. The best questions often raise some anxiety, which narrows thinking, so keep warmth high [source: Padesky 2019]. It is working when the client says something new that you did not expect and re-rates the belief lower.

8. Standard evaluation questions (the J. Beck question set)

Purpose / when: a structured set the client can learn and use alone on thought records.
Steps: (0) Ask permission to look at the thought, and take belief (0–100) and emotion (0–100) ratings first [source: Beck Institute training deck, Miller 2020]. Then work through these questions, paraphrased from J. Beck (2020) [source: a secondary summary of the book (Shortform), not the book itself]: (a) What is the evidence for and against the thought? (b) Is there another explanation or way of seeing it? (c) What is the worst that could happen and how would I cope, what is the best, and what is most realistic? (d) What is the effect of believing this thought, and what could change if I changed my thinking? (e) What would I tell a friend who had this thought? (f) What should I do about it? Not every question is needed. Pick the ones that fit the thinking error. Afterwards, have the client draw a conclusion that takes all the evidence into account, write a response they believe, then re-rate belief and emotion [source: Miller 2020]. Give the client the list on a card.
Example language (illustrative): "Let's take one question at a time. What tells you this thought is true? … And is there anything that doesn't quite fit it?"
Watch for: mechanical box-ticking, where the answers are given but belief does not move. If so, slow down, look for the meaning under the thought, or switch to an experiment.

9. Examining the evidence

Purpose / when: for thoughts that are factual claims ("I always mess up", "He's angry with me").
Steps: (1) Write two columns: evidence that supports the thought, and evidence that doesn't fully support it. (2) Collect the supporting evidence first and fully, which shows you take the client seriously. (3) Separate facts from interpretations ("he frowned" is a fact; "he was annoyed" is an inference). (4) Gather disconfirming evidence with specific prompts: exceptions, what others have said, what they would notice on a better day, small things they discount. (5) Summarise both columns. (6) Ask for a statement that takes all of the evidence into account [source: Greenberger & Padesky 2016; J. Beck 2020].
Example language (illustrative): "You've given me good reasons for the thought. Now, has there been anything, even small, that doesn't fit with 'I always mess up'?"
Watch for: the client discounting every counter-example ("that doesn't count"). Name it as possible disqualifying the positive and explore it. It is working when the balanced statement includes evidence from both columns.

10. Generating alternative explanations and reattribution

Purpose / when: for personalisation, self-blame, and mind-reading about others' behaviour.
Steps: (1) List every possible reason the event happened, including unlikely ones. (2) Ask the client to assign rough percentages. (3) For self-blame, list all contributing factors (circumstances, other people, chance, the client), which leads into the responsibility pie (No. 14). (4) Decide whether any explanation can be checked [source: Beck et al. 1979, "reattribution"; J. Beck 2020].
Example language (illustrative): "If we brainstorm, what are all the reasons a friend might not answer a message for a day?"
Watch for: the client generating alternatives but believing none. Pair it with a test (ask the friend; wait and see). It is working when the original explanation's percentage drops.

11. Decatastrophising (the "what if" technique)

Purpose / when: for catastrophic predictions in anxiety, worry or panic.
Steps: (1) Ask what the worst outcome would be, to make the feared catastrophe explicit. (2) Estimate how likely it is, sometimes by listing the chain of steps needed and estimating each step's probability [uncertain: probability-chain method associated with anxiety-disorder protocols]. (3) Ask the client to imagine the worst actually happened and walk through how they would cope: who would help, what they would do, how they would feel in a week, a month, a year. (4) Ask for the best possible outcome and the most realistic outcome. (5) Plan coping for the realistic outcome [source: J. Beck 2020; Beck, Emery & Greenberg 1985].
Example language (illustrative): "Suppose you did freeze during the talk. What would happen in the next minute? … And then? … How would you get through the rest of that day?"
Watch for: with OCD, health anxiety and GAD, repeated probability estimates can turn into reassurance-seeking. Shift to coping and to behavioural tests [synthesis, consistent with Salkovskis's appraisal model]. It is working when anxiety ratings fall and the client spontaneously adds "and I'd cope."

12. Double standard / friend perspective (distancing)

Purpose / when: for harsh self-criticism and perfectionist standards.
Steps: (1) Ask what the client would say to a close friend in exactly the same situation, with exactly the same thought. (2) Write that response down. (3) Ask what makes the standard different for them. (4) Explore whether the friend's response could apply to the client [source: Burns 1980 (double-standard method); J. Beck 2020].
Example language (illustrative): "If your best friend had snapped at her kids after a twelve-hour shift and said 'I'm a terrible mother', what would you tell her?"
Watch for: "it's different for me" (often a core belief, so note it for later). Compassion that the client can say but not feel. It is working when the client can say the kind response to themselves with some belief.

