Socratic dialogue and thought records

In short

• Socratic dialogue is not a way to win an argument with a thought. It is a series of questions that lets a person reach a new conclusion from their own information. In 1993 Christine Padesky contrasted two styles: "changing minds" (the therapist already knows the answer and steers towards it) and "guiding discovery" (the therapist genuinely doesn't know).
• Padesky's model has four stages: informational questions, listening, summarising, and a synthesising question. The most common mistake is to stop before the fourth.
• A thought record is the same dialogue on paper. Situation, emotions rated 0–100, the "hot" thought, evidence for and against, a balanced thought, and a re-rating. A good outcome is a drop of roughly 10–50% in the emotion, not a drop to zero.
• CBT is not positive thinking. The aim is an accurate thought, not a pleasant one. In lab experiments, people with low self-esteem felt worse after repeating "I'm a lovable person".
• The evidence is honestly mixed. CBT as a whole works well for depression. In one observational study, more Socratic questioning was linked to a bigger drop in symptoms by the next session. But direct evidence for the method is thin, and component studies have not shown that the thought-challenging part adds much.
• The strongest case for the skill is protection after therapy ends. People who learned cognitive therapy skills well and used them on their own were less likely to relapse into depression over the following year.
• It matters not only whether homework is done, but how well. Both the amount and the quality of homework completion are linked to therapy outcomes.
• Most "CBT apps" are not thought records. Of 36 apps presented as cognitive restructuring tools, 21 had no step for writing an alternative thought.

Timeline

How the techniques for working with thoughts took shape, and how they were tested. Details and sources are in the chapters below.

1. 1979

   Beck, Rush, Shaw and Emery, "Cognitive Therapy of Depression": the Daily Record of Dysfunctional Thoughts, forerunner of the thought record

   Cognitive line

2. 1980

   David Burns, "Feeling Good": cognitive therapy reaches a mass audience

   Cognitive line

3. 1983

   Padesky's seven-column thought record: "rational response" becomes "balanced thought"

   Cognitive line

4. 1993

   Padesky's London address: "changing minds or guiding discovery?" and the four stages

   Cognitive line

5. 1994

   Padesky describes core-belief work: the continuum, the positive data log, the historical test

   Cognitive line

6. 1995

   First edition of Greenberger and Padesky's "Mind Over Mood"

   Cognitive line

7. 1996

   Jacobson and colleagues: behavioural activation alone does as well as full cognitive therapy

   Evidence

8. 2000

   Kazantzis and colleagues' first meta-analysis of homework

   Evidence

9. 2003

   Bennett-Levy: behavioural experiments feel more convincing than thought records

   Evidence

10. 2007

    Longmore and Worrell: "Do we need to challenge thoughts?"; Strunk and colleagues: CT skills and relapse risk

    Evidence

11. 2009

    Wood and colleagues: positive self-statements and low self-esteem

    Evidence

12. 2015

    Clark and Egan: no direct evidence for the method; Braun and colleagues: first data linking questions to symptom change

    Evidence

13. 2016

    Kazantzis and colleagues' meta-analysis: both quantity and quality of homework matter

    Evidence

14. 2019

    Padesky in Berlin: "Socratic dialogue" rather than "Socratic questions"; a review finds 110 different thought record forms

    Cognitive line

15. 2022

    App review: 21 of 36 "restructuring tools" lack an alternative-thought step

    Evidence

How to read these notes: evidence labels

Every claim here carries a label that tells you how solid it is.

The label comes first in brackets, followed by the source. Proven by research means trials, meta-analyses or other empirical studies (we say so when it is a single, small or observational study). Clinical consensus means professional standards, manuals and accepted CBT practice. The author's own account is what the creator of a method says; it has not been independently checked. Documented fact covers dates, publications and names. Interpretation is a reading of the facts, or a point still under debate. Sources disagree means both versions are given. Not verified means we could not open the primary source, so treat it with care. All dialogue and thought record examples here are composites, not stories of real people.

What cognitive techniques are and where they fit in CBT

Before you can change a thought you have to find it, and before you find it you need to agree who looks for it, and how.

Earlier episodes in the series covered the cognitive model (episode 2), cognitive distortions (episode 3) and how a therapist builds a picture of the case and structures a session (episode 4). This one is about the next step: what you do with a thought once you have noticed it.

Cognitive techniques are the part of CBT that works with the content of thoughts. That means automatic thoughts (quick phrases that flash through the mind in response to an event: "I've ruined everything"), intermediate beliefs (rules and assumptions: "if I make a mistake, people will stop respecting me") and core beliefs (broad conclusions about oneself, other people and the world: "I'm incompetent"). The umbrella term is cognitive restructuring: thoughts are explored, evaluated and, where useful, replaced with more accurate ones clinical consensus; Clark, 2014, cited in Erhardt et al., 2022.

It helps to separate two things.

