Atlas›Domain VIII · Creative & expressive arts therapies›Guide 070
Guide VIII-070

Behavioural music therapy

Clifford K. Madsen, Vance A. Cotter, Clifford H. Madsen Jr.1968United States
A 12 · L 80 · M 815 interventions in the Trinity explorer
Client cardCompare
[source: …] where a claim comes from[uncertain] not confirmed in a primary source[synthesis] the writer's reconstruction

1. Introduction

Version followed: Clifford K. Madsen's behavioural (applied-behaviour-analysis) approach as restated in his 1999 address to the 9th World Congress of Music Therapy, which names the 1968 Journal of Music Therapy paper with Vance Cotter (Madsen, Cotter & Madsen, 1968) as its starting point; with Madsen's work on approval and disapproval (Madsen & Alley, 1979; Madsen, 2000) and Standley's (1996) meta-analysis of music as reinforcement. Where Madsen's sources give no procedure, the session process is built from the stages of a case run on Suzanne Hanser's data-based model (Hanser, 1999), as reported by Árkosi (2024), and from published single-case studies of contingent music and song cues (McCarty et al. 1978; Graber Juhnke 2015; Sato & Sonoyama 2017). Hanser's handbook (1999; 3rd edition 2018) is cited only through Árkosi [uncertain].

Sources support: the 1968 paper's citation (Journal of Music Therapy 5(3), 69–71, doi 10.1093/jmt/5.3.69) and Madsen's account of it as the origin of the approach; the four functions of music (cue, time and body-movement structure, focus of attention, reward) as stated in Madsen's 1999 address; the Madsen & Alley (1979) findings (abstract); and the US MT-BC credential rules (CBMT). Unverified [uncertain]: the content of the 1968 paper beyond its citation; the exact title and content of Madsen's 1981 NAMT behavioural guide for music therapy with people with intellectual disabilities; and the content of Hanser's handbook. Madsen does not link each of the four functions to a worked procedure, and no fixed approval-ratio target (such as 80:20) is published in the sources used. Published single-case studies now support parts of 4.5 (contingent music) and 4.6 (music as a cue); the remaining step-by-step session procedures in section 4 are the writer's operationalisation of the published principles, marked [synthesis].

Origin: behavioural music therapy dates from Madsen's 1968 paper. It was written by Clifford K. Madsen, Vance A. Cotter and Clifford H. Madsen Jr. Madsen says it was developed partly at Parsons State Hospital, Kansas. Madsen was based at Florida State University. [source: JMT issue 5(3) contents; Madsen 1999 address]

What it is. Behavioural music therapy is applied behaviour analysis that uses music. Madsen calls it the use of music and of the therapist's own self to change behaviour, built on learning principles and on control of the client's environment. Other names he gives for it are "applied behavior analysis" and "music cognitive/behavior modification". [source: Madsen 1999 address]

How change happens. Learning is a change in behaviour. Behaviour changes when the therapist arranges the consequences of responses on purpose, following reinforcement principles. Madsen says music can be used "as a cue", as "a time and body movement structure", "as a focus of attention" and "as a reward". The therapist sets explicit, measurable, socially appropriate goals, measures behaviour before and during treatment, and shows change with data that others can check and replicate. Madsen uses "behaviour" broadly, to cover emotional and cognitive responses as well as motor acts. [source: Madsen 1999 address; Madsen 2000]

Who it is for. The published examples cover people with intellectual disability in institutions, a juvenile offender treated at home with his mother, school pupils, children with emotional disturbance on school buses, young children with autism or intellectual disability in transitions, and premature infants (contingent music). Madsen's papers do not define the target populations in detail. [source: Madsen 1999 address; Madsen & Alley 1979; McCarty et al. 1978 abstract; Graber Juhnke 2015 abstract; Sato & Sonoyama 2017 summary page; infant use: uncertain]

Professional frame.

  • Who may practise. This is a music therapy method. Use it only as a credentialled music therapist, or jointly with one. In the US the credential is the MT-BC (Music Therapist-Board Certified), awarded by the Certification Board for Music Therapists (CBMT). Candidates must complete the academic and clinical training requirements of an AMTA-approved music therapy programme, or their equivalent, including a supervised internship, and then pass the CBMT written exam. Certification lasts five years and is renewed with 100 recertification credits, 3 of them in ethics. [source: CBMT certification page; CBMT Candidate FAQs; CBMT Candidate Handbook] CBMT's pages give no fixed internship-hour figure. In the UK, music therapists register with the HCPC, with BAMT as the professional body [uncertain].
  • Training. Madsen says the principles are few but applying them is complex. It needs extensive training, a solid grounding in behavioural principles, and analytical and creative skill. [source: Madsen 1999 address]
  • Behaviour-analytic interventions. Interventions that use extinction, punishment or restriction should also follow local behaviour-analysis and restrictive-practice rules. Where these apply, involve a board-certified behaviour analyst or the equivalent [synthesis].
  • If you are a psychotherapist without music therapy training. You can use the generic behavioural parts (defining targets, recording data, giving praise, using a preferred song as a reward). Do not call this music therapy, and refer clients for music therapy proper [synthesis].