13. Advantages–disadvantages (cost–benefit) analysis

Purpose / when: for thoughts, rules or behaviours where usefulness matters more than truth ("I should worry to be prepared", "I must never show weakness"); also when the client is ambivalent about changing a belief.
Steps: (1) Two columns: advantages of believing or acting on this, and disadvantages. (2) Draw out the advantages fully, because they explain why the belief persists. (3) Weigh them, e.g. by percentage split. (4) Ask whether a modified belief could keep the advantages and reduce the costs [source: Burns 1980; J. Beck 2020; Leahy 2017].
Example language (illustrative): "What does the rule 'never ask for help' do for you? … And what does it cost you?"
Watch for: the client concluding the belief is worth keeping. Respect that and look for a more flexible version rather than arguing. It is working when the client drafts a modified rule.

14. Responsibility pie chart

Purpose / when: for excessive self-blame, guilt, and inflated responsibility (including OCD).
Steps: (1) Have the client rate their own responsibility (0–100%) without drawing the pie. (2) List every person or factor that contributed, and put the client last. (3) Draw a circle and assign slices to each factor in turn. (4) The client's slice goes in last, from what remains. (5) Compare it with the first rating and discuss [source: Greenberger & Padesky 2016; Salkovskis's OCD work — uncertain].
Example language (illustrative): "Before we give you a slice, let's give everyone else theirs. How much belongs to the icy road? To the other driver?"
Watch for: use with care where the client really did cause harm. The goal is accurate responsibility, not none. It is working when the final slice is smaller than the first estimate and guilt drops.

15. Defining terms (semantic technique)

Purpose / when: for global labels ("failure", "loser", "selfish").
Steps: (1) Ask the client to define the term precisely. (2) Ask whether anyone they know fully meets the definition. (3) Check the client against every part of the definition. (4) Replace the label with a specific, behaviour-level description [source: Burns 1980; J. Beck 2020].
Example language (illustrative): "What would someone have to do, or be, to count as a 'failure' as a person? … Does that describe you across your whole life?"
Watch for: definitions that move to stay out of reach. It is working when global labels are replaced by "I did X, which I regret."

16. Full thought record (dysfunctional thought record / seven-column record)

Purpose / when: phase 2 onward. This is the main between-session tool for doing restructuring independently.
Steps: The versions differ. Beck et al. (1979) use five columns: situation, emotion, automatic thoughts, rational response, outcome [uncertain]. Mind Over Mood uses seven: situation; moods with ratings; automatic thoughts with the hot thought circled; evidence for the hot thought; evidence against; alternative or balanced thought with a belief rating; re-rating of moods [source: Greenberger & Padesky 2016; the seven-column layout (trigger, emotion, automatic thought, evidence for, evidence against, alternative thought, outcome) is confirmed in the Wills 2008 extract, which also lists Burns's Mood Log and REBT's ABCDE form as other versions]. Procedure: (1) Fill in one record together in the session using a fresh example, close to the event. (2) Introduce the columns gradually: monitoring first (No. 5), evidence columns once the client can find hot thoughts; Wills warns against moving to full records too quickly [source: Wills 2008]. (3) Ask the client to write the balanced thought as a sentence or two that combines both evidence columns. (4) Re-rate. (5) Review at least one record in the next session, with attention to what helped.
Example language (illustrative): "Reading your evidence for and against, could you write one or two sentences that take both columns into account? How much do you believe that, 0 to 100?"
Watch for: records done intellectually with no change in emotion. Records used to criticise oneself. Overuse as a ritual, especially in OCD. Do not copy the published forms; draw a blank grid with the client, or use the client's own licensed workbook. It is working when emotion re-ratings drop by about 10 or more points and the client starts doing the process in their head.

