Aggression Replacement Training

1. Introduction

Version followed: Glick & Gibbs (2011), Aggression Replacement Training: A Comprehensive Intervention for Aggressive Youth, 3rd edition (Research Press), as described by the publisher, the federal Title IV-E Prevention Services Clearinghouse and a state implementation FAQ. Earlier versions are Goldstein & Glick (1987) and Goldstein, Glick & Gibbs (1998, revised edition). The manual itself is not drawn on, so session-level detail comes from published descriptions and trainer slides [uncertain where flagged].

Aggression Replacement Training (ART) is a manualised, closed, co-facilitated group programme, developed by Arnold Goldstein and Barry Glick in the USA in the 1980s for aggressive adolescents (mainly ages 11 to 18, chiefly in juvenile justice, schools and residential care). It is a structured programme with three linked curricula rather than a single therapy. It treats aggression as "simultaneously a behavioral, cognitive, and emotional phenomenon" [source: Research Press product page], so it trains three things in parallel:

• Social Skills Training (behavioural: what to do instead). Uses Goldstein's Skillstreaming method.
• Anger Control Training (affective: what not to do, and how to calm arousal). A seven-link anger control chain.
• Moral Reasoning (cognitive: why to choose prosocial behaviour). Gibbs's social decision-making meetings.

How change happens: young people with aggressive histories are seen as lacking prosocial skills, being quick to anger and reasoning at an immature, self-centred moral level. Change comes from (a) learning concrete replacement skills through modelling, role-play, feedback and real-world homework, (b) learning to spot and lower anger before acting, and (c) peer discussion that moves reasoning toward fairness and taking other people's perspective. Peers are the main vehicle of learning [source: Wikipedia; Research Press; Amendola & Oliver slides].

2. Framework

Format [source: ACF Clearinghouse; Georgia CJCC FAQ; Research Press]

• 30 sessions over 10 weeks: three sessions a week, one per component, so each component gets 10 sessions. Sessions usually last about 1 hour (the Clearinghouse gives a range of 45 to 90 minutes).
• Group of 8 to 12 young people; 8 to 10 is preferred, to allow for drop-out.
• Closed group, because skills build week on week. Run by two trained facilitators, the same pair throughout.
• Room: U-shaped seating, a flip chart or whiteboard, skill cards and posters, a folder for each participant.
• Alternative schedule (Georgia): two 2-hour sessions a week. Day 1 is Social Skills, then the first half of the Moral Reasoning meeting. Day 2 is the second half of the same Moral Reasoning problem situation, then Anger Control. If you use this schedule, keep that sequence.
• [uncertain] Georgia's FAQ says "25 sessions" in one place and 30 hours elsewhere. Follow the 30-session structure in the 3rd edition.

Phases

[source: Research Press table of contents; Georgia CJCC curriculum table]

Weekly moral reasoning problem situations [source: Research Press; Georgia]: Jim (or Emilio), Jerry (or Latoya), Mark (or Ishan), George (or Enzio), Sam (or Carmen), Leon (or Cheri), Reggie (or Big Bear), Alonzo (or Tara and Lashonda), Juan (or Lin), Antonio (or Emma). The 3rd edition offers alternate situations in Appendix D. Their content is not reproduced here.

Within-session progress criteria. Session plans are fixed, but each component has an internal rule: in Social Skills and Anger Control, every youth should get at least one role-play turn over time and a homework assignment. Missed sessions must be made up before the next session [source: Georgia FAQ]. [synthesis] Before a youth's turn ends, check that they did every skill step (or chain link) in the role-play, and repeat the role-play if they did not.

Facilitators and fidelity. Two facilitators trained in the model. The Clearinghouse lists a minimum of a high school diploma. Accreditation involves a 5-day training covering all three components, and UCCI offers a 3-day end-user training. There are three tiers: group facilitator, trainer of facilitators, master trainer. The manual's Appendix B contains session evaluation checklists for fidelity. Washington found that 6 of its 20 courts were not delivering the model competently, and those courts did not reduce reoffending [source: ACF; UCCI; Georgia; WSIPP].