Hearing safety.

  • Music therapists can be exposed to too much noise at work. A dosimetry case study of a therapist working with people with learning disabilities found exposures above UK workplace standards. It recommended a noise risk assessment, control of exposure, hearing protection and, where indicated, health surveillance. [source: MacMahon & Page 2015, abstract]
  • NIOSH guidance, as reproduced in a university information sheet, gives a daily maximum of 8 hours at 85 dB and 15 minutes at 100 dB. The safe time falls steeply as loudness rises. [source: TTU "Protecting Your Hearing Health"]
  • In practice: keep recorded music used as a reinforcer at moderate volume, and use a volume limit on headphones. Watch clients who cannot report discomfort, such as non-verbal clients or those with sensory sensitivities or hyperacusis. Keep drums and amplified instruments away from ears, and offer ear defenders [synthesis].

Music as a trauma trigger.

  • A song chosen as a "reward" can be linked to abuse, loss, gang or prison life, or a religious or cultural injury. Contingent interruption of a loved piece can also feel like punishment.
  • Screen for music the client associates with distress. Let clients veto any piece, and keep a do-not-play list [synthesis].

2. Framework

Madsen's framework is a set of principles applied through a data cycle, not a fixed number of sessions. The six phases below are the writer's arrangement [synthesis]. They combine Madsen's elements (explicit goals, consequences, shaping, data, generalisation) [source: Madsen 1999 address] with the stages of a case run on Hanser's data-based model as reported by Árkosi (2024): background data, observation, a music questionnaire, an individual session, a client profile, goal setting, planning, implementation and evaluation [source: Árkosi 2024]. Árkosi names observation, goal-oriented planning and objective measurement as the model's main pillars and gives no labelled, ordered phase list; Hanser's own sequence is not reproduced here [uncertain].

Phase 1: Referral and assessment.

  • Goal: understand the problem behaviour, the client's strengths, and the client's music history and preferences.
  • What to gather: records; reports from carers or teachers; direct observation in the natural setting (for example a class or the home); a first individual music session; a music-preference questionnaire (listening habits, preferred genres, musical activities); and a written client profile. [source: Árkosi 2024]
  • Move on when: you have at least one target you can define and a list of preferred music [synthesis].

Phase 2: Pinpointing targets and setting objectives.

  • Set explicit, measurable, socially appropriate goals. [source: Madsen 1999 address]
  • Write two kinds of objective: musical objectives and non-musical (therapeutic) objectives. Write short-term objectives for each session and long-term ones for the whole course. [source: Árkosi 2024]
  • Move on when: each target has an operational definition that two observers could score the same way [synthesis].

Phase 3: Baseline observation.

  • Measure each target before intervening. [source: Árkosi 2024; Madsen 1999 address]
  • Move on when: the baseline is stable, or clearly worsening and needing prompt action. Three to five data points is a common rule of thumb [synthesis].

Phase 4: Music therapy strategy (intervention).

  • Choose which music function or functions to use (cue, structure for time and body movement, focus of attention, reward) [source: Madsen 1999 address] and the contingency (what earns music or praise). The sources do not say how to match a function to a target; choose by what the target needs (a signal, pacing, competing attention or a consequence) [synthesis].
  • Build in shaping steps and plan generalisation, for example by treating the client in the setting where the problem happens. [source: Madsen 1999 address]

Phase 5: Evaluation.

  • Plot the data and compare it with baseline, using single-subject or experimental designs. [source: Madsen 1999 address]
  • Árkosi reports Hanser's inter-observer reliability formula with an 85% threshold. Below that, further measurement was needed. Her own case reached 89.28% agreement. [source: Árkosi 2024 citing Hanser 1999]
  • Decision rule: if the data show no trend after several sessions, change one variable at a time (the reinforcer, the schedule, the size of the shaping step, or the cue) [synthesis].

Phase 6: Generalisation, fading and termination.

  • Fade contingent music, shift towards natural reinforcers such as praise and the client's own playing, and hand procedures over to carers or teachers.
  • In Madsen's juvenile case the mother was trained to give the reinforcement herself. Follow-up was intermittent, both short- and long-term. [source: Madsen 1999 address]
  • Árkosi's case had no formal termination phase and ended with a recommendation to continue. [source: Árkosi 2024]

Typical format and dose (examples, not norms).