17. Behavioural experiments

Purpose / when: to test predictions and beliefs through action. Use them when verbal evaluation leaves residual belief ("I know it but don't feel it") or when a thought is a prediction that can be tested [source: Bennett-Levy et al. 2004; Beck et al. 1979; J. Beck 2020].
Steps: (1) State the belief or prediction to test, with a belief rating, and make it specific and observable. (2) If useful, state an alternative belief. (3) Design the experiment together: what, where, when, who. (4) Specify exactly what result would support each belief. (5) Anticipate problems, including safety behaviours that would spoil the test. Agree to drop them, or deliberately compare with and without. (6) Do the experiment, in session if possible. (7) Record what happened. (8) Debrief with Socratic questions (what did you predict, what happened, what does that tell you about the belief), then re-rate. (9) Plan the next experiment. Variants: the survey method (ask others what they think, e.g. whether they notice blushing) [source: Burns 1980; Bennett-Levy et al. 2004]; observation experiments (watch how others respond); act-as-if experiments.
Example language (illustrative): "Your prediction is that if you say you disagree in the meeting, people will look away and ignore you. How would we know if that happened? What would count as it not happening?"
Watch for: experiments that are too big (not completed) or unconvincing ("I got lucky"). Agree in advance how luck will be judged. Hidden safety behaviours. For exposure-type experiments, use graded steps, with informed consent and a clear right to stop. Assess medical contraindications for interoceptive tasks (e.g. heart or respiratory conditions, pregnancy) and clear them with the client's physician when in doubt [synthesis]. It is working when belief ratings fall after the debrief and the client designs the next test.

18. Problem-solving when the thought is accurate

Purpose / when: when evaluation shows the thought is largely true ("My job is at risk").
Steps: (1) Acknowledge it plainly. (2) Ask what it means to them, which often uncovers a distorted conclusion ("so I'm worthless") that can be evaluated. (3) Move to structured problem-solving: define the problem, generate options, weigh pros and cons, choose one, plan the steps, review [source: J. Beck 2020]. (4) If nothing can be done, work on acceptance and coping [synthesis].
Example language (illustrative): "It does look as if the contract won't be renewed. Suppose that's so. What does it mean to you? And what's in your power this week?"
Watch for: the therapist pushing a reframe of a true thought, which invalidates the client. It is working when distress shifts from helplessness to action.

19. Coping cards

Purpose / when: to carry balanced responses into moments of high emotion, when thinking is hardest.
Steps: (1) Write the key automatic thought on one side of a card or phone note. (2) On the other side, write the balanced response in the client's own words, plus one or two actions. (3) Agree when to read it: on a regular schedule and when the trigger appears. (4) Revise it as belief changes [source: J. Beck 2020].
Example language (illustrative): "What words from today do you most want to have with you when you're about to cancel on your friends? Let's put them on a card."
Watch for: cards used as reassurance in anxiety disorders, or read without engagement. It is working when the client reports reading the card and then taking the planned action.

20. Intellectual–emotional role-play (point–counterpoint)

Purpose / when: for the "head–heart" gap, when the client knows the balanced view but believes the old thought emotionally. Usually used later, after the client is skilled with thought records.
Steps: (1) Get the client's permission. (2) The client argues the "emotional" side, voicing the old belief. The therapist argues the "intellectual" side, using the client's own evidence from earlier work. (3) Then swap roles, so the client argues the balanced side against the therapist voicing the old belief. (4) Debrief and re-rate. Make clear that the therapist is voicing the client's thoughts and does not hold that view [source: J. Beck 2020 — name as recalled, uncertain; Young's schema "point–counterpoint" is a related variant].
Example language (illustrative): "I'll play the part of you that says you're unlovable, using the arguments you've told me, and you answer back with what we've found. Remember I'm only voicing it, I don't believe it. Ready?"
Watch for: the client feeling attacked or overwhelmed. Stop and debrief if this happens. Don't use it with fragile alliances. It is working when the client answers with growing conviction and belief in the old view falls.

4C. Assumptions and core beliefs

[source for 4C: Padesky 1994 for Nos. 23–26a; J. Beck 2020; Burns 1980; Beck et al. 1979]

21. Downward arrow (vertical arrow)

Purpose / when: to go from an automatic thought to the assumption and core belief beneath it, once themes recur.
Steps: (1) Take a hot thought. (2) Assume for now that it is true. (3) Ask what that would mean, what is bad about it, or what it would say about them. (4) Repeat for each answer until a global statement about self, others or the world appears, or until the emotion peaks. (5) Write out the chain. Ask about meaning, not causes [source: Burns 1980; J. Beck 2020].
Example language (illustrative): "Let's suppose for a moment it's true that your report had errors. If so, what would that mean? … And if that were true, what would that say about you?"
Watch for: rising distress. Go slowly and summarise compassionately at the end. Watch for loops that keep returning to the same level. It is working when the chain ends in a short, emotionally charged belief the client recognises ("Yes, that's it").