3. Guiding principles

• Replace, don't just suppress. For every aggressive behaviour you target, teach and practise a prosocial alternative [source: Research Press; Wikipedia].
• Three channels at once. Each week covers behaviour, affect and cognition. Do not drop a component to "catch up" on another [source: Research Press].
• Learn by doing. Skills are taught through modelling, role-play, performance feedback and transfer training, not lectures. Keep talk short and get youth on their feet [source: ACF; Amendola & Oliver].
• Peers are the main teachers. Peer feedback and group consensus carry more weight than adult instruction, so seat and structure the group for peer exchange [source: Wikipedia; Campbell slides].
• Personal relevance. Every role-play is built on a real, recent or upcoming situation from the youth's own life, not a made-up one [source: Wikipedia; Amendola & Oliver].
• Transfer is the goal. Homework, Hassle Logs and caregiver home notes carry skills into home, school and unit life. Collaborate with parents, teachers and staff whenever possible [source: Georgia; ACF].
• Structure and fidelity. Closed group, fixed sequence, consistent co-facilitators, a group size cap, required make-up sessions. Poor delivery cancels the benefit [source: Georgia; WSIPP].
• Firm but not demeaning. Facilitators apply consequences and keep the agenda without humiliating youth [source: Georgia FAQ].
• Facilitate moral discussion; don't preach. In Moral Reasoning the facilitator questions and steers, letting more mature peers challenge less mature ones [source: Amendola & Oliver; Campbell].
• Don't reinforce immature reasoning. Record it, but praise only mature reasons. Reshape immature answers into mature form wherever you can [source: Campbell slides].
• Benign confrontation. Challenge contradictions through questions, in a way that does not escalate the youth or make them withdraw [source: Amendola & Oliver].
• Cultural sensitivity and differentiation. Adapt delivery to participants' communication styles and accessibility, and to cultural factors; avoid stereotyping [source: ACF; Georgia].
• Target the right youth. Deliver ART to youth at moderate to high risk with real deficits in skills, anger control or moral reasoning [source: WSIPP; Georgia]. [synthesis] Do not dilute the group with low-risk youth.
• Prepared facilitators. Before each session, review the lesson, prepare flip charts and handouts, and rehearse modelling displays with your co-facilitator [source: Georgia FAQ].

4. Intervention inventory

All example language is illustrative, written for this guide; none of it is manual text. The manual's own scripts, skill cards, Hassle Log form and problem-situation sheets are copyrighted. Use the originals from the 3rd edition and the publisher's download page.

A. Programme set-up

A1. Pre-group individual orientation

• Purpose: engage each youth before the closed group starts; explain the purpose, procedures and rules; screen fit [source: ACF Clearinghouse].
• Steps:
  1. Meet the youth once, individually, with one of the group's facilitators.
  2. Explain the three components in plain terms and what a session looks like (role-plays, logs, discussions).
  3. Go through the group rules and the attendance and make-up policy.
  4. [synthesis] Ask for one or two recent incidents where anger cost them something. These become the youth's first role-play material and reasons to take part.
  5. [synthesis] Note reading level, language and any trauma, cognitive or attention needs so you can adapt handouts and role-plays.
• Example language (illustrative): "This group is about getting what you want without it costing you: fewer write-ups, fewer fights, more say in what happens to you. We practise, we don't lecture."
• Watch for: refusal or flat compliance. [synthesis] Link the programme to the youth's own goals (release, getting back to school, keeping a placement), not to the court's goals.

A2. Group rules and session routine

• Purpose: create a predictable, safe structure for role-play and discussion [source: Amendola & Oliver; Campbell slides].
• Steps: In the first session of each component, introduce the goals and the rules for taking part. In Moral Reasoning, go over the discussion rules at the start of every meeting (they are printed in the manual). Keep a consistent opening: review homework or Hassle Logs, then the new content, then role-play or discussion, then assign homework.
• Example language (illustrative): "One person talks at a time, no put-downs, what's said here stays here unless someone's safety is at risk."
• Watch for: [synthesis] rules imposed without any youth input are more often tested. Inviting a youth to add a rule builds ownership; the core rules are non-negotiable.