  • Madsen and Madsen (1968), juvenile case: treatment at home. Sessions moved from listening alone, to listening plus questioning, to a first verbal exchange of about 10 minutes, each followed by a guitar lesson. Session lengths were adjusted in steps (for example, listening cut from 30 to 20 minutes while questioning rose from 3 to 8 minutes). The counsellor spent about 48 hours over four weeks, front-loaded. [source: Madsen 1999 address]
  • Árkosi (2024), Hanser model: 25 weekly 45-minute individual sessions over a school year, with 5 scheduled measurement points. [source: Árkosi 2024]
  • Group work: the approach also runs in groups and classrooms, measured with approval-ratio and off-task data. [source: Madsen & Alley 1979]

3. Guiding principles

  • Behaviour is what you can observe, broadly defined. Write every goal as something that can be seen or recorded, including emotional and cognitive responses. Practical meaning: "calmer" becomes "fewer than 2 shouting episodes per 30-minute session". [source: Madsen 1999 address]
  • Learning means a change in behaviour. Practical meaning: judge a session by measured change, not by how it felt. [source: Madsen 2000]
  • Arrange consequences explicitly. Practical meaning: decide in advance what earns music or praise, and deliver it every time at first. [source: Madsen 1999 address]
  • Contingent beats continuous. In Standley's meta-analysis, music given contingently was more effective than continuous music and more effective than contingent non-music rewards. Practical meaning: do not play music all session "for mood" if it is meant to reinforce. Make access depend on the target behaviour. [source: Standley 1996 abstract]
  • Do not dilute music with other rewards. Pairing music with food, approval or visual stimulation reduced its effect. Practical meaning: use music on its own as the consequence, and test your own combination with data. [source: Standley 1996 abstract]
  • Keep a high approval ratio and few reinforcement mistakes. Behaviourally trained therapists showed high approval ratios, and their clients were rarely off-task. The sources give no numeric target ratio (see 4.10). Practical meaning: praise the target behaviour often and specifically. Do not praise or attend to misbehaviour by mistake. [source: Madsen & Alley 1979 abstract; Madsen 2000]
  • Music can be the reward and the thing being learned. Practical meaning: access to playing (for example a guitar lesson) can reward non-musical work. Musical skill is itself a legitimate objective. [source: Standley 1996 abstract; Madsen 1999 address]
  • Shape by successive approximations. Practical meaning: raise demands in small steps the client can manage. Grade difficulty so the client succeeds. [source: Madsen 1999 address; Árkosi 2024]
  • Treat in the natural environment where possible. Practical meaning: work where the problem occurs (home, classroom, ward) so the change carries over. [source: Madsen 1999 address]
  • Results must be publicly verifiable and replicable. Practical meaning: record data, graph it, check reliability with a second observer, and use single-case designs. [source: Madsen 1999 address; Árkosi 2024]
  • The therapist creates, selects and improvises music to fit the client. Practical meaning: the behavioural frame does not mean canned music. Musicianship matters. [source: Madsen 1999 address]
  • Sustained intensity. Madsen's teaching research found that "sustained control" of the interaction ("intensity") could be rated reliably, and that keeping it up over time was the hard part. Practical meaning: pace, eye contact and quick feedback all count, so review your own consistency over weeks. [source: Madsen 2000]
  • Involve the people around the client. Practical meaning: train parents, teachers or staff to deliver the contingencies. [source: Madsen 1999 address; Árkosi 2024]
  • Least restrictive, client-dignified practice. Prefer positive reinforcement to withdrawal, punishment or deprivation. Respect client choice of music, and get consent or assent for every contingency [synthesis].

4. Intervention inventory

The sources give the procedures below as principles and case examples. Madsen's 1999 address lists the four music functions without linking each to a worked example, and no published step-by-step session protocol for the method is available. Where a published study describes a procedure (contingent music in 4.5, song cues in 4.6), the step carries its source tag. Unless a step carries a source tag, the step-by-step wording is the writer's operationalisation [synthesis]. Example therapist language is illustrative throughout.