22. Identifying and modifying intermediate beliefs (rules and assumptions)

Purpose / when: phase 3. Use it when the same automatic thoughts recur across situations.
Steps: (1) Spot assumptions in themes, in "should" statements, from the downward arrow, or by offering the first half of a conditional ("If I don't do everything perfectly…") for the client to complete. (2) Turn rules ("I must…") into conditional assumptions ("If I don't…, then…"), because a prediction can be tested. (3) Rate belief. (4) Examine advantages and disadvantages (No. 13), the evidence (No. 9), and the double standard (No. 12). (5) Test with behavioural experiments (No. 17), e.g. deliberately submitting work that is good enough but not perfect. (6) Draft a more flexible alternative assumption ("If I do a reasonable job, most people will be satisfied, and if not, I can cope"). (7) Rehearse it on cards [source: J. Beck 2020].
Example language (illustrative): "You seem to live by something like 'If I say no, people will leave.' Shall we find out how true that is?"
Watch for: rules that protect against a core belief, so loosening them can raise anxiety at first. Warn the client about this. It is working when the client acts against the old rule and reports the outcome.

23. Core belief identification and core belief worksheet

Purpose / when: phase 4. Use it when core beliefs maintain problems or drive relapse.
Steps: (1) Identify the core belief: follow the affect and ask what the event says about the client, about other people, and about their life or the world; use sentence completions ("I am…", "People are…", "The world is…"); belief questionnaires such as the Dysfunctional Attitude Scale or Young's schema questionnaire can be a starting point; put the belief in the client's own words and ask for linked images or memories [source: Padesky 1994]. J. Beck groups negative self-beliefs into helpless, unlovable and worthless categories [source: J. Beck 2020, via secondary summary]. (2) Present it as a hypothesis and ask the client to rate belief. (3) Explain why it resists evidence with Padesky's prejudice metaphor: the client names a prejudice someone else holds, describes how contrary evidence gets ignored, distorted or treated as an exception, considers what would change that person's mind (usually observation and record-keeping), then applies this to their own belief [source: Padesky 1994]. (4) Build the new belief early by asking how the client would like themselves, others or the world to be, using admired people if they are stuck, in the client's own words; expect the wording to change over weeks. Versions differ: Padesky (1994) states old and new beliefs as absolutes ("I am lovable"), because a hedged new belief gets absorbed as an exception [source: Padesky 1994]; J. Beck (2020) is usually described as aiming for a more realistic, moderate new belief [uncertain]. Choose one approach and explain it to the client. (5) Track over weeks on a core belief worksheet: weekly 0–100% ratings of the old and new beliefs, evidence that supports the new belief, and evidence that seemed to support the old belief reframed to fit the new one (Padesky 1994 credits the worksheet to J. Beck, 1992, unpublished). (6) Add "acting as if" the new belief were true [source: J. Beck 2020, uncertain].
Example language (illustrative): "From what we've seen, there may be a deep belief something like 'I'm a burden.' Does that fit? If you held a more balanced view of yourself, how might it be worded?"
Watch for: rejecting the new belief because it feels "fake." Slow change is expected. Do not reproduce the published worksheet; build a simple table in session. It is working when the new belief's rating rises gradually and the client spontaneously notices disconfirming evidence.

24. Positive data log

Purpose / when: to counter the information-processing bias that keeps a core belief going.
Steps [source: Padesky 1994, unless marked]: (1) Explain the rationale with the prejudice metaphor (No. 23): the old belief filters what is noticed. (2) State the new belief. (3) The client logs, every day, each observation consistent with the new belief, however small or easily dismissed; a confidence rating beside an entry (e.g. 20%) lets them log data they do not yet fully believe. (4) Expect them to have trouble seeing such data at first: watch for it in session and help them write it down. (5) Build categories of evidence together (e.g. looking after myself, saying what I think) and add new ones over time. (6) Rate the new belief weekly on a 0–100% continuum (No. 25) and keep the log going for at least six months. If they struggle, ask what they would count as evidence if it happened to a friend [synthesis].
Example language (illustrative): "This week, every time something happens that fits 'I'm competent at some things', even making a good cup of coffee, jot it down. Small ones count."
Watch for: "yes, but" discounting. Tackle it as a thought with No. 9 and No. 12. Slow change is normal (see Phase 4). It is working when the entries grow in number and range and the client starts noticing supporting data without prompting [source: Padesky 1994, for the last point].

