The target behavior is the headline. The chain is the story.
A client tells you, “I blew up again.” Your task is not to explain why.
It is to reconstruct how the behavior became increasingly likely,
what it accomplished, and where future change may become possible.
You’ll practice
Defining target behavior precisely
Separating prompting events from vulnerability factors
Slowing down apparently sudden behavior
Tracking thoughts, emotions, body sensations, urges and actions
Investigating function without assuming it
Identifying immediate and delayed consequences
Finding realistic intervention points
Using solution analysis without hindsight blame
Including relational and environmental context
Knowing when the chain is detailed enough
Approx. 30–40 minutes
?
↓
?
↓
?
↓
SEND
What happened before that?
And before that?
Case orientation
Meet Devon
Devon is 29 and works in nonprofit communications.
Three months ago, an eighteen-month relationship with Jules ended.
The breakup was not clean. Devon and Jules still occasionally text:
sometimes warmly, sometimes ambiguously, sometimes not at all.
Devon
“I did the thing again.”
“Sent a giant paragraph.”
“Actually four giant paragraphs.”
“And then I blocked them.”
“And then I unblocked them.”
“I don't know why I did it.”
Decision 1 · Define the target
Devon
“You can say it.”
“That this is ridiculous.”
What is the first task?
“What exactly happened first?”
“How many messages?”
“When did the blocking happen?”
Clinical Support
“I sent a message” was seven behaviors.
11:48
Message 1
11:52
Message 2
11:58
Check read receipt
12:06
Messages 3 + 4
12:11
Block Jules
12:26
Unblock Jules
12:28
Check online status
Summary
“I lose control with my ex.”
Target
Four escalating messages between 11:48 PM and 12:06 AM, followed by blocking, unblocking and checking online status.
Specificity creates analyzable behavior.
Decision 2 · Where does the chain begin?
Devon
“They posted a picture.”
“At a bar.”
“With somebody.”
Is the post the start of the chain?
Clinical Support
Prompting event is not the same thing as vulnerability.
Five hours sleep
No lunch
Work criticism
Cancelled plans
Two drinks
Alone at night
Recent breakup + stacked vulnerabilities
+
See Jules's post
↓
Less flexibility when the event appears
Vulnerability ≠ cause.
Decision 3 · Collect before correcting
The photo shows Jules at a bar with another person's arm around them.
Devon
“I immediately knew they were dating.”
What do you map next?
Clinical Support
Reconstruction mode and evaluation mode are different tasks.
Reconstruction mode
“What went through your mind?”
Evaluation mode
“What evidence supports that?”
Both can be useful. During chain reconstruction, sequence comes first.
Decision 4 · Separate internal links
Devon
“I was angry.”
“Really angry.”
What do you do?
Clinical Support
Separate links when doing so adds information.
Thought “They never cared.”
↓
Emotion Humiliation · anger · fear · sadness
↓
Body Heat · shaking · chest tension
↓
Urge Make them answer
↓
Action Open text thread
Categories help observation. They are not rules about how experience must unfold.
Scene · The first action was not sending
The “impulsive” behavior took 34 minutes to develop.
11:14
See post
11:16
Check Jules's profile
11:18
Search tagged photos
11:22
Open old text thread
11:27
Find message: “I'm not trying to date anyone right now.”
11:31
Screenshot message
11:34
Open Notes app
11:36
Draft “Interesting how quickly things change.”
11:37
Delete draft
11:39
Check profile again
11:43
Start longer message
11:48
Send
Devon
“Okay, that's horrifying.”
Decision 5 · Hindsight
How do you respond?
Clinical Support
More links do not mean more blame.
EarlierLater
Working hypothesis: as the chain progresses, activation, cognitive narrowing and investment in the interpretation may increase.
Intervention availability may therefore change across the chain.
Decision 6 · What did checking do?
Devon
“I wanted proof.”
“Maybe proof I was wrong.”
“Or proof I was right.”
How do you conceptualize the checking?
Clinical Support
One behavior. Several possible functions.
CHECK PROFILE
Reduce uncertainty
Seek reassurance
Seek confirmation
Feel active
Maintain connection
Prepare confrontation
Function is a hypothesis to test collaboratively.
Decision 7 · Chains can loop
11:48 PM
“Interesting that you needed ‘space’ for three months and suddenly you're wrapped around someone at a bar.”
No response.
11:52 PM
“Actually don't answer. I get it.”