• The process: how the therapist talks. In CBT this is guided discovery, where the person reaches a new view on their own, using their own information. Its main vehicle is Socratic dialogue. It is the style through which everything else is done, not a separate technique. Guided discovery and Socratic questioning are part of the CBT competence framework used to train therapists in the UK clinical consensus; UCL CORE, Roth & Pilling.
• The tools: what is done on paper and in life. Thought records, rating scales, evidence for and against, alternative explanations, the double standard, decatastrophising, the downward arrow, the responsibility pie, the continuum, behavioural experiments.

Behind both sits collaborative empiricism: a thought is treated as a hypothesis that therapist and client test together. The therapist is a co-investigator, not a judge deciding whether the person thinks correctly clinical consensus; Padesky, 1993; UCL CORE.

Experts disagree on whether the Socratic method is the heart of CBT. In a survey of 15 CBT researchers, some called it a central feature of therapy and others a helpful but optional extra (interpretation; Clark & Egan, 2018). We come back to this in the chapter on evidence.

Changing minds or guiding discovery: Padesky's four stages

The best question a therapist can ask is one they don't know the answer to.

Two conversations with the same thought

Imagine someone says, "I'm a failure."

In the first conversation the therapist replies: "But you have a job, a family, a degree. Could a failure do all that? So you're not a failure." The person nods. The thought has been "disproved" in three moves.

In the second, the therapist asks: "When did you last think that? What happened then? What does 'failure' mean to you?" They don't know where this will lead, and a few minutes later the person says something the therapist didn't expect.

The psychologist Christine Padesky (later co-author of the self-help book "Mind Over Mood", 1995) built a keynote address at the European Congress of Behavioural and Cognitive Therapies in London on 24 September 1993 around this difference documented fact; Padesky, 1993. She called the first style changing minds: the therapist already has a conclusion and leads the client there in a "one-two-three-aha!" pattern. The second is guiding discovery: helping the person see what they already know but haven't noticed the author's own account; Padesky, 1993.

Her working rule: "If you are not truly curious to know the answer, don't ask the question" the author's own account; Padesky, 1993. And from the same talk: "We are not simply fixing problems but also teaching ways of finding solutions" the author's own account; Padesky, 1993.

Why does this matter beyond the therapy room? Because it is how we talk to ourselves. When the inner voice says "I've ruined it", we usually either believe it or argue back: "No, you're doing great." Both are attempts to change a mind. The third way is to question the thought as if you really don't know the answer.

What makes a question Socratic

In Padesky's definition, Socratic questions are questions that the author's own account; Padesky, 1993:

1. the client has the knowledge to answer;
2. draw attention to relevant information that is currently out of focus;
3. generally move from the concrete to the more abstract;
4. help the client apply the new information to re-evaluate an earlier conclusion or build a new one.

The four stages

The stage names come from Padesky's 1993 address documented fact; Padesky, 1993.

Stage 1 usually starts by getting specific. "I always mess things up" is a label; "yesterday at 4:40 pm I sent a report with a mistake in a table" is a situation you can work with.

Stage 2 is not a pause before the next question; it is work in its own right. A client's unexpected word often matters more than the therapist's plan. Padesky warns that "if you are too confident of where you are going, you only look ahead and miss detours" the author's own account; Padesky, 1993.

Stage 3, according to Padesky, is where one of the most common mistakes happens: therapists summarise too rarely the author's own account; Padesky, 1993.

Stage 4 is the point of the whole exercise. Padesky notes that therapists often "stop short of this critically important final stage": the facts are gathered, but the person is never invited to link them back to the original thought the author's own account; Padesky, 1993.

Common mistakes

According to Padesky the author's own account; Padesky, 1993:

• asking questions the person cannot answer (for example, "what are you feeling right now?" to someone who struggles to identify emotions; asking about body sensations may work better);
• "building a case": waiting for answers that confirm a planned conclusion and not hearing unexpected ones;
• summarising too little;
• stopping before synthesis.

The psychologist Scott Waltman adds a mistake he calls provided discovery: the therapist simply tells the client the conclusion they are "supposed" to reach. He also warns against using thought records to prove to someone that their thinking is distorted the author's own account; Waltman, 2020.

What changed 25 years later

In 2019, at the Ninth World Congress of Behavioural and Cognitive Therapies in Berlin, Padesky revisited the model documented fact; Padesky, 2019. The main updates the author's own account; Padesky, 2019:

• "Socratic dialogue" rather than "Socratic questions", to stress that it is a joint process.
• Information can be created through action, not only asked for. She highlights four kinds: interactive writing, behavioural experiments, imagery and role play.
• Summaries in the client's own words. The therapist should be a "humble parrot", using only the client's exact words, metaphors and images. Written summaries help.
• The synthesising question need not be about testing a thought. It can be "How would you like to be?" or "What are the next steps you could take?"
• Silence is an intervention too. The therapist needs the courage to stay silent while the person searches for a meaningful answer.

Padesky herself notes that we don't yet know whether the four-stage model makes therapists' learning easier; that hasn't been tested the author's own account; Padesky, 2019.

Another version of the same steps

Waltman describes a similar sequence the author's own account; Waltman, 2020:

1. Targeting: pick the thought most tied to the person's difficulties and underlying beliefs, rather than working through everything.
2. Understanding their view: grasp why the belief makes sense from where the person stands; this is the basis for validation (acknowledging that the experience is understandable).
3. Curious exploration: what might the person be missing, including because of avoidance?
4. Integration: bring both sides together into a balanced thought, then plan how to test it in daily life.