4.1 Music-preference and reinforcer assessment

  • Purpose: find out which music is reinforcing for this client, and which music is aversive or linked to trauma. Use it in Phase 1, and again whenever a reinforcer stops working.
  • Steps:
    1. Give a music questionnaire to the client, or to a carer: listening habits, preferred genres, and preferred musical activities (listening, singing, playing particular instruments). [source: Árkosi 2024 citing Hanser 1999]
    2. Watch the client in a free-choice music session. Note what they approach, how long they stay with it, and what they ask for again [synthesis].
    3. Run brief choice trials: offer two options (for example 30 seconds of song A or song B, or guitar or drum) and record what the client picks. Rank the options [synthesis].
    4. Ask directly, and ask carers, about music to avoid. Build a do-not-play list [synthesis].
    5. Write the reinforcer menu and the do-not-play list into the client profile. [source: Árkosi 2024 (profile)] [synthesis (lists)]
  • Example language (illustrative): "Which songs do you put on when you want to feel good? Is there any music you never want to hear in here?"
  • Watch for: preferences that change from week to week (reassess); preferences that are carers' guesses; music that raises arousal when calm is the goal. It is working when the top-ranked item reliably raises the behaviour it follows.

4.2 Pinpointing target behaviours and writing objectives

  • Purpose: turn the referral problem into explicit, measurable, socially appropriate goals. [source: Madsen 1999 address]
  • Steps:
    1. List the complaints. For each one, name the observable behaviour, the setting and the time window [synthesis].
    2. Choose a behaviour to increase (for example task completion, appropriate speech, time in seat) and, where needed, one to decrease (for example verbal abuse, tics) [synthesis].
    3. Write an operational definition, including what counts and what does not [synthesis].
    4. Write musical and non-musical objectives, each split into short-term (per session) and long-term targets. [source: Árkosi 2024]
    5. Agree the targets with the client, or with the guardian plus the client's assent, and with the teacher or carer who will continue the work. Árkosi's case used a written agreement with the school. [source: Árkosi 2024] [synthesis (consent)]
  • Example language (illustrative): "Let's pick one thing we can count together. How about the number of times you finish a worksheet before music time?"
  • Watch for: vague targets ("self-esteem"). Árkosi tracked self-confidence on an arbitrary scale starting at 10, which shows how hard such constructs are to measure; prefer behavioural indicators. [source: Árkosi 2024] It is working when two people can score the same session and agree.

4.3 Behavioural observation, baseline and graphing

  • Purpose: document the behaviour before and during treatment, so decisions follow data. [source: Madsen 1999 address; Árkosi 2024]
  • Steps:
    1. Choose a recording method. Use frequency counts for discrete acts, duration for on-task or listening, interval or time-sampling for off-task in groups. Madsen and Alley recorded the percentage of time off-task [source: Madsen & Alley 1979 abstract]; the choice of method is [synthesis].
    2. Record the baseline before introducing contingencies. [source: Árkosi 2024]
    3. Graph every session, with the baseline and intervention phases marked. [source: Árkosi 2024 (graphed measurement points)]
    4. Get a second observer for some sessions and calculate agreement. Hanser's threshold, as reported, is 85%. [source: Árkosi 2024 citing Hanser 1999]
    5. Use a single-subject design (for example AB, reversal or multiple baseline) to show the change came from the intervention. [source: Madsen 1999 address (single-subject designs)] Specific design choice is [synthesis].
  • Example language (illustrative): to the client, "I'm going to keep a tally on this sheet. You can see it at the end."
  • Watch for: observer drift, and recording that eats into therapy time (video helps). Plot data the same day. It is working when you change the plan because of the graph rather than impressions.

4.4 Music as contingent reinforcement (reward)

  • Purpose: increase a target behaviour by making access to preferred music depend on it. This is Madsen's "reward" function, and in Standley's meta-analysis it was more effective than continuous music or non-music rewards. [source: Madsen 1999 address; Standley 1996 abstract]
  • Published example: in Madsen and Madsen (1968), guitar lessons, including improvisation, were contingent on completing assigned work tasks, and the guitar was withdrawn after non-compliance. [source: Madsen 1999 address]
  • Steps:
    1. Pick the top reinforcer from 4.1, and decide the contingency ("when X, then music") [synthesis].
    2. Begin with a continuous schedule: music after every instance or every completed task [synthesis].
    3. Deliver the music promptly. State the contingency simply before and after [synthesis].
    4. Use music on its own as the consequence, because pairing it with food or other stimuli reduced its effect. [source: Standley 1996 abstract]
    5. Once the behaviour is stable, thin the schedule (every second task, then a session-end reward), and lengthen the work required (shaping, 4.9) [synthesis].
    6. Hand over to natural reinforcers (praise, the client's own enjoyment of playing) and to carers [synthesis].
  • Example language (illustrative): "Finish these three problems, and then we'll play the next verse of your song on guitar." Then afterwards: "You finished all three, so here's the guitar."
  • Watch for: satiation (rotate pieces); using music the client is indifferent to; withholding music the client needs for regulation. Contingent music should never be the client's only access to music in their life [synthesis]. It is working when the target rate rises above baseline and stays there as the schedule thins.