A3. Caregiver and staff home notes

• Purpose: support transfer by telling significant adults which skill is being learned [source: Georgia FAQ; Research Press, Appendix C].
• Steps: Each time a new social skill is introduced, send home (or to unit staff or teachers) the Staff/Caregiver Social Skills Home Note. It covers why the skill matters, its steps and the youth's homework. Similar forms exist for Anger Control and Moral Reasoning. Ask adults to prompt and praise the skill when they see it.
• Example language to caregivers (illustrative): "This week Jordan is practising Keeping Out of Fights. If you see him step away from an argument, tell him you noticed."
• Watch for: adults who punish skill attempts (for example, a youth's complaint treated as backchat). [synthesis] Coach those adults directly.

B. Social Skills Training (Skillstreaming)

The four core procedures are modelling, role-playing, performance feedback and transfer training [source: ACF; Research Press; Amendola & Oliver]. One skill is taught per week.

B1. Social Skills session procedure

• Purpose: the standard way every skill is taught. Trainer slides list nine steps [source: Amendola & Oliver slides].
• Steps:
  1. Define the skill. Name it, discuss why it matters, and put the steps on a poster or skill card.
  2. Model the skill. The two facilitators act out at least one realistic scene using all the steps, thinking aloud at the cognitive steps.
  3. Establish trainee need. Each youth names where, when and with whom they need the skill. Record these on the flip chart.
  4. Select a role-player (the main actor), usually a volunteer.
  5. Set up the role-play. The main actor picks a co-actor who resembles the real person. Gather details (where, when, what was said) and set the scene.
  6. Conduct the role-play. The main actor follows the steps, with one facilitator pointing at each step on the poster. Assign each observer a step to watch.
  7. Performance feedback in this order: co-actor, observers, facilitators, then the main actor's own self-evaluation.
  8. Assign skill homework (transfer): when, where and with whom the youth will try the skill before the next session.
  9. Select the next role-player and repeat until each youth has had a turn, across sessions if needed.
• Example language (illustrative): "Who has had a time this week when someone accused you of something? Malik, you said the unit staff. Pick someone here who can play him."
• Watch for: role-plays that drift into rehearsing aggression, or the main actor skipping steps. Pause, point to the step and restart. Working well looks like youth naming the steps without prompting and bringing real situations.

B2. Modelling displays

• Purpose: show the skill done well, in a scene youth recognise [source: Amendola & Oliver; Research Press, Skillstreaming overview].
• Steps: Plan and rehearse with your co-facilitator beforehand. Use settings from the youth's world (unit, school, street, home). Show every step in order. Speak the thinking steps aloud. Show a positive outcome. [synthesis] Two short scenes beat one long one.
• Example language (illustrative), during the "think" step: "I'm thinking: if I swing now, I lose my home pass. What else can I do?"
• Watch for: facilitators hamming it up so the model looks fake or mocking. Keep it realistic.

B3. Role-play (main actor and co-actor)

• Purpose: rehearse the skill behaviourally in the youth's own situation [source: Wikipedia; Amendola & Oliver].
• Steps: Use a real past or upcoming situation. The main actor briefs the co-actor on how the other person acts. Keep it short. One facilitator coaches from beside the main actor; the other watches the group. If a step is missed, stop, prompt and replay.
• Example language (illustrative): "Freeze. Which step are you on? Right, step two. Try that line again."
• Watch for: the co-actor provoking beyond the brief. Use "Freeze" and re-brief.

B4. Performance feedback

• Purpose: reinforce correct step use and shape the next attempt [source: Amendola & Oliver].
• Steps: In order: co-actor (how it felt to receive), observers (did their assigned step happen?), facilitators (praise specific steps, then one improvement), and the main actor (self-evaluation). Keep feedback behavioural and tied to the steps.
• Example language (illustrative): "Dez, you were watching step three. Did he say how he felt before saying what he wanted?"
• Watch for: put-downs dressed up as feedback. Redirect to step-specific comments.