Waltman stresses that there are no "silver bullet" questions that disprove any and every belief, and calls Socratic questioning among the hardest skills for a psychotherapist to learn the author's own account; Waltman, 2020.

What it looks like in practice: "I never finish anything"

A composite example: what each question does, and how it differs from an argument.

A composite client; let's call her Lana. The theme is everyday self-criticism about abandoned projects. This is not a real person's story. The right-hand column shows Padesky's stage and what the question is doing.

Notice what is missing. The therapist never says "that's overgeneralisation" and never lists Lana's achievements. The strongest discoveries — the rule "only flawless counts" and the link "I finish when there's an outside deadline" — came from her. The therapist couldn't have suggested them, because they didn't know.

How not to do it

The same situation in a changing-minds style (composite):

What went wrong:

• the therapist knows the answer in advance and is "building a case" (Padesky, 1993);
• the therapist states the conclusion, which is provided discovery (Waltman, 2020);
• the "yes, but…" is pushed past rather than explored, so the rule "only flawless counts" never surfaces;
• there is no summary and no synthesis in the client's words;
• the label "irrational" turns the conversation into a contest about who is right.

Cheat sheet: what each question does

Does the Socratic method work? What we actually know

It is called a cornerstone of CBT, but testing started late and the data are still thin.

An awkward review

In 2015 the Australian researchers Gavin Clark and Sarah Egan published a review of the Socratic method in CBT. Their conclusion: the method is described as an important part of CBT, but no empirical case for it has been made, and no direct evidence shows that it is beneficial in CBT (interpretation; Clark & Egan, 2015). A year earlier, a published expert panel that included Padesky herself acknowledged that even the definition and purpose of Socratic dialogue remain unresolved documented fact; Kazantzis et al., 2014.

The first supporting data

The same year, Daniel Strunk's team at Ohio State University published a study proven by research, a single observational study; Braun et al., 2015:

• 55 adults with major depressive disorder, 16 weeks of cognitive therapy;
• video recordings of the first three sessions were rated by independent observers on a five-item scale built for the study (inter-rater agreement ICC = 0.77, where 1 would be perfect agreement);
• the therapists were four graduate students with one to two years of clinical experience.

The result: after sessions in which the therapist used more Socratic questioning than usual for that client, depressive symptoms fell more by the next session. A one-standard-deviation increase corresponded to a score about 1.5 points lower on the BDI-II depression scale (which runs from 0 to 63). The link held after accounting for the therapeutic alliance (the quality of the working relationship). There was no difference between clients, though: those who on average received more questioning did not do better overall.

The authors list the limits themselves: without an experimental manipulation, causation can't be established; only early sessions were studied; the findings apply to cognitive therapy for depression; and the mechanism — whether questions change thoughts or simply increase engagement — was not tested documented fact; Braun et al., 2015.

Using the same sample, a later analysis looked at who benefits more. The link with symptom change was stronger in people who made especially gloomy predictions about the future and in those not taking antidepressants; the effects were small proven by research, a single study; Vittorio et al., 2021. A further paper by the same group, judging by its title, describes cognitive change as a mediator of this effect; we did not access its data (not verified; Vittorio et al., 2022, Behaviour Research and Therapy, 150, 104035 — only the bibliographic record was available).

Experiments outside the clinic

• In an analogue experiment (a model of therapy, not real therapy), 144 people watched videos of "sessions". The Socratic style was rated higher than a didactic (explaining) one: the questions seemed more helpful, people felt more supported in their autonomy and were more willing to do the tasks. Part of the effect was explained by alliance and empathy proven by research, an analogue study; Heiniger et al., 2018.
• In an online experiment with 69 participants, a 15-minute "Socratic" presentation changed a common irrational belief more than a control reading did, but not more than a standard explanation (psychoeducation) proven by research, a small analogue study; Harrison et al., 2018.

What follows

An honest summary: Socratic dialogue is an accepted CBT style with a plausible rationale and some encouraging early data, but it is not a proven "main mechanism" of therapy (interpretation; based on the sources above).

Thought records: the same dialogue on paper

A worksheet is a way to ask yourself the same questions when no therapist is in the room.

Padesky writes that most of the worksheets in "Mind Over Mood" guide clients through a Socratic dialogue with themselves the author's own account; Padesky, 2019. The thought record is the central one.

Where it came from

• 1979. Beck and colleagues' manual "Cognitive Therapy of Depression" introduces the Daily Record of Dysfunctional Thoughts documented fact; Beck et al., 1979. It is usually described as five columns: situation, emotions, automatic thoughts, rational response, outcome. We give this layout from secondary descriptions; we did not consult the book not verified.
• 1983. Padesky creates a seven-column thought record and replaces "rational response" with "alternative/balanced thought". According to Padesky and Greenberger, Beck later described "rational response" as a misstep because it risked making cognitive therapy look dogmatic the author's own account; Padesky & Greenberger, 2020, as cited by Psychology Tools.
• 1995 and 2016. The first and second editions of Greenberger and Padesky's self-help book "Mind Over Mood", in which the thought record is a core tool documented fact; Guilford Press.