4.5 Contingent interruption of music (removing music after an unwanted behaviour)

  • Purpose: reduce an unwanted behaviour by briefly stopping preferred music whenever it occurs. Barrett (1962) used this to reduce multiple tics. [source: Madsen 1999 address citing Barrett 1962] Published example: on three school buses carrying children with emotional disturbance, the team recorded a baseline without music, then a baseline with ordinary background music, and then made the music contingent on behaviour, introduced on one bus at a time a week apart. Fighting and out-of-seat time, recorded as the percentage of time in inappropriate behaviour, fell once the music became contingent [source: McCarty, McElfresh, Rice & Wilson 1978, abstract]. Not every study finds an effect: with third graders doing a desk task, neither continuous background music nor contingent interruption or presentation of music changed task performance or body movement [source: Wolfe 1982, abstract].
  • Steps:
    1. Record a baseline first, without music and then with non-contingent music, so that you can tell the effect of the contingency from the effect of music alone [source for this sequence: McCarty et al. 1978 abstract]. Then play preferred music during a defined activity [synthesis].
    2. When the target behaviour occurs, stop or mute the music at once, with no lecture [synthesis].
    3. Resume it as soon as the behaviour has been absent for a short set interval [synthesis].
    4. Reinforce an alternative behaviour at the same time (4.4, 4.9) [synthesis].
  • Example language (illustrative): before you start, "The song keeps playing while hands stay on the table. If they leave the table, the song pauses until they're back."
  • Watch for: distress, escalation or a burst of the behaviour when music stops. Involuntary behaviours such as tics need medical and neurological input, so do not treat them as wilful [synthesis]. Get consent or assent and agreement from the team, because this is a punishment-type procedure [synthesis]. It is working when the behaviour falls without rising distress.

4.6 Music as a cue

  • Purpose: a musical signal sets the occasion for a behaviour (for example a transition, a turn to speak, or tidying up). Madsen names "a cue" as one of music's four functions but gives no worked example [source: Madsen 1999 address]. Published single-case studies show the use: a therapist sang a transition song while walking with each child through a daily transition between parts of a building, and transitions were shorter with the song than without it (reversal design, three children with autism aged 4 to 8) [source: Graber Juhnke 2015, thesis abstract]. In a Japanese day-treatment study, one composed song served as the prompt to start tidying toys away before a group activity, and a second song as the reward for finishing; all three children improved on baseline, though which function helped differed between children [source: Sato & Sonoyama 2017, NZJMT, summary page]. Step wording below is otherwise [synthesis].
  • Steps:
    1. Choose a short, distinctive musical signal (a chord, chime or song phrase), or a short song kept for one transition only [song form: Graber Juhnke 2015; Sato & Sonoyama 2017; choice of signal synthesis].
    2. Present the signal right before or during the behaviour you want (the start of tidying, the walk to the next room), and reinforce compliance. If the cue alone does not work for a child, try the song as a reward for finishing instead, as the function of the music can differ between children [source: Sato & Sonoyama 2017 summary page; timing wording synthesis].
    3. Fade the verbal prompt until the musical cue alone works [synthesis].
    4. Teach carers to use the same cue in the client's natural settings [synthesis].
  • Example language (illustrative): "When you hear this tune, it's time to put the instruments in the box."
  • Watch for: cues that are overused and lose their meaning; cues that sound like sounds the client finds aversive. It is working when the behaviour follows the cue without verbal prompts.

4.7 Music as a structure for time and body movement

  • Purpose: rhythm, tempo and song form organise how long an activity lasts and how movement is sequenced. This is one of Madsen's four functions. [source: Madsen 1999 address] The applications below are [synthesis].
  • Steps:
    1. Match the length of the task to the length of a piece (for example "work until the song ends") [synthesis].
    2. Use a steady beat to pace repetitive motor tasks or exercises, and grade the tempo [synthesis].
    3. Signal phases of a session with consistent musical sections (a hello song, a work segment, a goodbye song) [synthesis].
  • Example language (illustrative): "We'll keep going until the song finishes; you'll hear it slow down at the end."
  • Watch for: tempos that over-arouse; rigid structure for clients who need flexibility. It is working when time on task or movement accuracy improves inside the musical frame.