B5. Skill homework and transfer training

• Purpose: move the skill from the group into real life [source: ACF; Skillstreaming overview].
• Steps:
  1. Each youth writes the skill, its steps, and where, with whom and when they will try it.
  2. After trying it, they record what happened, which steps they used, and rate their own performance.
  3. Review the homework at the start of the next session. A successful attempt can become the next role-play; a failed one becomes a re-rehearsal.
  4. The Skillstreaming system has a second, more advanced report that adds self-reward [source: Skillstreaming overview; uncertain whether ART uses both levels].
• Example language (illustrative): "Not 'I'll try it sometime.' Who, where, when? Tomorrow at breakfast with Ms R?"
• Watch for: homework not done. See Section 5.

B6. The ten ART curriculum skills (one per week)

The skills, in 3rd-edition order [source: Research Press; Georgia FAQ], are taken from the 50-skill Skillstreaming the Adolescent list. Each skill has 3 to 6 numbered steps printed on skill cards. The step summaries below are paraphrased only to show what each skill covers and are [uncertain]: not checked against the manual. Teach from the official skill cards.

• Week 1, Making a Complaint: decide what the problem is and who to tell; choose the time and place; state the complaint calmly; suggest a solution; listen to the reply.
• Week 2, Understanding the Feelings of Others: watch the other person's face, body and words; name what they might feel; decide whether and how to show you understand.
• Week 3, Getting Ready for a Difficult Conversation: picture the situation; think about how you and the other person will feel; plan what to say and what they might answer; pick the best approach.
• Week 4, Dealing with Someone Else's Anger: listen; try to understand what the angry person is saying and feeling; decide whether to speak or leave; deal with the anger without escalating.
• Week 5, Helping Others: notice that someone may need help; decide whether and how to help; offer; follow through.
• Week 6, Keeping Out of Fights: stop and think about why you want to fight; consider the outcome; think of other ways to handle it; choose and act.
• Week 7, Dealing with an Accusation: think about what you've been accused of and why; decide whether it's fair; choose how to answer (deny calmly, explain, apologise, offer to make up); respond. Note: the Research Press page lists "the angry behavior cycle" as week 7's skill, while Georgia lists Dealing with an Accusation [uncertain]. Follow your edition's table.
• Week 8, Dealing with Group Pressure: think about what the group wants and why; decide what you want to do; plan how to tell them; tell them [steps of this skill as published by Research Press, paraphrased].
• Week 9, Expressing Affection: decide whether you have caring feelings for someone; decide whether they'd want to know; choose how and when; express it.
• Week 10, Responding to Failure: decide whether you failed and what went wrong; think about what you could do differently; decide whether to try again and plan it.

For each skill, run B1 to B5. [synthesis] In juvenile justice groups, Making a Complaint and Dealing with an Accusation often bring up grievances against staff. Treat these as legitimate role-play material, not as defiance.

C. Anger Control Training

One link of the anger control chain is taught each week and then added to the rehearsal of all the links learned so far. The chain is: triggers, cues, anger reducers, reminders, thinking ahead, social skill, self-evaluation [source: Wikipedia; Amendola & Oliver slides; Research Press].

C1. Anger Control session procedure

• Purpose: the standard format for every Anger Control session [source: Amendola & Oliver slides].
• Steps:
  1. Review the previous week's homework and Hassle Logs.
  2. Define the new concept and where it sits in the chain.
  3. Facilitators model the chain up to and including the new link.
  4. Establish need: youth choose real incidents from their Hassle Logs.
  5. Select the main actor and give each observer a link to watch.
  6. Select a co-actor and conduct the role-play. The main actor speaks each internal step aloud ("bubble-talking").
  7. Feedback in this order: co-actor, observers, facilitators, main actor.
  8. Next role-player.
  9. Assign homework: a Hassle Log entry using the chain so far.
• Example language (illustrative): "Say the inside stuff out loud so we can hear it: what's your cue, what's your reducer?"
• Watch for: youth acting out the angry part with gusto and rushing the coping part. Slow the role-play down at the links.