What it is made of

Most versions include five core elements: the situation that triggered distress; named emotions; automatic thoughts; an adaptive alternative thought; and an assessment of the outcome, i.e. how much the distress changed clinical consensus; Erhardt et al., 2022.

The seven-column version spells this out in more detail. Below is the logic of the columns in our own words clinical consensus; based on Padesky & Greenberger as cited by Psychology Tools. The layout and text of published worksheets are copyrighted, so we don't reproduce them.

How seven columns differ from five: evidence for and against gets its own columns, and "rational response" becomes "balanced thought" clinical consensus; Psychology Tools, after Padesky & Greenberger, 2020. In the first weeks, short three-column forms (situation — emotion — thought) are often used: learn to notice first, then test clinical consensus.

Using the ratings

• Emotion strength is rated to give a baseline for measuring change clinical consensus; Psychology Tools. In CBT a 0–100 scale is typical.
• Belief in the new thought is rated from 0 to 100% clinical consensus; Psychology Tools.
• Order. Recent formats often name and rate emotions before writing automatic thoughts, since a strong feeling leads more easily to the hot thought clinical consensus; Psychology Tools.
• What to expect. When a thought record goes well, shifts of 10–50% are typical clinical consensus; Psychology Tools, citing Padesky & Greenberger, 2020, and Waltman et al., 2021. The goal is relief, not zero emotion.
• If very little changes, possible reasons include: the analysis was too superficial or too directive; the new thought wasn't credible enough; another distressing thought remains; or the original thought reflects a core belief that needs deeper work clinical consensus; Psychology Tools.

Apps: recording is not the same as processing

In a 2022 review, researchers found 246 apps by keyword search and narrowed them to 36 that fit the description of cognitive restructuring tools. Twenty-one of them had no step for writing an alternative thought: they only logged situations, emotions or thoughts. Of the 15 apps in the final analysis, 7 included all five core elements. The authors found no study directly testing whether these apps help proven by research, a review of apps; Erhardt et al., 2022.

The practical takeaway: if an app has no step where you test the thought and write a new one, it is a mood diary, not a thought record.

A completed thought record: "Let's discuss tomorrow at 10"

A composite example following the seven-column logic, with a breakdown and common mistakes.

Mira, 32, works in marketing (a composite example). She sent her manager the quarterly report. An hour later the reply came: "Let's discuss tomorrow at 10." No "thanks", no emoji.

After the record. Mira prepares three questions about the section she's unsure of. She also sets up a small experiment: at the start of the meeting she will ask what matters most to her manager in the report, and write down what he actually says. Before the meeting she records her prediction: "he'll say it's all bad" — belief 60%. This is how a thought record turns into a real-life test; behavioural experiments are covered in episode 6.

What was done well

• Emotions came before thoughts, each in one word and rated.
• The hot thought is not the first one but the most charged: about incompetence, not "he didn't like it".
• The "for" column was filled in honestly. Without it, the balanced thought would sound like self-persuasion.
• The "against" column holds facts, not reassurance. Not "it'll be fine" but "he meets about every report". The last item is the double standard (see below).
• The balanced thought allows for likely edits. "I'm great and the report is perfect" would be positive thinking, and fragile.
• The emotions dropped, but not to zero (anxiety 85 → 45). Some anxiety before a meeting with your manager is normal.
• It ends in action, not just a conclusion on paper.

Common mistakes when filling one in

Why rephrase questions: thoughts phrased as questions ("what if?") are hard to test against evidence, so they are first turned into statements clinical consensus; Psychology Tools, "What to Try When CBT Thought Records Don't Work".

CBT is not positive thinking

The aim is an accurate thought, not a pleasant one. And sometimes the accurate thought confirms that the problem is real.

When people hear "CBT works with thoughts", they often take it to mean "swap bad thoughts for good ones". It doesn't.

The affirmation experiment

In 2009 the psychologist Joanne Wood and colleagues ran a series of experiments with the phrase "I'm a lovable person". Among people with low self-esteem, repeating it — or thinking about why it was true — left them in a worse mood than skipping the statement or considering both how it was and wasn't true. Among people with high self-esteem, mood improved, but only a little. The authors concluded that positive self-statements may help some people but backfire for the very people who seem to need them most proven by research; Wood et al., 2009.

A caveat: these were lab experiments with non-clinical samples. They don't show that affirmations harm everyone. But they illustrate why CBT isn't built on positive phrases. Waltman makes the same point from clinical practice: purely positive reframes can be brittle if they don't fit the reality of the person's life the author's own account; Waltman, 2020.

What if the thought is true?

A balanced thought isn't always "softer". If the facts confirm a problem — the person really is facing injustice, a threat, discrimination or abuse — restructuring must not turn into "it's all in your head". The conversation moves to problem-solving: what can be done, who can help, what is needed for safety clinical consensus. Clumsy "challenging" risks sounding like "you're thinking wrong", which is invalidation: dismissing the person's experience clinical consensus; Waltman, 2020.