4.8 Music as a focus of attention (incompatible response)

  • Purpose: use listening, or making music together, to hold attention and to compete with a problem response. Madsen names "a focus of attention" as one of music's four functions. In the 1968 juvenile case, mother and son listened to recorded music in the background during communication sessions, and shared listening was paired with talking to replace intense emotional exchanges; the account calls listening "incompatible with intense emotional reactions". [source: Madsen 1999 address] The address does not say which function this case illustrates; reading it as the focus-of-attention function is my inference [synthesis].
  • Steps (as in the case):
    1. Start with joint listening only, with no demands. [source: Madsen 1999 address]
    2. Add brief, structured questions while the music plays, and gradually lengthen the talking segment. In the case, listening went from 30 to 20 minutes as questioning grew from 3 to 8 minutes. [source: Madsen 1999 address]
    3. Move to a first verbal exchange of about 10 minutes, followed by the reinforcer (a guitar lesson), then lengthen the sessions. [source: Madsen 1999 address]
    4. Before step 2, coach the parent in reinforcing appropriate speech (4.11). [source: Madsen 1999 address]
  • Example language (illustrative): to the parent, "While the track plays, ask one easy question, then just listen. If he answers calmly, tell him you liked hearing that."
  • Watch for: music that the dyad associates with past conflict; one party using the music to avoid talking altogether. It is working when calm talk lengthens and blow-ups fall.

4.9 Shaping by successive approximations and graded tasks

  • Purpose: build a behaviour the client cannot yet perform by reinforcing closer and closer approximations to it. In the juvenile case, work tasks became longer and more complex day by day, and session segments were lengthened in steps. [source: Madsen 1999 address] Árkosi graded the difficulty of musical tasks (for example colour notation before standard notation, and hands separately before hands together) so that each task could be achieved. [source: Árkosi 2024]
  • Steps:
    1. Define the end behaviour and the client's current level from the baseline [synthesis].
    2. List small steps between the two. Musical skills often split naturally, for example one hand, then two hands, then reading notation [synthesis].
    3. Reinforce each step until it is reliable, usually over 2 or 3 consecutive sessions, then require the next step [synthesis].
    4. If the client fails, drop back one step and make the steps smaller [synthesis].
  • Example language (illustrative): "Last week you played the right hand all the way through. Today we add just the first two left-hand notes."
  • Watch for: steps that are too big (errors, avoidance) or too small (boredom). It is working when the client meets each criterion within a few sessions.

4.10 Therapist approval and disapproval; the instructional sequence

  • Purpose: deliver social reinforcement precisely. Madsen's group codes therapist or teacher behaviour as academic or social approval and disapproval, and as "errors", such as approving a child who is misbehaving. Behaviourally trained music therapists kept high approval ratios, and their clients were rarely off-task. [source: Madsen & Alley 1979 abstract; Madsen 2000]
  • Steps:
    1. Structure each teaching unit as a sequence of three parts. [source: Madsen 2000 (Instructional Sequence Observation Form)]
      • Instruction: use the client's name, give task-specific information, or model the task.
      • Client response: correct or incorrect.
      • Immediate feedback: specific approval wherever possible, rather than general ("try again") or non-specific feedback.
    2. Give factual content directly, rather than asking questions the client cannot yet answer, so errors stay low. Keep open questions for opinions and reasoning. [source: Madsen 2000]
    3. Aim for far more approvals than disapprovals. Madsen & Alley (1979) report "high approval ratios" for behaviourally trained therapists, but the abstract gives no figures. None of the sources, including Madsen (2000) and the abstracts of Hanser (1982) and Zhukov (2008), gives a fixed target ratio (for example the often-quoted 80:20). Treat any number as a rule of thumb, not as Madsen's published standard [uncertain]. Ignore minor off-task behaviour where safe, and approve on-task peers or the client's next appropriate act [synthesis].
    4. Keep the pace and intensity sustained: brief instructions, quick feedback, and little dead time. [source: Madsen 2000]
  • Example language (illustrative): "Sam, watch my hand: strum down four times." After the client strums: "Yes, four even strums, right on the beat."
  • Watch for: reinforcement mistakes, such as attending to shouting or praising during misbehaviour; general praise ("good") that does not name the behaviour. It is working when a tally of your approvals and disapprovals shows a high ratio and the client's off-task time falls.

4.11 Parent or carer training through role-play

  • Purpose: transfer the contingencies to the people who live or work with the client. In the 1968 case, the mother learned to give verbal reinforcement (praising appropriate speech) through role-play, in three 10-minute practice sessions. [source: Madsen 1999 address]
  • Steps:
    1. Explain the target and the reinforcement rule in one or two sentences [synthesis].
    2. Model it: you play the parent and the parent plays the child [synthesis].
    3. Swap roles. The parent practises while you give specific feedback, using the same approval rules as in 4.10 [synthesis].
    4. Repeat in short blocks. The case used three 10-minute sessions. [source: Madsen 1999 address]
    5. Have the parent record one simple daily measure at home [synthesis].
  • Example language (illustrative): "Let's practise. I'll be your son and grumble a bit, and you praise the first calm sentence I say."
  • Watch for: the parent only noticing problems; mixed messages between carers; carer stress and capacity. It is working when the parent gives approvals unprompted and home data improve.