C2. ABCs of anger (week 1)

• Purpose: introduce the analysis of anger episodes into Antecedent (what led up to it), Behaviour (what you did) and Consequence (what happened next). Explain the goals and rules of Anger Control [source: ACF; Amendola & Oliver; Research Press].
• Steps: Explain A-B-C on the board. Each youth breaks down a recent incident into A, B and C. Point out that changing B changes C, and that later sessions give tools for the moment between A and B. Take an initial A-B-C self-assessment from each youth.
• Example language (illustrative): "A is what set you off, B is what you did, C is what it cost or got you. We can't always control A. This group is about B."
• Watch for: youth placing all blame on A ("he made me"). Accept A and keep returning to the choice at B.

C3. Hassle Log and triggers (week 2)

• Purpose: self-monitoring of conflicts, and identifying triggers. Direct (external) triggers are things others do or say. Indirect (internal) triggers are the youth's own interpretations and self-talk [source: Amendola & Oliver; ACF; Wikipedia].
• Steps:
  1. Introduce the Hassle Log: a short structured form completed after each conflict, covering where, what happened, who was involved, what the youth did, how angry they were, and how well they handled it. Use the manual's form; it is not reproduced here.
  2. Explain external versus internal triggers and give examples.
  3. Role-play a logged incident, naming both trigger types.
  4. Homework: complete Hassle Logs between sessions.
• Example language (illustrative): "He bumped you, that's the outside trigger. What did you say to yourself right after? That's the inside trigger, and that's the one we can work on."
• Watch for: blank or invented logs. Accept spoken logs from youth with literacy problems [synthesis].

C4. Cues and anger reducers (week 3)

• Purpose: notice physical warning signs early, then lower arousal before acting [source: Amendola & Oliver; Wikipedia].
• Steps:
  1. Cues: each youth names their own body signals (for example, a hot face, clenched fists, a fast heartbeat, a tight jaw) and the earliest one they notice.
  2. Teach and practise the three anger reducers:
     • Deep breathing: several slow breaths, attention on the breath.
     • Backward counting: counting down slowly, silently, ideally turning away.
     • Pleasant imagery: picturing a calm, pleasant scene.
  3. Each youth picks the reducer that suits them. Role-play: trigger, then cue, then reducer.
• Example language (illustrative): "First warning light? Fists. The second you feel your fists, that's your signal to start counting back."
• Watch for: reducers practised only when calm. [synthesis] Rehearse them after a brief provocation in the role-play so they work under arousal.

C5. Reminders (week 4)

• Purpose: brief self-instructions that guide coping under provocation, such as "chill out" or "stay cool". The method comes from self-instructional training (the slides cite Meichenbaum) [source: Amendola & Oliver; Wikipedia].
• Steps: Explain how self-talk can feed anger or cool it. Model reminders said silently at the moment of provocation. Each youth writes two or three of their own reminders. Role-play the chain so far with the reminder spoken aloud (bubble-talking).
• Example language (illustrative): Reminders a youth might choose: "Not worth it", "Stay calm, stay in control."
• Watch for: borrowed reminders that feel fake. Use the youth's own words.

C6. Thinking ahead (week 5)

• Purpose: use likely consequences to inhibit aggression [source: Amendola & Oliver; Wikipedia].
• Steps: Teach "if-then" thinking: "If I do this, then that will happen." List the short-term and long-term consequences, and the internal (how I'll feel) and external (what others will do) ones. Rank which matter most to the youth. Role-play the chain with an if-then statement.
• Example language (illustrative): "If I hit him, then I'm back in the room for a week and my mum doesn't visit. Is that worth it?"
• Watch for: youth who see consequences only as punishment by others. Help them find internal consequences too (for example, losing respect for yourself).

C7. Self-evaluation (week 6)

• Purpose: after a conflict, judge how well you handled it. Self-reward for what went well; self-coaching for what to do differently [source: Amendola & Oliver; Research Press].
• Steps: Model self-rewarding statements and self-coaching statements. Add self-evaluation as the closing link of the chain role-play. Log self-evaluation in the Hassle Log.
• Example language (illustrative): Self-reward: "I walked away. That was hard and I did it." Self-coaching: "I waited too long to breathe. Next time, start counting at the first sign."
• Watch for: harsh self-criticism. Shape it into specific coaching.