Tools for evaluating a thought

A short set of techniques a therapist chooses to fit the situation. Not all at once, and not all for working on your own.

The double standard

Most people judge themselves more harshly than they would a close friend in the same situation. The therapist makes the gap visible: "Would you say that to a friend?" the author's own account; Burns, 2018.

David Burns, author of "Feeling Good", says he first saw the technique in the 1970s, during his residency at the University of Pennsylvania, used by Dr Maultsby, and later adapted it the author's own account; Burns, 2018. So it is fair to say Burns popularised the technique rather than invented it.

Burns also has a paradoxical version: the therapist plays the client's friend and describes the client's problem as their own; the client responds with a compassion they don't give themselves; then together they check whether that response is true, and the client writes it down. Burns says it works when there is good empathy and the person really does hold a double standard, and rarely works for people who hate both themselves and other people the author's own account; Burns, 2018.

Alternative explanations

"How else could this be explained? If this happened to a friend, how would you explain it?" The person lists three to five possibilities and, if useful, estimates how likely each is clinical consensus.

Decatastrophising

"What if the worst happens?" The term is linked to Albert Ellis and his idea of catastrophising, and the technique was also developed within Beck's cognitive therapy not verified; based on a secondary source. The steps clinical consensus:

1. Name the worst-case scenario specifically.
2. Estimate how likely it is.
3. "If it happens, how would I cope? Who and what would help? What would it look like in a week, a month, a year?"
4. Name the best and the most likely scenarios.
5. Make a plan for the most likely one.

A limitation: in generalised anxiety disorder, working through the content of one worry often triggers the next, so therapy there focuses more on the worry process itself clinical consensus; Psychology Tools.

The downward arrow

A technique for finding the belief beneath an automatic thought clinical consensus; Neenan & Dryden:

1. Take the automatic thought.
2. Provisionally accept it as true. It isn't challenged along the way, or the underlying belief won't surface.
3. Ask: "If this were true, what would it mean for me? What's the worst part of it?"
4. Repeat until the answer becomes a general statement about yourself, others or the world.
5. Then work with each link.

The technique was first mentioned by Beck and colleagues (1979) and elaborated by Burns in "Feeling Good" (1980) (documented fact; Wenzel, 2012, in Standard and innovative strategies in cognitive behavior therapy, InTech). Burns calls his version the "Individual Downward Arrow Technique" and says he developed it the author's own account; Burns, Feeling Good blog, 2017. We did not check the book's text for the name "vertical arrow" not verified.

The responsibility pie

When someone blames themselves "one hundred percent", they are asked to list all the factors that affected the outcome: other people, circumstances, chance, information that wasn't available at the time. Each factor gets a slice of the circle, and their own share goes in last. Then they discuss what they see and re-rate their responsibility clinical consensus; Psychology Tools. The technique has roots in cognitive therapy for obsessive-compulsive disorder, where an inflated sense of responsibility is considered a key mechanism documented fact; van Oppen & Arntz, 1994. We could not establish who first devised the "pie" itself.

The goal is a realistic share of responsibility, not zero: legitimate responsibility isn't removed clinical consensus; Psychology Tools.

The continuum

Beliefs like "I'm a failure" live in a black-and-white world: either a failure or perfect. A continuum turns this into a scale, which makes room for doubt clinical consensus; James & Barton, 2004. The person names the ends of the scale themselves, often on a line about 10 cm long clinical consensus; James & Barton, 2004.

Padesky suggests several moves the author's own account; Padesky, 1994:

• rate yourself on a scale of the new, desired belief (for example, 0–100% "I'm lovable") rather than the old one, so any shift is easier to see;
• criteria continua: break a global idea ("a successful person") into specific qualities and rate each — often it turns out the person's rules never let them rise above zero;
• a positive data log: write down, every day, even small observations that support the new belief; this may take at least six months;
• a historical test: gather evidence for and against the belief across different periods of life.

The continuum and core-belief work are a step beyond working with automatic thoughts. They are done after case conceptualisation and when the person is stable clinical consensus; James & Barton, 2004.

Do we even need to challenge thoughts?

The most awkward question in the topic, and the answer may lie in what stays with you after therapy.

"I get it in my head, but I don't feel it"

A common complaint: it makes sense, but the anxiety is the same. The psychologist James Bennett-Levy asked trainee therapists to try thought records and behavioural experiments on themselves. They rated the experiments as more powerful and convincing for changing beliefs. His explanation draws on the Interacting Cognitive Subsystems model (Teasdale and Barnard): a thought record works more at the level of "I understand", action at the level of "I feel" proven by research, one small study of trainees; Bennett-Levy, 2003. Hence the practice of ending a thought record with "how can I test this in real life?" It has been reported that in an experiment with a non-clinical sample, a behavioural experiment changed belief immediately while a thought record did so only after a week, but we could not open the original paper not verified; McManus et al., 2012.

The 1996 component study

Neil Jacobson and colleagues randomly assigned 150 outpatients with major depression to three groups proven by research, RCT; Jacobson et al., 1996:

1. behavioural activation only — work on activities and daily schedule;
2. behavioural activation plus skills for modifying automatic thoughts;
3. full cognitive therapy, including work on core beliefs.

