Using disagreement, “resistance,” and rupture as clinical information in CBT.
Collaboration is easiest to claim when the client agrees with you.
Your task is not to eliminate pushback. It is to decide what the pushback is telling you
about the goal, formulation, intervention, delivery, relationship, power, and your own stance.
Approx. 35–45 minutes
Premature interpretation
Client pushes back → “resistance”
Clinical inquiry
Client pushes back → what does this tell us?
Case orientation
Meet Jordan
Jordan is 31, Black, nonbinary, and works as a communications strategist for a nonprofit.
They sought CBT for anxiety, perfectionism, reassurance-seeking, and increasing work avoidance.
Important task
↓
“I have to get this right.”
↓
Anxiety
↓
Overprepare / recheck / delay
↓
Short-term relief
↓
Deadline pressure
↓
“See? I can't handle this.”
Last session, you assigned a monitoring worksheet.
Jordan
“I didn't do it.”
“And honestly, you keep making me do homework.”
“You keep making me do homework.”
Decision 1 · Homework noncompletion
What do you ask next?
What did the task become?
Jordan
“Like another thing I was supposed to do correctly.”
“I already have a boss.”
Therapist intended
Observe perfectionism.
Task became
Another performance demand.
Before treating noncompletion as a barrier, ask what the task became.
Decision 2 · Agreement vs compliance
Previous session:
Therapist: “Great. Try this three times before next week.”
Jordan: “Sure.”
How should you interpret the “sure”?
Clinical Support
Compliance is not the same thing as collaboration.
Compliance
“Sure.”
Collaboration
“I understand why we're doing this.”
“I think it could be useful.”
“I have concerns about the format.”
A client can complete every worksheet and still be minimally collaborative.
Decision 3 · Therapist reaction
Possible therapist data
Internal thought: “I'm trying to help you.”
Second thought: “You're the one who said you wanted CBT.”
DefensivenessUrge to justifyMore formal tone
What do you do with your reaction?
Therapist signal
Countertransference is data—not instructions.
Defensiveness
Urgency to persuade
Desire to prove intervention was reasonable
Fear of being wrong
Ask: Am I becoming more invested in defending CBT than understanding Jordan?
Decision 4 · Rupture named
Jordan
“You keep saying CBT is collaborative.”
“But sometimes it feels like you already know what we're supposed to discover.”
What do you do?
Collaborative empiricism
Predetermined
“What evidence do you have that you looked incompetent?”
Implicit endpoint: You didn't.
Genuine inquiry
“What exactly happened?”
“Which part fits the evidence?”
“Which part goes beyond it?”
A Socratic question is not collaborative if only one answer is acceptable.
Decision 5 · Cognitive disagreement
Jordan revisits a presentation in which their director interrupted twice to correct details.
Jordan
“I hated the alternative thought.”
“Because I think I did look incompetent.”
What should happen next?
Real error, broader inference
Jordan had not reviewed one section.
Director corrected something basic.
Another manager noticed.
Accurate
“I was underprepared for that section.”
Overgeneralized
“Everyone now sees me as professionally incapable.”
CBT becomes more credible when it can tolerate evidence that supports part of the client's thought.
Formulation revision
Jordan
“I want to make a mistake without losing the next four hours of my life.”
Old target
Stop believing mistakes mean incompetence
Revised target
Reduce the cognitive and behavioral aftermath of mistakes
Decision 6 · What is the real target?
What intervention now follows most directly?
Post-error maintenance cycle
Mistake
↓
Body heat
↓
Mental replay
↓
Check Slack
↓
Ask for reassurance
↓
Review notes / search reactions
↓
Temporary relief
↓
Threat remains active
A rupture can improve formulation.
Decision 7 · New homework conversation
Therapist: “I have an idea for what we could test between sessions.”
Jordan
“Here we go.”
What do you do?
Co-designed monitoring
1
What happened?
2
What did I do in the next ten minutes?
3
What was I trying to get from that behavior?
Therapist: “Would you actually use it?”
Jordan: “If it's in my phone, probably.”
Collaboration occurs at the level of design.
Decision 8 · Client-generated data
Event: wrong date in team email
Behavior: checked sent email five times, opened colleague chat,
typed “Did that look stupid?”, deleted it, searched director's earlier emails.
Function: “I wanted proof nobody thought differently about me.”
Jordan
“The checking didn't actually tell me anything.”
“It just kept the whole thing alive.”
What is the next move?
Client-owned inference
Therapist conclusion
“The checking maintains your anxiety.”
Jordan's conclusion
“The checking didn't tell me anything. It kept the whole thing alive.”
Let the client own the inference.
Decision 9 · “I don't want to stop checking”
Jordan
“Because sometimes checking catches something.”
