Appearance
15.3 — CBT, Self-Applied
The model
Something happens. Between the event and your feeling, there is an interpretation (4.1) — and it is so fast that it feels like part of the event.
The interpretation is where the leverage is, because the event is frequently unchangeable and the feeling cannot be changed directly.
Four components, all connected: the situation, the thought, the feeling, and the behaviour. Change any one and the others move.
The thought record
The core technique. Seven columns, done in writing — and the writing is not optional, because the whole point is to get the thought outside your head where it can be examined.
1. Situation. What happened, in camera language (10.9). "Sent the report at 4. No reply by 7."
2. Feeling, and intensity out of 100. One word (4.3). "Anxious, 75."
3. Automatic thought. The exact sentence that went through your mind. Verbatim, not a summary. "He hated it and he's deciding what to do about me."
Then mark the hot thought — the one carrying most of the feeling. There is usually one.
4. Evidence for it. Facts only, not feelings. "He usually replies within an hour."
5. Evidence against it. This is the column you will not want to fill in, and it is the one that does the work. Force at least three.
"He was in meetings all afternoon.""He said last week the work was good.""Twice before I thought this and it was nothing.""He's slow with everybody, not just me."
6. A balanced alternative. Not a positive thought — a more accurate one, that accounts for all the evidence. "I don't know why he hasn't replied. It is most likely he is busy. If there is a problem I will hear tomorrow and I will deal with it then."
7. Feeling again, out of 100. "Anxious, 40."
It will rarely go to zero. A drop of twenty or thirty points is a good result, and it is enough to change what you do next.
The distortions
Named patterns that make thoughts inaccurate. Learning to spot which one you are running is faster than testing every thought from scratch.
All-or-nothing. No middle ground. If it isn't perfect it's a failure.
Overgeneralisation. One event becomes a rule. Always. Never.
Mental filter. Nine good comments and one criticism, and only the criticism is retained.
Discounting the positive. That doesn't count, anybody could have done it.
Mind reading. Assuming you know what they think (10.1).
Fortune telling. Predicting the worst as though it were established.
Catastrophising. Following the chain to disaster.
Emotional reasoning. I feel like a fraud, therefore I am one. The most common and the most convincing.
Should statements (5.1). Rigid rules, applied to yourself or others.
Labelling. I'm an idiot rather than I made a mistake (4.8).
And personalisation. Taking responsibility for things outside your control (5.8).
Most people run two or three of these repeatedly. Find yours and you have found most of your pattern.
The questions that test a thought
When you cannot find evidence against it, use these.
What would I say to a friend who told me this about themselves? The most reliable of the set (1.7).
What is the evidence, as a camera would record it?
What is the most likely outcome, not the worst?
Have I thought this before? What actually happened?
Will this matter in a year?
What would somebody who disagreed with me say?
And: even if the worst happened, what would I actually do? This last one is important. Anxiety avoids the answer, and answering it usually reveals a survivable and manageable situation (6.9).
The behavioural half, which is the stronger half
CBT is not only about thoughts, and the behavioural side frequently does more.
Behavioural experiments. Turn the thought into a prediction and test it. "If I say I disagree in the meeting, people will think I am difficult." Then do it, and write down what actually happened. One real test beats twenty thought records, because it supplies evidence rather than argument (2.6).
Activity scheduling (8.8). For low mood: plan the activity, do it regardless of feeling, rate mood before and after. The rating matters — people with depression consistently predict they will not enjoy things, and the record shows otherwise.
Exposure (14.3). For anxiety: graded approach, stay until it drops, no safety behaviours.
And dropping safety behaviours (6.9), which is the single highest-value behavioural intervention for anxiety.
The deeper layer
Underneath automatic thoughts sit core beliefs — global statements about yourself, others or the world. I am not good enough. People let you down. I have to be perfect to be acceptable.
Find yours with the downward arrow. Take an automatic thought and ask repeatedly: and if that were true, what would it mean?
He hasn't replied. → He's unhappy with my work. → He'll think I'm not up to it. → I'll be found out. → "I'm not actually good enough and everybody will eventually see it."
That is the core belief, and it is usually old (13.10).
They change slowly, through accumulated evidence rather than through argument (8.7). The practical technique is a running evidence log — a note where you record, daily, anything that contradicts the belief, however small. Weeks of small entries do what no single insight does.
Making it work
Do it in writing. Held in your head it becomes rumination (5.5).
Do it soon after, not weeks later.
Aim for accuracy, not positivity. CBT is not about thinking positively. If the thought is true — you did handle that badly — then the work is problem-solving, not disputation (4.8).
And expect it to feel artificial for a few weeks, then to become faster, then eventually to happen without writing.
What to do
Do one full thought record this week, on paper, seven columns.
Identify your two commonest distortions from the list.
Run one behavioural experiment — a real prediction, tested, with the result written down. This is worth more than the rest of the page.
And start the evidence log against your core belief if you have found it.
Next: 15.4 covers the approach for the thoughts that are true and the situations that cannot be changed.