The full therapy was no better at the end of treatment or at six-month follow-up. Behavioural activation without thought work changed negative thinking as effectively as full cognitive therapy. One detail adds weight: all groups were treated by four experienced cognitive therapists, and the authors note a clear allegiance in favour of cognitive therapy documented fact; Jacobson et al., 1996. A limitation: this is a single trial in depression and shouldn't be generalised to every disorder.

A critical review

In 2007 Richard Longmore and Michael Worrell published a review with the blunt title "Do we need to challenge thoughts in cognitive behavior therapy?" Their conclusion: component studies offer little evidence that cognitive interventions add benefit, and there is little evidence that changing thoughts is what causes improvement (interpretation; Longmore & Worrell, 2007). This position has been contested, and the debate continues.

The case on the other side

Daniel Strunk and colleagues followed 35 people who had responded to cognitive therapy for depression. A lower risk of relapse over the next year was predicted by their cognitive therapy coping skills and by in-session signs that they were applying its principles on their own. This held after accounting for symptom severity at the end of treatment. General self-esteem did not predict relapse risk proven by research, small sample; Strunk et al., 2007.

The bottom line

CBT for depression works overall: in a meta-analysis of 409 trials (52,702 patients), the effect against control conditions was g = 0.79, while the difference from other bona fide psychotherapies was close to zero, g = 0.06 proven by research; Cuijpers et al., 2023. How much the thought work itself contributes is an open question (interpretation; Jacobson et al., 1996; Longmore & Worrell, 2007). Perhaps the value of cognitive techniques lies less in relieving symptoms faster and more in a skill that stays after therapy ends (interpretation; based on Strunk et al., 2007). That echoes Padesky's 1993 point: therapy teaches ways of finding solutions.

Homework: what happens between sessions

If a thought record is a dialogue without the therapist, it happens between sessions. Here the data are fairly consistent.

All three are meta-analyses proven by research; Kazantzis et al., 2000; Mausbach et al., 2010; Kazantzis et al., 2016. The coefficient r shows the strength of a relationship (0 = none, 1 = perfect); g shows the size of a difference (0.2 is usually considered small, 0.5 medium, 0.8 large). An important caveat: these are mostly correlations. It may be that people who are already improving do their homework better.

The practical point: what matters is not just "done or not done" but how it was done — whether the person grasped the process. That is why in CBT homework is reviewed at the start of the next session clinical consensus.

What good homework looks like clinical consensus: designed together; linked to the session's goal; specific — when, where, how much; started as a trial run in the session; obstacles discussed in advance; undone homework is data, not failure.

The limits of the method, and how to recognise a good CBT therapist

The thought record is a strong tool, but not a universal one, and not a substitute for help.

When the method needs adapting or adding to

• Cognitive load. A thought record requires writing, concentration and abstract thinking. With severe depression, with children or with cognitive impairment, it is simplified (three columns, spoken, by voice) or therapy starts with behavioural activation clinical consensus.
• Not every thought is testable. Questions ("what if?"), statements about feelings and truisms don't lend themselves to weighing evidence; they are rephrased first clinical consensus; Psychology Tools.
• Real problems. If the thought is true, what's needed is not restructuring but problem-solving, safety and validation clinical consensus.
• The head–heart gap. Behavioural experiments, imagery work and attention to the tone of the inner voice can help clinical consensus; Psychology Tools.
• Cultural context. Where people expect advice from a specialist ("just tell me what's right"), questions can feel like evasion. It helps to explain why the therapist asks rather than tells: a conclusion the person reached themselves usually holds better than one handed to them (interpretation; the author's observation, no sources gathered).
• Other approaches. Third-wave therapies (ACT, metacognitive therapy) suggest not testing the content of thoughts but changing one's relationship to them, or working with the worry process; they are covered in episode 13 clinical consensus.

How to tell whether a CBT therapist works in the spirit of the method

These pointers come from descriptions of the method and the competence framework clinical consensus; Padesky, 1993; UCL CORE:

• they ask more than they explain;
• they summarise in your words and ask you to correct them;
• they don't say "you're thinking wrong" and don't argue;
• they don't rush to a conclusion, and let you put the new thought into words yourself;
• they end with a question about the next step or a real-life test.

Warning signs are being argued with, lectured, or led to an answer decided in advance. The CBT competence framework explicitly says to avoid debate, persuasion, "lecturing" and "cross-examining", and to create doubt in place of certainty clinical consensus; UCL CORE.

Myths: a quick check

All the myths from these notes in one table.