What happens next in the chain?
Clinical Support
Behavior changes the environment.
Send message
↓
No response
↓
“They're ignoring me.”
↓
Check phone
↓
More distress
↓
Send again
Consequences can become the next prompting event.
Decision 8 · Block / unblock
Devon
“Blocking made me feel like I was the one ending it.”
“For about ten minutes.”
“Then I unblocked because what if they replied?”
What is the function of blocking?
Clinical Support
Opposing behaviors can solve different momentary problems.
Block
Temporary control
Less exposure to uncertainty
Reversal of who rejects whom
Unblock
Restore possibility of contact
Resume monitoring
Preserve possibility of reply
Ask: “What problem was each behavior trying to solve at that moment?”
Decision 9 · Consequences
Devon
“Right after I sent the fourth message, I felt relieved.”
“They knew.”
“I felt powerful for maybe five minutes.”
Which consequences matter?
Clinical Support
What did the behavior solve for five minutes?
Now
Relief
Action
Expression
Temporary control
Later
Shame
Sleep loss
Missed work
More checking
Reassurance-seeking
Completed chain · Version 1
Vulnerability Poor sleep · no lunch · work criticism · alcohol · isolation · recent breakup
↓
Event Jules posts ambiguous photo
↓
Thought “I was disposable.”
↓
Emotion / body Humiliation · anger · fear · heat · shaking
↓
Urge Make them answer
↓
Behavior Investigate → draft → send
↓
Event No response
↓
Behavior Send again → block → unblock → check
Immediate consequences
Relief · action · expression · temporary control
Delayed consequences
Shame · sleep loss · work disruption · more monitoring
Chain Analysis → Solution Analysis
Understanding the sequence is not the endpoint.
Where could the sequence change?
Decision 10 · Where should intervention begin?
Devon
“Okay. What was I supposed to do instead?”
Clinical Support
Do not place the entire treatment at the end of the chain.
○ Vulnerability
↓
○ Environment
↓
○ Interpretation / early activation
↓
○ Checking loop
↓
○ Drafting
↓
○ Sending
↓
○ Post-message escalation
Possible does not mean necessary.
Decision 11 · Environment vs willpower
Devon
“Muting Jules feels childish.”
“Like I should be strong enough to see things and just handle it.”
How do you respond?
Clinical Support
The same environmental change can serve different functions.
Could function as avoidance
“I cannot tolerate knowing anything about Jules ever again.”
Could function as stimulus control / boundary
“For two weeks, I want fewer involuntary cues while I practice responding differently.”
The behavior alone does not reveal its function.
Solution analysis
Eat dinner before drinking.
Mute Jules for two weeks.
Label: “I am moving from uncertainty into investigation.”
One profile check maximum.
Thirty-minute Notes delay for long messages.
Devon
“This looks like I'm being managed by airport security.”
Decision 12 · Skill dumping
What do you do?
Selected solution plan
Selected
Mute Jules for two weeks
Purpose: reduce unplanned exposure during the acute breakup period.
Selected
Thirty-minute Notes delay
Purpose: create time between urge and contact.
Not selected
One-profile-check limit
Devon predicts they will not follow it reliably.
Treatment planning also includes deciding which plausible strategies not to use.
Nine days later
Friend
“Uh, just saw Jules at Mira.”
Friend
“They're definitely with someone.”
Devon feels the same hot rush and opens Jules's text thread.
They write six sentences.
Draft message
↓
Copy into Notes
↓
30 minutes
↓
Delete
Devon
“So I succeeded.”
“Except then I spent forty minutes interrogating my friend.”
Decision 13 · Behavior migration
How do you evaluate this?
Clinical Support
Behaviors can migrate.
Old path
Uncertainty ↓ Check Jules ↓ Message Jules
New path
Uncertainty ↓ Question friend ↓ Seek details
Same function? Maybe. The next task is to test that hypothesis rather than assume it.
Decision 14 · What intervention is emerging?
Devon
“I wanted details.”
“Who? How close were they standing? Did they leave together?”
“I wanted certainty.”
What intervention is emerging?
Clinical Support
Chain analysis identifies intervention points. It does not dictate one intervention.
Cognitive work
Meaning attributed to Jules's behavior
Exposure
Allow uncertainty without investigation
More chain work
Clarify function before choosing
Context complication
The chain is not only inside Devon.