“If you tell me to stop checking everything, I'm not doing it.”
How do you respond?
Functional distinction
Functional checking
Confirm recipient.
Check attachment.
Review key facts once.
Confirm accessibility requirements.
Reassurance-driven checking
Re-open sent email repeatedly.
Inspect reactions.
Ask whether you looked foolish.
Search for reputational evidence.
The treatment target is not carelessness.
Decision 10 · Ambivalence
Importance
8/10
Confidence
4/10
What are you doing with these questions?
Ambivalence
Wants change because
Time cost.
Mental exhaustion.
Work disruption.
Wants checking because
Feels protective.
Reduces uncertainty.
Sometimes catches real errors.
Motivated for what? Ambivalent about what?
Decision 11 · Therapist overcorrection
After the earlier rupture, you begin asking permission before nearly every intervention.
Jordan
“You don't have to ask me permission to speak.”
What happened?
Repair flexibility
Before rupture
Too much direction
Flexible collaboration
Structure + choice + challenge + feedback
After overcorrection
Too little direction
Repair should increase flexibility—not produce the opposite rigidity.
Decision 12 · Session feedback
Therapist: “What worked today?”
Jordan
“Mapping the checking.”
“I didn't love the motivation scale thing.”
What do you do?
Feedback changes treatment
Jordan
“It felt sales-y.”
“The confidence question helped, though.”
Keep
Self-efficacy inquiry
Revise
Motivational framing
Feedback only matters if it can change treatment.
Decision 13 · Identity and authority
Jordan
“I've had clinicians act like disagreement proves I'm avoidant.”
Jordan describes prior healthcare encounters in which assertiveness was documented as
aggressive, guarded, or noncompliant.
“So when you say ‘collaborative,’ I'm still waiting to see what happens when I say no.”
What should change in your formulation?
Power-aware collaboration
Therapist authority
↓
Client expects cost for disagreement
↓
Compliance
↓
Therapist misreads compliance as buy-in
Trust is tested behaviorally when the client says no.
Decision 14 · The therapist is wrong
Jordan explains that some of their checking is related to disability-accessibility requirements
for public-facing materials. Your formulation had grouped part of that checking into perfectionism.
What do you do?
Repair
Therapist: “I think I collapsed two different behaviors because they looked similar.”
“That's on me.”
Jordan
“Okay.”
Getting it wrong is not necessarily the rupture. What happens next may be.
Decision 15 · Exposure calibration
You propose delaying reopening one low-stakes sent email for thirty minutes.
Jordan
“Thirty is too much.”
“Ten.”
What do you do?
Exposure calibration
10 minutes
“I'll be uncomfortable, but I think I can redirect to another task.”
30 minutes
“I'll spend the whole time counting until I can check.”
Useful first experiment:
Return attention to another work task for ten minutes without reopening the sent message.
Collaboration does not mean the therapist has no position.
Decision 16 · “I failed your experiment”
Jordan
“I checked at four minutes.”
“There. I failed your experiment.”
What is the next question?
Experiment data
Minute 1
Strong urge to check
Minute 2
Opens another work document
Minute 3
Urge falls slightly
Minute 4
Director walks past → urge spikes → Jordan checks
The experiment produced data.
It did not produce ten minutes. Those are not the same thing.
Decision 17 · Therapy as live functional analysis
Jordan
“I think I know why I called it your experiment.”
“If it's yours, then I can fail you.”
What do you do?
Therapy becomes data
Therapist assigns
↓
Jordan predicts evaluation
↓
Tries to give the “right” response
↓
Avoids disagreement
↓
Feels pressure
↓
Homework / Socratic work becomes performance
The therapeutic relationship can become part of the functional analysis.
Decision 18 · Later disagreement
Jordan
“No. I don't think revisiting that presentation is useful anymore.”
You disagree. You think the event may still contain useful learning.
What should collaboration look like now?
Collaboration with disagreement
Therapist: “I think there may still be something useful there.”
“But I may be hanging onto it because it fits my formulation. Tell me why you think it's done.”
Jordan
“We already learned the thing.”
“I don't need another autopsy.”
Revised target:
Repair a mistake without spending two days hating myself.
A formulation that cannot be disagreed with is not collaborative.
Final integration
Jordan
“I don't feel like I have to agree with you for therapy to work.”
“And weirdly, I trust you more when you disagree with me now.”
What would you do next?
Pattern synthesis
Your clinical attention tended to foreground…
This is not a score. It describes what your decisions tended to foreground.
Relational climate across the case
More guarded
Completed relational map
Presenting problem perfectionism · work avoidance · post-error rumination
Relational insight “If it's your experiment, I can fail you.”
↓
Therapy becomes functional data
↓
Later disagreement held collaboratively
↓
Outcome disagreement no longer equals rupture
Core formulation
Not
Jordan is resistant to CBT.