C8. Angry behaviour cycle (week 7)

• Purpose: youth recognise how their own behaviour provokes others, and change it [source: Amendola & Oliver; Research Press].
• Steps: Ask youth to list things they do that annoy or anger others (for example, staring, teasing, insulting, interrupting). Each picks one to change. Model and role-play the change, using the chain.
• Example language (illustrative): "We've talked about what other people do to set you off. What do you do that sets them off?"
• Watch for: defensiveness. Start with facilitators naming their own provoking habits [synthesis].

C9. Using a social skill in the chain, and full-chain rehearsal (weeks 8 and 9)

• Purpose: add a Skillstreaming skill (for example, Keeping Out of Fights or Dealing with an Accusation) as the prosocial action after thinking ahead, then rehearse the whole sequence [source: Amendola & Oliver; Research Press; Georgia].
• Steps: Week 8: model and role-play the full chain, choosing a skill already learned in Social Skills. Week 9: more full-chain rehearsals. The trainer slides say week 9 is rehearsed without the thinking-ahead step [uncertain: the rationale is not given in the sources]. Every youth should do at least one full-chain role-play.
• Example language (illustrative): "Trigger, cue, reducer, reminder, think ahead, now which skill? Then how did you do?"
• Watch for: youth who can recite the chain but freeze in the role-play. Shorten it to cue, reducer, skill, and build back up.

C10. Overall review (week 10)

• Purpose: consolidate and carry forward [source: Amendola & Oliver; Research Press].
• Steps: Review Hassle Logs across the 10 weeks, recap each technique, rehearse the full chain, and reinforce participation. [synthesis] Have each youth write a personal chain card (their trigger, cue, reducer, reminder, key consequence and skill) to keep.

D. Moral Reasoning (social decision-making meetings)

The four phases come from the 3rd edition: introduce the problem situation, cultivate mature morality, remediate moral developmental delay, consolidate mature morality [source: Research Press; ACF Clearinghouse]. The detailed steps below follow a trainer's teaching slides, which cite the 1998 revised edition [source: Campbell slides], and the Amendola & Oliver slides. Background theory: Kohlberg's moral stages as adapted by Gibbs. Immature reasoning is either power-based ("might makes right") or self-interest-based ("you scratch my back") [source: Wikipedia].

D1. Teaching thinking errors

• Purpose: give the group shared words for the self-serving distortions that justify aggression [source: Wikipedia; Amendola & Oliver].
• Steps: Early in the programme, introduce Gibbs's four thinking errors:
  • Self-centred: only my wants and views count.
  • Assuming the worst: expecting hostility from others, or the worst outcome.
  • Blaming others: it's their fault.
  • Minimising/mislabelling: shrinking the harm or putting a harmless label on it.
  The Amendola & Oliver slides also list attribution of hostile intent as an anger-arousing error. Put the list on a poster. In every meeting, the group (not just facilitators) names the errors they hear. Page references for the manual's discussion rules and thinking-errors material are in the Campbell slides.
• Example language (illustrative): "'Everybody does it.' Which thinking error is that?"
• Watch for: labels used as insults. Keep naming the thought, not the person.

D2. Phase 1: Introduce the problem situation

• Purpose: set up the week's dilemma so that everyone understands it and feels it is relevant [source: Campbell slides; Research Press].
• Steps:
  1. Go over the discussion rules.
  2. Hand out the problem-situation sheet. Each sheet has a short story and a set of questions; each question asks what the character should do (for example, persuade the friend, let it happen, or can't decide), with variations.
  3. Ideally youth answer the questions before the meeting [uncertain: pre-meeting completion is the EQUIP/ART convention, not confirmed in the sources].
  4. A group member reads the situation aloud.
  5. Establish the "real" problem, so everyone agrees on what the dilemma is.
  6. Connect it to members' own lives.
• Example language (illustrative): "What's really going on here? What's the hard choice? Has anyone been in something like this?"
• Watch for: the group getting lost in story details. Bring it back to the central decision.