4.12 Systematic desensitisation and relaxation with music

  • Purpose: reduce anxiety or anger by pairing relaxing music with graded exposure to a feared or provoking situation. Madsen lists desensitisation and relaxation among the techniques that can use music. Elements of desensitisation were used to reduce anger in the 1968 case. [source: Madsen 1999 address] The procedure below is standard behavioural practice, not taken from Madsen [synthesis].
  • Steps:
    1. Choose calming music with the client, and check it is not linked to distress [synthesis].
    2. Teach relaxation (paced breathing or muscle relaxation) with that music until the client can relax reliably [synthesis].
    3. Build a graded hierarchy of situations together, from mild to strong [synthesis].
    4. Present each item in imagery or in vivo while the music plays. Move up only when distress at the current step is low, as the client rates it [synthesis].
    5. Fade the music, or turn it into a portable coping cue such as a playlist [synthesis].
  • Example language (illustrative): "Let the music set your breathing. Now picture the first step on our list, your mum asking about homework. Rate how tense you are, 0 to 10."
  • Watch for: contraindications for exposure. These include acute psychosis, active suicidality, current danger at home, and dissociation, which needs trauma-specialist pacing. Always get consent and let the client stop at any point. It is working when distress ratings fall and the client climbs the hierarchy [synthesis].

4.13 Other techniques Madsen lists as usable with music

  • What the source says: Madsen's 1999 list also names conditioning, counter-conditioning, extinction, role-play, conditioned avoidance, self-disclosure, emotive imagery, modelling, negative practice, expressive-rational approaches and stimulus deprivation. He says all of them can use music to produce measurable change. He gives no music-specific procedures for them. [source: Madsen 1999 address] Everything below is [synthesis].
  • Modelling. Demonstrate the target behaviour, such as a musical skill or a social greeting in song, and reinforce imitation. Illustrative: "Watch me play it once, then it's your turn."
  • Role-play. Rehearse social or conflict situations, using a song structure or improvised musical dialogue as the frame. Reinforce the appropriate responses.
  • Emotive imagery. Pair preferred music with positive imagery, for example to reduce fear in medical procedures with children. Check how the imagery and music affect the client.
  • Extinction. Stop reinforcing a behaviour that was maintained by attention, and always reinforce an alternative at the same time. Expect a temporary burst, and do not use extinction for dangerous behaviour.
  • Counter-conditioning. Pair a previously aversive context with enjoyable music, for example in dental or hospital settings.
  • Restrictive procedures: conditioned avoidance, negative practice, stimulus deprivation. These are historical behaviour-modification techniques. Do not use them in routine music therapy. Any restrictive or aversive procedure needs ethics or behaviour-support review, documented consent, and a least-restrictive-alternatives rationale under current codes. Never withhold a client's ordinary access to music, food, company or comfort as a contingency [synthesis].

4.14 Therapist self-monitoring with video feedback

  • Purpose: train and maintain your own reinforcement skills. In Madsen and Alley's pilot, music therapy students learned reinforcement skills with videotape feedback and self-analysis of each session. [source: Madsen & Alley 1979 abstract]
  • Steps:
    1. Video-record a session, with consent [synthesis].
    2. Code a 10–20 minute excerpt: approvals, disapprovals, reinforcement mistakes, and the client's off-task time. Madsen's research coding used 20-minute excerpts checked by two experts. [source: Madsen 2000]
    3. Calculate your approval ratio and the client's off-task percentage, and set one change for next session [synthesis].
    4. Repeat monthly, because keeping intensity up over time is the hard part. [source: Madsen 2000]
  • Watch for: storing video without consent or securely; self-criticism without a concrete next step. It is working when your ratios hold steady across weeks.

4.15 Generalisation and maintenance

  • Purpose: make the change last outside sessions. Madsen's case treated the client at home, where the problem occurred, with intermittent short- and long-term follow-up. [source: Madsen 1999 address]
  • Steps:
    1. Work in the natural setting, or bring its people and cues into the session. [source: Madsen 1999 address]
    2. Thin the reinforcement schedules and move from contingent music to natural social reinforcers [synthesis].
    3. Train carers (4.11) and use the same musical cues in other settings (4.6) [synthesis].
    4. Schedule follow-up probes and record the same measures [synthesis].
  • Example language (illustrative): "We'll meet monthly now. Keep the tally going at home, and we'll check the graph together."
  • Watch for: relapse when the contingencies stop. Reinstate them briefly, then fade them more slowly.