25. Continuum technique (scaling)

Purpose / when: for all-or-nothing core beliefs ("I'm a failure", "I'm bad").
Steps [source: Padesky 1994, unless marked]: (a) Rating on the new-belief continuum: draw a one-way 0–100% scale for the new belief (e.g. 0–100% lovable), not a two-way scale running from old belief to new, because then every movement counts as progress; rate weekly, ideally noting the highest and lowest point that week. (b) Criteria continua: ask the client to rate themselves globally on the new quality (often near 0%); list what they mean by it (the criteria); define each criterion's ends in absolute terms (100% = perfect, 0% = complete absence, not "only a little"); have the client rate themselves on each criterion (usually mid-range); compare the global and specific ratings and explore the "scoring rules" by which one negative wipes out the positives; use the criteria later for planning. (c) Two-dimensional charts: when a belief links two ideas ("getting close to people is painful"), rate examples on both dimensions, plot them on a grid and check whether the two really go together; or graph the belief (e.g. worth rises only with perfection) and look for real data points off the line. Variant: placing other people (acquaintances, public figures) along the line to calibrate the anchors is a common teaching variant not described in Padesky 1994 [uncertain].
Example language (illustrative): "Who would be at 100% 'a good mother'? What would she do? … And where would you put yourself, given what you've listed?"
Watch for: the client moving the anchors to stay at the bottom. Work explicitly on the criteria. It is working when the client places themselves away from the extreme and the global label softens.

26. Historical test of a core belief

Purpose / when: for long-held core beliefs. Use it after a working alliance and some symptomatic improvement.
Steps [source: Padesky 1994, which credits the method to Young 1984]: (1) Name the core belief. (2) Divide the client's life into periods, often by natural transitions (schools, moves, jobs), starting with infancy and toddlerhood, where clients rarely judge themselves harshly. (3) For each period, list evidence that seemed to support the old belief and evidence that did not; outside sources can help (reading about typical child development, asking family). (4) Write a summary for each period, then an overall summary, which often holds the seeds of the new belief. (5) Leave blank pages for periods the client cannot remember, and note sharp shifts before and after them. (6) For beliefs about others or the world, also recall people and events that were positive or reliable. (7) Re-rate both beliefs. Reframing old evidence with adult understanding (e.g. how a child was treated) belongs in the summary discussion [synthesis].
Example language (illustrative): "Let's look at age 6 to 11. What happened then that seemed to show you were unlovable? … Knowing what you know now, how else could we understand that?"
Watch for: traumatic memories surfacing. Pace the work, attend to safety, and consider trauma-focused methods if indicated. It is working when the client re-evaluates past events with more compassion.

26a. Early-scene role-play (psychodrama) for core beliefs

Purpose / when: to re-evaluate a core belief at an emotional level by re-enacting an early scene in which it was activated. Padesky (1994) suits it to clients who avoid affect; clients who dissociate under strong affect may do better with continua or the historical test [source: Padesky 1994].
Steps [source: Padesky 1994]: (1) Choose an early scene where the belief was activated and have the client describe it. (2) Enact it: the client plays themselves as a child; the therapist plays the significant other, following the client's directions and the key words and actions they recall. (3) Process it: emotions, beliefs activated, what the client did or held back, links to current life, alternative explanations. (4) Replay the scene from another viewpoint (e.g. the parent's or a bystander's), with the therapist playing the child. (5) Re-enact it in the client's adult voice: speaking up, expressing feelings, defending the right to make mistakes; scripting can help. (6) Rehearse until the new response feels real in belief, emotion and behaviour. Keep the daily logs and continua going alongside.
Example language (illustrative): "Tell me how your father stood and what he said, and I'll play him as closely as I can. You be yourself at eight. Afterwards we'll talk about what came up."
Watch for: the aim is re-evaluation, not catharsis [source: Padesky 1994]. Intense affect, dissociation and trauma memories: get consent, agree a stop signal, pace the work and leave time to ground the client before the session ends [synthesis]. It is working when the client can answer the old scene from an adult perspective and new-belief ratings rise.

4D. REBT variant: disputing irrational beliefs

[source for 4D: Turner 2016 and Wikipedia: Rational emotive behavior therapy for the ABC model, the three disputing types, rational-belief components and healthy vs unhealthy emotions; Ellis 1962; Ellis & Dryden 1997; Dryden & Neenan 2004; DiGiuseppe et al. 2014]

27. ABC assessment

Purpose / when: first REBT sessions, for each problem.
Steps: (1) Clarify C, the unhealthy negative emotion and behaviour. REBT treats unhealthy negative emotions (e.g. anxiety, depression, unhealthy anger, shame, guilt) and healthy negative emotions (e.g. concern, sadness, annoyance, disappointment, remorse) as different in kind, not just in intensity; the healthy ones go with goal-directed action [source: Turner 2016; Wikipedia: REBT; the exact emotion pairs follow Dryden's work, uncertain]. (2) Find the critical A: the aspect of the event the client is most disturbed about. Accept it as true for now. (3) Assess the irrational beliefs at B: a rigid demand (must), plus awfulising, low frustration tolerance ("I can't stand it") or global rating of self or others [source: Wikipedia: REBT]. (4) Teach the B–C connection, not A–C.
Example language (illustrative): "Let's assume for now he did criticise you in front of everyone. What were you demanding about that, that made you so ashamed rather than just disappointed?"
Watch for: getting stuck in disputing the A (that is Beckian work, and REBT defers it). It is working when the client can state their own "must".