Not
The therapist should simply follow Jordan's preferences.
Not
The relationship matters more than technique.
Jordan entered CBT with a perfectionistic pattern in which visible errors trigger anxiety,
rumination, checking, reassurance-seeking, and fears of being judged as incompetent.
That same pattern entered therapy: assignments became tests of being a good client,
Socratic questions were experienced as having predetermined correct answers, and disagreement
felt risky because Jordan had prior experiences of clinicians pathologizing dissent.
The therapist initially contributed by moving too quickly into structured tasks and by overgeneralizing
some functional checking as perfectionism. When Jordan pushed back, the most useful response was neither
abandoning CBT nor intensifying persuasion, but treating the disagreement as information.
That allowed the formulation to become more precise, the behavioral targets more functional,
and the therapy relationship itself to reveal the same performance dynamics treatment was trying to change.
Disagreement can improve formulation when the therapist is willing to be wrong.
Collaboration and Rupture Decision Rule
What to ask when the client pushes back
Goal, formulation, intervention, format, pace, therapist language, relationship?
Avoidance, lack of motivation, fear, defiance? Consider alternatives.
A nod is not always collaboration.
Learning exercise, demand, test, evidence of being a “good client”?
Do not ask the client to surrender it before understanding what it provides.
Keep the mechanism. Revise the delivery.
Include that in the formulation when it is functionally relevant.
Say so when appropriate.
Use the data before redesigning the task.
Map it functionally.
When the rationale remains coherent, the target is shared, and discomfort is serving the learning goal.
When the goal is not shared, the formulation no longer fits, the intervention is producing the wrong learning, or the therapist is defending the model more than observing the client.
Reference
Ambivalence
What does the client want from both changing and not changing?
Low readiness
Is the intervention ahead of willingness?
Poor treatment fit
Is the intervention targeting the right mechanism in a usable way?
Alliance rupture
Has something between therapist and client begun interfering with trust or engagement?
Avoidance
Is the client escaping contact with an experience that maintains the problem?
Boundary-setting
Is the client appropriately declining something they do not want?
Therapist defensiveness
Has the clinician become invested in being right, helpful, liked, or obeyed?
Structural mistrust
Does caution toward the therapist make sense within prior institutional experience?
Do not use “resistance” when a more precise formulation is available.
Transcript bridge 1
Pacing before challenge
Validate before challenge
Trust
Timing
Emotional pacing
Could the same cognitive intervention become helpful or unhelpful depending on when and how it is introduced?
Transcript bridge 2
Not rushing to reframe
Historical truth
Contextual pain
Premature optimism
Reframe timing
What happens when the therapist reaches interpretation before the client feels understood?
Transcript bridge 3
Power and positionality
Who defines success?
Who sets the pace?
Who interprets disagreement?
What does authority mean here?
How is therapist expertise used without making the formulation immune to challenge?
Transcript bridge 4
Countertransference and supervision
Feeling criticized
Helplessness
Frustration
Anger
Disengagement
Rescue
Racial / cultural impasse
Power / privilege activation
Which therapist reaction would make you most likely to misread disagreement?
Debrief
Reflect on pushback, rupture, and collaboration.
Why was “you keep making me do homework” more than nonadherence?
What did the task become?
How was “sure” different from agreement?
Instructor note: Look for the idea that homework itself entered the perfectionism cycle.
Why did it matter that Jordan really had made an error?
How did the rupture improve the formulation?
What became the new target?
Instructor note: The target shifts from proving competence to changing the aftermath of real or perceived mistakes.
What did defensiveness tell you?
How did the therapist later overcorrect?
Why can excessive permission-seeking become therapist-centered?
What made session feedback useful?
How did therapist acknowledgement improve the formulation?
Why is “I misunderstood” compatible with expertise?
Instructor note: Push learners to distinguish loss of certainty from loss of clinical authority.
How did prior institutional experiences shape compliance?
Why can disagreement be especially risky for some clients?
How should that change therapist behavior without becoming overcautious?
When should the therapist persist with discomfort?
When should they change course?
Why is “the client is avoiding” sometimes accurate but dangerous as a default explanation?
Instructor note: Emphasize falsifiability. If every no proves avoidance, the formulation cannot lose.
Hold the question
If every disagreement with the treatment can be interpreted as evidence of the client's pathology,
what evidence could ever show the therapist that they are wrong?
No response is required.
Continue with The CBT Dive
See what happens between the worksheet and the person completing it.
The transcripts extend this module into real clinical conversations where pacing, trust,
challenge, power, therapist reaction, and client disagreement shape whether structured CBT techniques can actually be used.
Transcripts, Notes and Reflections from The CBT Dive Video Podcast