D3. Phase 2: Cultivate mature morality

• Purpose: bring the most mature answers and reasons to the front first, so they set the tone [source: Campbell slides; ACF].
• Steps: Go through the questions. For each one, first ask the youth who gave a responsible answer to give their reasons. Record mature answers and reasons on one side of a T-chart on the flip chart. Reshape less mature but salvageable answers into mature form, and record them on the mature side. Praise mature reasons specifically.
• Example language (illustrative): "Tasha, you said he should stop his friend. Why? So, because the owner gets hurt too. Let's put that up."
• Watch for: facilitators supplying the mature reasons themselves. The reasons must come from peers.

D4. Phase 3: Remediate moral developmental delay

• Purpose: challenge self-centred, power-based or distorted reasoning through peers and perspective-taking [source: Campbell slides; Amendola & Oliver].
• Steps:
  1. Record immature answers that cannot be reshaped on the other side of the T-chart. Do not reinforce them.
  2. Invite more mature members to respond, using the mature list to challenge.
  3. Perspective-taking: members who disagree examine each other's reasoning for thinking errors.
  4. Role-taking: ask the youth how they would feel, or what they would do, in the victim's or another character's position.
  5. Benign confrontation: point out contradictions through questions, without an attacking tone.
• Example language (illustrative): "You said stealing the car is no big deal. If it were your grandma's car, the one she drives to work, would it be a big deal?"
• Watch for: escalation or withdrawal by the challenged youth. Ease off, find a partial agreement and come back to it.

D5. Phase 4: Consolidate mature morality

• Purpose: turn mature answers into a group decision that members own [source: Research Press; ACF; Campbell slides].
• Steps: Where possible, make the more mature answers and their reasons the group's decision (a consensus, or at least a majority). Write the group decision on the chart. Praise the group and individual members for responsible decisions and mature reasoning.
• Example language (illustrative): "Looks like most of the group is going with 'talk him out of it' because people get hurt. Can we all live with that as the group's answer?"
• Watch for: fake consensus produced by pressure from a dominant member. Check quietly with dissenters.

E. Programme-level fidelity procedures

E1. Session evaluation checklists

• Purpose: self-monitoring and supervision of fidelity, using the manual's Appendix B [source: Research Press].
• Steps: After each session, the co-facilitators complete the checklist for that component. Supervisors observe sessions and use the same checklists. Washington's quality assurance process found that non-adherent sites lost the benefit [source: WSIPP].
• Watch for: [synthesis] drift in which sessions become talk-only groups without role-plays. This is the most common failure mode.

E2. Make-up sessions

• Purpose: keep the closed group's sequence intact [source: Georgia FAQ].
• Steps: If a youth misses a session, give them the missed content and have them make it up before the next session. [synthesis] Keep make-ups brief and one-to-one: model, one role-play, homework.

5. Handling problems

The 3rd edition has a chapter on motivation and resistance and one on generalisation, but its content is [uncertain]. Skillstreaming the Adolescent (McGinnis, 2012) handles group-member resistance with a three-level approach: universal, then targeted, then individual. At the individual level it uses functional behavioural assessment and behaviour intervention plans [source: Skillstreaming overview]. Except for points tagged to a source, everything below is [synthesis].