28. Disputing (logical, empirical, pragmatic; Socratic and didactic)

Purpose / when: the central REBT intervention, used once the ABC is clear.
Steps: Ask three kinds of questions about the irrational belief, then compare it with the rational alternative: empirical (Where is the evidence that you must? Does reality require it?), logical (Does it follow that because you strongly want something, it must happen?), pragmatic or functional (Where does holding this demand get you?) [source: Turner 2016 names the three types; question wording illustrative]. Disputing can be Socratic (questions) or didactic (short explanation), and may use humour, metaphor or self-disclosure [source: DiGiuseppe et al. 2014 — uncertain]. Build the rational belief with its four parts: a flexible preference ("I'd strongly prefer…, but it doesn't have to…"), anti-awfulising (bad, not awful), high frustration tolerance (hard but bearable), and acceptance of self or others as fallible, without global rating [source: Turner 2016]. Assign rational-belief rehearsal and behavioural homework that acts against the demand.
Example language (illustrative): "You'd very much like your father to approve of you, which makes sense. Does it follow that he absolutely must?"
Watch for: the forceful, directive style can rupture the alliance with some clients. Adjust the vigour to the client [synthesis]. It is working when the client voices the preference version with conviction and their emotion shifts from unhealthy to healthy (e.g. anxiety to concern).

29. Rational-emotive imagery

Purpose / when: for emotional practice of a new belief after disputing.
Steps: (1) The client vividly imagines the critical A. (2) They let the unhealthy emotion arise. (3) They change the feeling to the healthy negative emotion (e.g. disappointment instead of shame), using the rational belief. (4) They notice what they told themselves. (5) Practise daily [source: Ellis & Dryden 1997; technique originated by Maultsby — uncertain].
Example language (illustrative): "Picture the moment your proposal was rejected. Let yourself feel the shame. Now see if you can change it to strong disappointment, without the shame. What did you say to yourself to do that?"
Watch for: overwhelming affect; use with care after trauma. It is working when the client can shift emotion within the image and reports doing so in real situations.

Not covered as separate entries: imagery rescripting of early memories, mindfulness and defusion. Some core-belief protocols use these methods, but they belong to other modalities. Defusion is the third-wave recasting of the same targets [synthesis].

5. Handling problems

The source for each item is tagged. Items marked [synthesis] are this guide's own clinical guidance, not taken from the primary texts.

Client can't identify thoughts ("I just feel bad"). Use the eliciting methods in No. 2: catch emotion shifts in the session, use imagery replay or role-play, offer the opposite thought, ask about meaning. Accept images and "felt sense" as cognitions. If verbal access stays poor, start with behavioural work and come back later [source: J. Beck 2020; Beck et al. 1979].

"I know it in my head but don't believe it" (head–heart gap). Check whether you have the hot thought or a peripheral one. Look for the underlying meaning or core belief. Move from verbal methods to behavioural experiments and the intellectual–emotional role-play. Rate belief often so that partial change is visible. Explain that the felt sense of belief lags behind repeated practice [source: J. Beck 2020; Bennett-Levy et al. 2004].

Client feels invalidated, criticised or "told they're wrong". This is the main known risk of mechanical use. Response: stop questioning. Reflect the emotion and say that the thought makes sense given their history. Name what happened ("I may have come across as if I was trying to talk you out of how you feel"). Ask for feedback. Then return with an openly curious stance, starting with the evidence for the thought. Clients resist therapists they see as too persuasive, so acknowledge their doubts rather than arguing [source: Padesky 1993 for the stance; Wills 2008 extract on over-persuasion; repair wording is synthesis].

"Yes, but" / discounting every counter-example. Treat the discounting itself as a thought to examine (disqualifying the positive). Ask what would count as evidence and agree on it before collecting data. Use the double standard. Explain the discounting with the prejudice metaphor and run a positive data log with confidence ratings, so half-believed data still get recorded (No. 23–24). Do not argue harder: increasing therapist pressure usually increases "yes, buts" [source: J. Beck 2020; Padesky 1994 for the prejudice metaphor and data logs; Wills 2008 extract: clients resist therapists seen as too persuasive].

Thought turns out to be true. Stop evaluating it and move to meaning (downward arrow), problem-solving or acceptance (No. 18). Don't reframe real losses, discrimination or danger [source: J. Beck 2020; discrimination point synthesis].

