Andre Ivanoff Describes Chain Analysis For
Identifying the Causes of Problem Behaviors
By Lynn Mollick & Milton Spett
On Sunday January 11, forty NJ-ACT members braved icy driveways and slippery streets to attend Andre Ivanoff’s workshop on using chain analysis to address difficult problems and stuck treatments.
Developed by Marsha Linehan, chain analysis is a technique for uncovering the variables that control behavior, and then using this information to plan treatment interventions. It is similar to functional analysis, but looks beyond observable antecedents and consequences to include thoughts, feelings, and emotions, as well as cultural and contextual influences. Chain analysis is especially helpful in treating episodic problem behaviors such as cutting, binge eating or drinking, temper outbursts, and suicide attempts.
Begin a chain analysis by clearly identifying a specific instance of the problem behavior. Ask the patient to tell you about the incident “like it is a movie.” You can let patients tell their stories uninterrupted, or ask questions to clarify details. A thorough chain analysis consists of identifying the following components:
1. Trait vulnerabilities. Personality traits that increase the likelihood that the problem behavior will occur (e.g. a tendency toward depression or a history of abuse).
2. State vulnerabilities. Temporary states of the patient that increase the likelihood that the problem behavior will occur (e.g. being hungry, tired, or pre-menstrual).
3. Protective factors. Patient characteristics that decrease the likelihood that the problem behavior will occur (e.g. intelligence, familiarity with the surroundings).
4. Context. The external situation in which the problem behavior occurred, including cultural expectations and values.
5. Cue. The specific external event that begins the chain of events leading to the problem behavior.
6. Links. Specific steps between the cue and the problem behavior. There are 5 kinds of links: actions, thoughts, emotions, physical sensations and behaviors, and additional external events.
7. The problem behavior. A specific incident.
8. External & internal outcomes. Consequences -- emotions, cognitions, or external factors that change as a result of the problem behavior. Outcomes can be positive or negative.
Use SUDS (subjective unit of distress) to assess the importance or intensity of each chain component (maximum distress = 100; no distress = 0). When did the situation go wrong? Which links were protective and which exerted a negative effect? Work with patients to understand the external and internal outcomes, the functions of the chain. What problems does it solve? What is avoided as a result of the chain?
Getting the Pertinent Details
Doing a chain analysis requires accurate observation of many details. If the patient’s recall is poor:
1) Obtain information from other observers;
2) Initially ask for events, and fill in thoughts, feelings, and physiological responses later;
3) Re-enact the chain either in the environment in which it occurred or by role playing; or
4) Teach mindfulness or distress tolerance to improve recall.
Good chains inform treatment. But be careful not to intervene too quickly. Do chains on several incidents that lead to the same or different outcomes. Look for similarities and differences in the cues, links and consequences. Ask patients which links and outcomes they think are the most powerful. Speculate with patients about possible consequences of paths not taken. Generate hypotheses about what changes in the chain would lead to a better outcome. Then implement these changes and do another chain analysis.
An Example of Chain Analysis
Dr. Ivanoff illustrated chain analysis with the following example -- a man on probation we’ll call “Morris.” His problem behavior was threats of aggression.
1. Trait vulnerabilities. History of aggressive behavior, belief that aggression solves problems, high sensitivity and reactivity to emotions, poor emotion regulation skills.
2. State vulnerabilities. Drinking, feeling anxious and irritable, belief that “my wife has been trying to pick a fight.”
3. Protective factors. Knowledge that his wife will call his probation officer if he threatens her, worries about what will happen if he returns to jail.
4. Context. Morris had invited three friends to his home to watch a football game on TV. They were drinking beer. Morris’s wife walked by and kicked his feet off the coffee table. His friends laughed. Morris felt mildly embarrassed (SUDS = 15). He muttered “bitch – always bossing me around” and experienced increased anger (SUDS = 50). His wife told him to come into the kitchen. He did, made a perfunctory apology, and returned to the game with his friends. He felt relieved (SUDS = 5).
5. Cue. His wife came out of the kitchen and said, “Get back in here and apologize to me the right way.”
6. Links. The three friends became silent and stopped watching the game. Morris felt very embarrassed (SUDS = 80). Morris thought “She really wants to pick a fight. She never lets me have any fun. She’s such a bitch” (SUDS = 90). “I should slap her. She deserves it. That will shut her up” (SUDS = 20). “If I slap her, she’ll call my probation officer and I’ll be in more trouble. I’m stuck” (SUDS = 50).
His wife repeated her command “I mean it, get in here now” (SUDS = 85). Morris went into kitchen and sat at the table with his hands over his ears, breathing deeply, trying to calm down (SUDS = 40). Morris heard his wife pick up the phone, thought she was calling his probation officer, and felt afraid (SUDS = 80). Morris experienced sweating, stomach distress and racing thoughts: “She’s ruined my evening. She always gets her way. She treats me like dirt. This is going to cause me a shitload of trouble. I might be back in the can” (thoughts, SUDS = 100).
7. The problem behavior. Morris yelled “Put that phone down, or I’ll hit you so hard you’ll wish you had.”
8. External & internal outcomes. She hung up the phone. His anger diminished and his racing thoughts stopped (SUDS = 15). She went into the bedroom, crying. He thought “good, she learned a lesson.“ His anger diminished further (SUDS = 5). He returned to his friends and chugged a beer. He noted looks of approval from his friends (SUDS = 0). His wife came out of the bedroom and calmly said “I’d like to talk.” He felt even better.
Other outcomes included satisfaction over stopping his wife from calling the probation officer, getting his wife to treat him nicely, relief from anger, relief from embarrassment, restoring “cred,” avoiding returning to jail.
Interventions Suggested by the Chain Analysis
Changing this patient’s threats may be very difficult because, as the chain reveals, the behavior has many positive outcomes, including Morris’s friends’ approval of his aggressive behavior. However, if Morris wants to stop threatening, many interventions are possible:
1. Emotion regulation training for Morris.
2. Getting Morris and his wife to agree that she will always call his probation officer when he makes threats,
3. Teaching Morris’s wife to ask for what she wants appropriately instead of aggressively kicking his legs off the coffee table.
4. Teaching Morris to be assertive instead of aggressive when his wife treats him inappropriately.
All rights reserved.