• Disruption or clowning in role-plays. Give the disruptive youth a job: observer for a specific step, or co-actor. Use the agreed rules and consequences calmly, without shaming [source for the no-demeaning stance: Georgia FAQ]. Shorten role-plays and raise the tempo. One facilitator can sit next to the youth. If the disruption continues, add a targeted plan: a brief individual check-in before the session, and a behaviour plan with the unit or school.
• "This is stupid" or active refusal. Link the skill to something the youth wants (a pass, release, keeping a placement, respect). Let them choose the situation. Offer the co-actor role first. Don't argue the case for the programme; ask what would make it worth their time.
• Withdrawn or inactive youth. Start with the observer role and a single step to watch. Then give a small speaking part as co-actor. Praise any participation. Check for depression, trauma reactions, language barriers or learning needs.
• Homework or Hassle Logs not done. Make the homework plan specific (who, where, when) and rehearse it in session. Accept spoken or drawn logs. Do the first log together at the end of the session. Ask unit staff, teachers or caregivers to prompt (home notes, A3). Reinforce attempts, not just successes.
• Skills that work in group but not outside. Ask staff and caregivers to reinforce real-life use. Use role-plays that mirror real settings, with co-actors who resemble the real people. Rehearse more than once. Build the youth's self-reward through self-evaluation (C7). Washington's model of court and probation delivery relies on staff in the youth's environment [source: WSIPP; Georgia].
• Moral Reasoning meeting dominated by negative peers. Bring forward mature voices first (D3). Don't reinforce immature reasons. Ask role-taking questions (D4). Break into pairs for a question, then return to the group. Prevent this at intake by mixing groups so there are enough mature members.
• Peer contagion or deviancy training (youth bonding over aggression stories). Keep tight structure. Don't allow glorifying war stories: redirect them to A-B-C analysis. Keep the group size cap [source: Georgia]. Leave out youth with serious instrumental aggression if they dominate. [uncertain] Problems with aggregating antisocial youth are noted in the wider literature; they are not covered in the sources.
• Overwhelm or high arousal during a role-play (for example, a role-play that revives a real fight or abuse). Stop ("freeze"). Use an anger reducer as a group (C4). Let the youth step out with one facilitator. Later, rerun the scene at lower intensity or with a different co-actor. Never script a role-play that reproduces victimisation; adapt it for trauma-exposed youth.
• Rupture with a facilitator (feeling unfairly treated or disrespected). Acknowledge it briefly in the moment. Repair one-to-one after the session. Use Making a Complaint (week 1) as a legitimate way for youth to raise grievances. Facilitators model owning a mistake.
• Disclosure of risk (suicidal thoughts, self-harm, abuse, or a plan to hurt someone) in a log or discussion. Follow the setting's safeguarding and risk procedures: do a private risk assessment the same day, inform the responsible clinician or supervisor, and for self-harm risk include safety planning and lethal-means counselling (secure storage or removal of means). Do not process this in the group.
• Attendance gaps or mid-group transfers. Make up missed sessions before the next one [source: Georgia]. Don't add new members to a closed group; refer them to the next cohort.
• Drift from the model (talk-only sessions, skipped components, changing facilitators). Use the session checklists (E1) and supervision. Non-adherent delivery did not reduce reoffending in Washington [source: WSIPP]. Keep the same facilitator pair throughout [source: Georgia].
• Literacy, cognitive or language barriers. Use picture skill cards and spoken Hassle Logs. Read problem situations aloud. Use fewer questions per meeting. Translated materials exist in several languages [source: ACF].
• Adult or girls' groups. Adaptations exist (for example UCCI's The Prepared Adult, and problem situations with female characters in the 3rd edition) [source: UCCI; Georgia]. Use the adapted materials rather than improvising.

Sources

Sources:

• Aggression Replacement Training (Wikipedia)
• Aggression Replacement Training, 3rd ed. (Research Press product page)
• Aggression Replacement Training (Title IV-E Prevention Services Clearinghouse)
• Aggression Replacement Training FAQ (Georgia Criminal Justice Coordinating Council)
• Amendola & Oliver, ART presentation slides (IIRP, 2013)
• Aggression Replacement Training slides (J. R. Campbell)
• ART Training Overview (University of Cincinnati Corrections Institute, 2021)
• Washington State's Implementation of ART for Juvenile Offenders: Preliminary Findings (WSIPP)
• Skillstreaming the Adolescent overview (Research Press, 2026)

Further reading (not drawn on):

• Glick, B., & Goldstein, A. P. (1987). Aggression Replacement Training. Journal of Counseling & Development, 65(7), 356–362.
• Goldstein, A. P., Glick, B., & Gibbs, J. C. (1998). Aggression Replacement Training: A Comprehensive Intervention for Aggressive Youth (rev. ed.). Research Press.
• Glick, B., & Gibbs, J. C. (2011). Aggression Replacement Training: A Comprehensive Intervention for Aggressive Youth (3rd ed.). Research Press.
• McGinnis, E. (2012). Skillstreaming the Adolescent (3rd ed.). Research Press.
• Barnoski, R. (2004). Outcome Evaluation of Washington State's Research-Based Programs for Juvenile Offenders. Washington State Institute for Public Policy [uncertain: key-source abstract].