Overwhelm or high distress in session. Cognitive work needs some reflective capacity. Slow down and use shorter questions. Ground the client in the room. Move from core beliefs back to concrete current situations or behavioural tasks. In severe depression, prioritise behavioural activation [source: Beck et al. 1979]. Do not do the downward arrow or historical test late in a session or with an unstable client [synthesis].

Non-adherence to thought records and homework. Start every homework task in the session. Make it smaller (one entry; a phone note; just the hot thought). Link it explicitly to the client's goal. Ask in advance what might get in the way, and rate likelihood of completion, adjusting if under about 90% [uncertain: J. Beck uses a similar likelihood check]. Explore cognitions about homework ("It won't help", "I'll do it wrong") as automatic thoughts. Always review whatever was done. Frame non-completion as useful data, not failure [source: J. Beck 2020; Kazantzis et al. 2005 on homework].

Stuck: belief does not shift over several sessions. Re-check the conceptualisation. Possibilities: wrong target level (an assumption or core belief maintaining it); secondary gain or high costs of change (do an advantages–disadvantages analysis of changing); safety behaviours spoiling experiments; untreated comorbidity (substance use, trauma, medical); or a problem that is mainly behavioural or situational. Review with supervision. Consider whether a different method fits better [synthesis, consistent with J. Beck 2020's guidance on difficult cases].

Resistance or reactance to questioning. Some clients experience questions as manipulative ("just tell me what you think"). Be transparent about why you ask questions. Offer a short didactic explanation when they need information. Give choices about which thought to work on and how. Lower the pace and number of questions [source: Padesky 1993 on balancing didactic and Socratic; rest synthesis].

Alliance ruptures. Use routine end-of-session feedback to detect them early. When you notice a withdrawal or confrontation marker, raise it, take your share of responsibility, explore the client's experience without defending yourself, and agree on an adjusted way of working. Restructuring can also be used collaboratively on thoughts about the therapist ("she thinks I'm hopeless"), but only after empathic repair, never as a way to dismiss the complaint [synthesis drawing on Safran & Muran 2000; J. Beck 2020 on thoughts about the therapist].

Rumination and over-analysis. For some depressed or worried clients, thought records become another form of rumination: long, repetitive, abstract. Signs include very long records and no drop in emotion. Response: time-limit the records, keep them to concrete situations, shift to behavioural experiments and to activity. Or consider a process-focused approach (rumination-focused CBT, metacognitive therapy, defusion), which works on how the client relates to thoughts rather than their content [synthesis].

Restructuring becomes reassurance or neutralising (OCD, health anxiety, GAD). Repeatedly evaluating intrusive thoughts or the likelihood of illness can work like reassurance-seeking or compulsive checking. Target the appraisals (responsibility, the meaning of having the thought, intolerance of uncertainty) rather than the content of obsessions. Do evaluation once, collaboratively, not on demand. Pair it with exposure and response prevention [synthesis consistent with Salkovskis's cognitive model of OCD — uncertain].

Trauma-related beliefs (self-blame, "the world is completely dangerous"). Use gentle Socratic dialogue around the meaning of the trauma, as in cognitive processing therapy and cognitive therapy for PTSD. Do not challenge the facts of what happened. Watch for hindsight bias in self-blame and use the responsibility pie. Use trauma-specific protocols where PTSD is present [synthesis; CPT and CT-PTSD not drawn on].

Psychotic or delusional beliefs. Ordinary restructuring done head-on tends to entrench delusions. Use the CBT-for-psychosis approach: engage first, normalise, examine the distress and the meaning the client gives it, use alternative explanations tentatively, and test gently with collaboratively designed experiments. Work alongside the psychiatric team [synthesis; specialist CBTp texts not drawn on].

Suicidal thinking and hopelessness. Hopelessness is a central target in Beck's model. Before any restructuring of hopeless thoughts, complete a standard suicide risk assessment and develop a collaborative safety plan: warning signs, internal coping, people and places for distraction, people to ask for help, professional and crisis contacts, and making the environment safe through lethal-means counselling (secure storage or removal of means). Then examine hopeless predictions with the evidence and alternative-explanation questions, and build "reasons for living" coping cards. Don't debate whether suicide is "rational"; work on the hopeless appraisals and current problems. Escalate care as local protocols require [source: Beck et al. 1979 on hopelessness; Stanley & Brown safety planning; procedure details synthesis].