Test yourself

1. How does "changing minds" differ from "guiding discovery"? In the first the therapist knows the answer and steers towards it; in the second they genuinely don't know, and the person draws the conclusion.
2. Name the four stages of Padesky's model. Informational questions, listening, summarising, synthesising question.
3. What did Braun and colleagues (2015) find, and why isn't it proof of cause? After sessions with more Socratic questioning, depressive symptoms fell more by the next session, even after accounting for alliance. But it was one observational study of 55 patients without an experimental manipulation.
4. How does a seven-column thought record differ from a five-column one? Evidence for and against the hot thought has separate columns, and "rational response" is replaced by "balanced thought".
5. Anxiety was 85 and after the thought record it's 50. Is that failure? No. A typical shift is 10–50%; the goal is relief, not zero.
6. Why isn't the thought challenged during the downward arrow? Otherwise the core belief won't surface.
7. What did Jacobson's component analysis (1996) show? Behavioural activation without thought work did as well as full cognitive therapy and changed negative thinking just as much.
8. What case for cognitive skills do Strunk's data (2007) make? People who learned the skills and used them on their own had a lower risk of relapse over the next year.

Key takeaways

• Socratic dialogue is curiosity, not argument: questions the person can answer, careful listening, summaries in their words, and a question that ties it all back to the original thought.
• A thought record is the same dialogue on paper. Name emotions before thoughts, fill in the "for" column honestly, and put facts rather than reassurance in the "against" column.
• A balanced thought is not a positive one. If the facts confirm a problem, the job is to solve the problem.
• A 10–50% drop in an emotion is already a result.
• The science is honest about it: CBT works overall, while the contribution of individual cognitive techniques is still being worked out. The strongest case for them is a skill that stays after therapy and is linked to a lower risk of relapse.
• A thought record ends in action: the next step is to test the conclusion in real life (episode 6).
• Heavy, daily thoughts, or thoughts of not wanting to live, are a reason to see a professional, not to reach for a worksheet.