Instagram
Mutual friends
Shared committee
Work stress
Alcohol availability
Phone access
Social expectations
Unavoidable contact
Decision 15 · Social environment
Devon and Jules are still on the same volunteer committee.
Several mutual friends also send updates about Jules without being asked.
Devon
“Everyone thinks they're helping.”
“But people keep giving me updates.”
What changes?
Clinical Support
A behavioral chain can cross intervention categories.
Environmental modification
Ask friends not to send unsolicited updates.
Assertiveness
Communicate that request.
Exposure planning
Prepare for unavoidable committee contact.
Cognitive work
Address what seeing Jules means to Devon.
Decision 16 · When to stop analyzing
Devon
“I can tell you what song was playing.”
“I remember which side of the couch I was sitting on.”
“The lamp was on—”
How much detail is enough?
Clinical Support
Precision is not exhaustiveness.
Continue asking when detail…
Identifies a meaningful link
Changes the functional hypothesis
Reveals a vulnerability
Clarifies consequence
Identifies an intervention point
Consider stopping when detail…
Does not change formulation
Does not change treatment planning
Becomes repetitive reconstruction
Begins to feel like evidentiary interrogation
Final integration · Ten minutes remain
Devon
“I think what surprises me is how early it started.”
“I keep thinking the problem was pressing send.”
“But I was already halfway down the hill before I opened the text box.”
Where would you go next?
Pattern synthesis
Your clinical attention tended to foreground…
This is not a clinical competence score.
It describes what your decisions tended to emphasize in this simulation.
Interpretation “They're dating.” · “I was disposable.”
↓
Activation Humiliation · anger · fear · heat · shaking
↓
Investigation Profile · photos · old texts · screenshot
↓
Contact Draft · send · no response · send again
↓
Regulation behavior Block · unblock · check
Immediate
Relief · action · expression · temporary control
Delayed
Shame · sleep loss · work disruption · more checking · reassurance-seeking
Working formulation:
Under stacked vulnerability and attachment threat, uncertainty initiates an investigation-and-contact sequence that provides brief action,
certainty-seeking, expression and control while producing longer-term shame and relational destabilization.
This formulation creates treatment options. It does not dictate one.
Chain Analysis Decision Rule
How to know what to ask next
What exactly happened? Make it observable where possible and bounded in time.
What made this moment harder to navigate than another moment?
What changed immediately before the chain accelerated?
Then what? Then what? Then what?
Thought, emotion, body sensation, urge, behavior and environment may be worth separating when the distinction adds information.
What happened immediately? What happened later? What changed for the client, other people or the environment?
What changed because of the behavior? What problem did it solve briefly? Keep the hypothesis tentative.
Early, middle, late, environmental, relational and internal intervention points may all matter.
Do not turn every link into homework. Select interventions based on impact, feasibility, fit and likely barriers.
When the behavior changes, ask whether the underlying function changed—or migrated elsewhere.
Adjacent interventions
Chain analysis
How did the behavior become increasingly likely?
Thought record
What interpretation occurred, and how well does it fit the evidence?
Behavioral experiment
What action could test a prediction?
Exposure
What can be learned by approaching uncertainty without the usual protective response?
Problem-solving
What real-world obstacle needs a workable plan?
Self-monitoring
What pattern occurs repeatedly across time and contexts?
Debrief
Reflect on the chain.
Reconstructing
Why define the target precisely?
Why distinguish prompting event from vulnerability?
Why record a thought before challenging it?
When did the apparently impulsive behavior actually begin?
Function
What did checking accomplish?
Why did blocking and unblocking both make sense in sequence?
Which immediate consequences may reinforce the pattern?
What did behavior migration teach you?
Intervention
Why is “don't send” a weak standalone plan?
When is environmental modification appropriate?
How do you prioritize among many solution points?
When should another CBT intervention take over?
Power / ethics / stance
When can detailed analysis become blaming?
How does hindsight distort judgments about available choice?
Which parts of Devon's chain occurred outside Devon?
How much detail is enough?
Take this question with you
If a chain analysis gives us twenty places where a client could theoretically have acted differently,
how do we prevent therapy from turning hindsight into blame?
Continue with The CBT Dive
See CBT slow down the moment between “trigger” and “behavior.”
The transcripts extend these teaching modules into longer clinical conversations where formulation has to happen in real time:
incomplete information, unexpected answers, relational reactions, uncertainty and all.
Transcripts, Notes and Reflections from The CBT Dive Video Podcast