Cultural, religious and value-based beliefs. Don't dispute faith commitments or cultural values. Separate the value ("family matters") from the distorted appraisal attached to it ("if my parents are disappointed, I am worthless"). Use the client's own tradition's resources as evidence where the client wants this. Be careful with Ellis-style disputing of "shoulds" that are moral or communal commitments [synthesis].

Clients with limited literacy, cognitive impairment, or children. Use spoken or pictorial records (faces, thought bubbles), simpler questions, and more experiments than verbal evaluation. Involve carers in prompts where appropriate [synthesis].

Therapist-side errors (self-check). Leading questions with a single right answer. Too many questions without summaries. Stopping before the synthesising question. Choosing low-emotion thoughts. Disputing before empathising. Delivering a reframe instead of letting the client synthesise. Never turning conclusions into action. Moving to full thought records too quickly. Consider auditing a recorded session with Padesky's Socratic Dialogue Rating Scale, whose manual is published on padesky.com [source: Padesky 1993 for too few summaries and skipped synthesis; Wills 2008 extract for rushing thought records; rating-scale manual not drawn on].

Sources

Sources:

Further reading (not drawn on): a University of Delaware lecture slide on questioning automatic thoughts (no author), the UCL guided-discovery competence page and a Mind Over Mood worksheet PDF.

Books and papers cited (those marked "drawn on" are used directly; the rest are from prior knowledge):

Beck, A. T. (1976). Cognitive Therapy and the Emotional Disorders. International Universities Press.
Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive Therapy of Depression. Guilford.
Beck, A. T., Emery, G., & Greenberg, R. L. (1985). Anxiety Disorders and Phobias: A Cognitive Perspective. Basic Books.
Beck, J. S. (2020). Cognitive Behavior Therapy: Basics and Beyond (3rd ed.). Guilford.
Bennett-Levy, J., Butler, G., Fennell, M., Hackmann, A., Mueller, M., & Westbrook, D. (2004). Oxford Guide to Behavioural Experiments in Cognitive Therapy. Oxford University Press.
Burns, D. D. (1980). Feeling Good: The New Mood Therapy. Morrow.
DiGiuseppe, R., Doyle, K. A., Dryden, W., & Backx, W. (2014). A Practitioner's Guide to Rational Emotive Behavior Therapy (3rd ed.). Oxford University Press.
Dryden, W., & Neenan, M. (2004). The Rational Emotive Behavioural Approach to Therapeutic Change. Sage.
Ellis, A. (1962). Reason and Emotion in Psychotherapy. Lyle Stuart.
Ellis, A., & Dryden, W. (1997). The Practice of Rational Emotive Behavior Therapy (2nd ed.). Springer.
Greenberger, D., & Padesky, C. A. (2016). Mind Over Mood (2nd ed.). Guilford.
Kazantzis, N., Deane, F. P., Ronan, K. R., & L'Abate, L. (Eds.) (2005). Using Homework Assignments in Cognitive Behavior Therapy. Routledge.
Kazantzis, N., Dattilio, F. M., & Dobson, K. S. (2017). The Therapeutic Relationship in Cognitive-Behavioral Therapy. Guilford.
Leahy, R. L. (2017). Cognitive Therapy Techniques: A Practitioner's Guide (2nd ed.). Guilford.
Padesky, C. A. (1993). Socratic questioning: Changing minds or guiding discovery? Keynote address, European Congress of Behavioural and Cognitive Therapies, London. (drawn on)
Padesky, C. A. (1994). Schema change processes in cognitive therapy. Clinical Psychology and Psychotherapy, 1(5), 267–278. (drawn on)
Padesky, C. A., & Kennerley, H. (2023). Dialogues for Discovery: Improving Psychotherapy's Effectiveness. Oxford University Press.
Safran, J. D., & Muran, J. C. (2000). Negotiating the Therapeutic Alliance. Guilford.
Young, J. E. (1999). Cognitive Therapy for Personality Disorders: A Schema-Focused Approach (3rd ed.). Professional Resource Press.
Padesky, C. A. (2019). Action, dialogue & discovery: Reflections on Socratic questioning 25 years later. Invited address, Ninth World Congress of Behavioural and Cognitive Therapies, Berlin. (drawn on)
Turner, M. J. (2016). Rational Emotive Behavior Therapy (REBT), irrational and rational beliefs, and the mental health of athletes. Frontiers in Psychology, 7, 1423. (drawn on)
Wills, F. (2008). Skills in Cognitive Behaviour Counselling & Psychotherapy. Sage. (chapter 4 extract drawn on)
Young, J. E. (1984). Cited in Padesky (1994) as the origin of the historical test of schema; full reference not checked.

← Previous · II-077Cognitive therapy Next · II-079 →Attribution therapy