Sources

1. Padesky, C. A. (1993). Socratic questioning: Changing minds or guiding discovery? Keynote address, European Congress of Behavioural and Cognitive Therapies, London, 24 September 1993. Author's publications page: https://www.padesky.com/clinical-corner/
2. 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, 18 July 2019. Author's publications page: https://www.padesky.com/clinical-corner/
3. Padesky, C. A. (1994). Schema change processes in cognitive therapy. Clinical Psychology and Psychotherapy, 1(5), 267–278.
4. Braun, J. D., Strunk, D. R., Sasso, K. E., & Cooper, A. A. (2015). Therapist use of Socratic questioning predicts session-to-session symptom change in cognitive therapy for depression. Behaviour Research and Therapy, 70, 32–37. https://pmc.ncbi.nlm.nih.gov/articles/PMC4449800
5. Vittorio, L. N., Braun, J. D., Cheavens, J. S., & Strunk, D. R. (2021). Cognitive bias and medication use moderate the relation of Socratic questioning and symptom change in cognitive behavioral therapy of depression. Cognitive Therapy and Research, 45(6). https://doi.org/10.1007/s10608-021-10224-6
6. Vittorio, L. N., Murphy, S. T., Braun, J. D., & Strunk, D. R. (2022). Using Socratic Questioning to promote cognitive change and achieve depressive symptom reduction: Evidence of cognitive change as a mediator. Behaviour Research and Therapy, 150, 104035. https://doi.org/10.1016/j.brat.2022.104035
7. Clark, G. I., & Egan, S. J. (2015). The Socratic Method in cognitive behavioural therapy: A narrative review. Cognitive Therapy and Research, 39(6), 863–879. https://doi.org/10.1007/s10608-015-9707-3
8. Clark, G., & Egan, S. (2018). Clarifying the role of the Socratic method in CBT: A survey of expert opinion. International Journal of Cognitive Therapy, 11(2), 184–199. https://doi.org/10.1007/s41811-018-0016-y
9. Heiniger, L., Clark, G., & Egan, S. (2018). Perceptions of Socratic and non-Socratic presentation of information in cognitive behaviour therapy. Journal of Behavior Therapy and Experimental Psychiatry, 58, 106–113. https://doi.org/10.1016/j.jbtep.2017.09.004
10. Harrison, L., Clark, G., Rock, A., & Egan, S. (2018). The impact of information presentation style on belief change: An experimental investigation of a Socratic Method analogue. Clinical Psychologist. https://espace.curtin.edu.au/handle/20.500.11937/72175
11. Kazantzis, N., Fairburn, C. G., Padesky, C. A., Reinecke, M., & Teesson, M. (2014). Unresolved issues regarding the research and practice of cognitive behavior therapy: The case of guided discovery using Socratic questioning. Behaviour Change, 31(1), 1–17. https://doi.org/10.1017/bec.2013.29
12. Waltman, S. H. (2020). Rethinking how we teach Socratic questioning. Psychotherapy Bulletin, 55(4), 18–21. https://societyforpsychotherapy.org/rethinking-how-we-teach-socratic-questioning-554/
13. UCL CORE. CBT Competences Framework: Guided Discovery and Socratic Questioning (Roth & Pilling). https://www.ucl.ac.uk/clinical-psychology/CORE/CBT_Competences/Specific_Competences/Guided_Discovery.pdf
14. Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive Therapy of Depression. Guilford Press.
15. Greenberger, D., & Padesky, C. A. (2016). Mind Over Mood (2nd ed.). Guilford Press. https://www.guilford.com/books/Mind-Over-Mood/Greenberger-Padesky/9781462520428
16. Psychology Tools. Thought Records in CBT: 10 Things Therapists Need to Know (summarising Padesky & Greenberger, 2020, The Clinician's Guide to CBT Using Mind Over Mood, 2nd ed.). https://www.psychologytools.com/articles/thought-records-in-cbt-10-things-therapists-need-to-know
17. Psychology Tools. What to Try When CBT Thought Records Don't Work. https://www.psychologytools.com/articles/what-to-try-when-the-cbt-thought-records-dont-work
18. Psychology Tools. Pie Chart — Responsibility. https://www.psychologytools.com/resource/pie-chart-responsibility
19. Waltman, S. H., Frankel, S. A., Hall, B. C., Williston, M. A., & Jager-Hyman, S. (2019). Review and analysis of thought records: Creating a coding system. Current Psychiatry Research and Reviews, 15(1), 11–19. https://doi.org/10.2174/1573400515666190123130725
20. Erhardt, D., Bunyi, J., Dodge-Rice, Z., Neary, M., & Schueller, S. M. (2022). Digitized thought records: A practitioner-focused review of cognitive restructuring apps. The Cognitive Behaviour Therapist, 15, e39. https://doi.org/10.1017/S1754470X22000320
21. Bennett-Levy, J. (2003). Mechanisms of change in cognitive therapy: The case of automatic thought records and behavioural experiments. Behavioural and Cognitive Psychotherapy, 31(3), 261–277. https://doi.org/10.1017/S1352465803003035
22. McManus, F., Van Doorn, K., & Yiend, J. (2012). Examining the effects of thought records and behavioral experiments in instigating belief change. Journal of Behavior Therapy and Experimental Psychiatry, 43(1), 540–547.
23. James, I. A., & Barton, S. (2004). Changing core beliefs with the continuum technique. Behavioural and Cognitive Psychotherapy, 32, 431–442. https://doi.org/10.1017/S1352465804001614
24. van Oppen, P., & Arntz, A. (1994). Cognitive therapy for obsessive-compulsive disorder. Behaviour Research and Therapy, 32(1), 79–87.
25. Burns, D. D. (2018, 16 April). 084: Paradoxical Double Standard Technique. Feeling Good Podcast. https://feelinggood.com/2018/04/16/084-paradoxical-double-standard-technique/
26. Neenan, M., & Dryden, W. Challenging thought: downward arrow procedure. Chapter in a Routledge book. https://www.taylorfrancis.com/chapters/mono/10.4324/9781315762470-91/challenging-thought-downward-arrow-procedure-michael-neenan-windy-dryden
27. Jacobson, N. S., Dobson, K. S., Truax, P. A., Addis, M. E., Koerner, K., Gollan, J. K., Gortner, E., & Prince, S. E. (1996). A component analysis of cognitive-behavioral treatment for depression. Journal of Consulting and Clinical Psychology, 64(2), 295–304. https://commons.clarku.edu/faculty_psychology/137
28. Longmore, R. J., & Worrell, M. (2007). Do we need to challenge thoughts in cognitive behavior therapy? Clinical Psychology Review, 27(2), 173–187.
29. Strunk, D. R., DeRubeis, R. J., Chiu, A. W. M., & Alvarez, J. (2007). Patients' competence in and performance of cognitive therapy skills: Relation to the reduction of relapse risk following treatment for depression. Journal of Consulting and Clinical Psychology, 75(4), 523–530. https://vivo.weill.cornell.edu/display/pubid17663607
30. Cuijpers, P., Miguel, C., Harrer, M., Plessen, C. Y., Ciharova, M., Ebert, D., & Karyotaki, E. (2023). Cognitive behavior therapy vs. control conditions, other psychotherapies, pharmacotherapies and combined treatment for depression: a comprehensive meta-analysis including 409 trials with 52,702 patients. World Psychiatry, 22(1), 105–115. https://doi.org/10.1002/wps.21069
31. Kazantzis, N., Deane, F. P., & Ronan, K. R. (2000). Homework assignments in cognitive and behavioral therapy: A meta-analysis. Clinical Psychology: Science and Practice, 7(2), 189–202. https://ro.uow.edu.au/hbspapers/2582
32. Mausbach, B. T., et al. (2010). The relationship between homework compliance and therapy outcomes: An updated meta-analysis. Cognitive Therapy and Research, 34(5), 429–438. https://pubmed.ncbi.nlm.nih.gov/20930925/
33. Kazantzis, N., Whittington, C., Zelencich, L., Kyrios, M., Norton, P. J., & Hofmann, S. G. (2016). Quantity and quality of homework compliance: A meta-analysis of relations with outcome in cognitive behavior therapy. Behavior Therapy, 47(5), 755–772. https://doi.org/10.1016/j.beth.2016.05.002
34. Wood, J. V., Perunovic, W. Q. E., & Lee, J. W. (2009). Positive self-statements: Power for some, peril for others. Psychological Science, 20, 860–866.

If you need help now

If your life is in danger, call your local emergency number: 112 in most of Europe, 911 in the US and Canada, 999 in the UK, 000 in Australia. For support lines by country, see Help now.

Image credits

This page uses no third-party images.