5. Handling problems

Madsen's sources deal with problems in only two ways: change the procedure based on the data, and grade the demands (shaping). Most of the guidance below is therefore the writer's synthesis, marked [synthesis]. Where a source supports a point, it is cited.

  • No change in the data (stuck).
    • First check measurement: is the definition clear, and is inter-observer agreement above about 85% (Hanser's threshold as reported)? [source: Árkosi 2024]
    • Then check the reinforcer. Reassess preference (4.1), and make sure music is contingent rather than continuous and not diluted with other rewards. [source: Standley 1996 abstract]
    • Then check the response effort: shrink the shaping steps (4.9).
    • Change one variable at a time, and give each change 3 to 5 sessions [synthesis].
  • Satiation or a weakening reinforcer. Rotate the pieces and activities, shorten each music "dose", and offer a choice from a menu [synthesis].
  • Resistance or opposition to "being controlled".
    • Make the contingency collaborative: let the client help choose the target, the music and the reward, and show them their graph.
    • Frame it as a deal the client can renegotiate.
    • Teenagers often respond to access to playing real music, as in the guitar lessons of the 1968 case, more than to listening rewards. [source: Madsen 1999 address (case)] [synthesis (framing)]
  • Escalation or an extinction burst when music is withheld or interrupted.
    • Expect a temporary rise. Keep the rule consistent, keep everyone safe, and reinforce the first calm alternative.
    • If the escalation is dangerous, stop the procedure and move to a positive-only plan with a behaviour-support review [synthesis].
  • Overwhelm or sensory overload.
    • Lower the volume and density, and offer ear defenders and a quiet corner.
    • Use predictable musical structure (4.7), and keep within noise guidance. [source: MacMahon & Page 2015; TTU hearing sheet] [synthesis]
  • Musical trauma triggers.
    • If a client freezes, cries, dissociates or becomes agitated in response to a piece, stop the music. Ground the client in the room by naming the present and checking their feet and breathing. Do not probe for the memory.
    • Add the piece to the do-not-play list and review the reinforcer menu.
    • For trauma-exposed clients, screen music before using it as a reinforcer or cue [synthesis].
  • Non-adherence at home (carers stop recording or reinforcing).
    • Simplify to a single measure and a single rule, and use role-play rehearsal (4.11).
    • Agree a minimum practice, and ask what gets in the way (time, stress, disagreement between carers).
    • In Árkosi's case, a written agreement with the school structured the involvement of others. [source: Árkosi 2024] [synthesis]
  • Ruptures in the relationship.
    • Behavioural framing can feel cold. Name the rupture, ask what felt unfair (often a contingency the client experienced as punishment), and adjust.
    • Keep your approval ratio high. Madsen links high approval ratios with low off-task rates. [source: Madsen & Alley 1979 abstract] [synthesis]
  • Criticism that the approach is dehumanising.
    • Madsen acknowledges that psychodynamic clinicians have objected that behaviourism ignores complexity, deals only with minor motor acts, and dehumanises clients.
    • His reply is that "behaviour" includes emotional and cognitive responses, and that what matters is whether methods are scientifically tested. [source: Madsen 1999 address]
    • In practice: choose socially meaningful targets, involve the client in choices, and avoid restrictive procedures [synthesis].
  • Children and young people.
    • Get the parent's or guardian's consent and the child's assent for every contingency, especially any withdrawal or interruption of music.
    • Follow local safeguarding and mandatory-reporting duties. Disclosures can surface during home-based work like the 1968 case.
    • Use praise, visual tallies and short music rewards suited to the child's developmental level. Never use access to music, food or company as leverage that becomes deprivation [synthesis].
  • Clients with intellectual disability or communication needs.
    • Use choice trials rather than verbal questionnaires (4.1), and observe signs of distress closely, since these clients may not report noise discomfort.
    • Involve support staff and any behaviour-support plan [synthesis].
  • Tics, movement or medical conditions as targets. Barrett's (1962) contingent-interruption study targeted tics. [source: Madsen 1999 address] Today, treat tics and medical behaviours only alongside medical or neurological care, and never as wilful misbehaviour [synthesis].
  • Suicide or self-harm disclosures. Stop the behavioural programme and carry out a standard risk assessment and safety plan. This covers lethal-means counselling (safe storage or removal), crisis contacts and escalation to the responsible clinician. Behavioural contingencies must never be applied to self-harm disclosures [synthesis].
  • Working outside your scope. If you are not a credentialled music therapist (MT-BC in the US; see Professional frame), keep to praise, data and simple preferred-music rewards, and refer clients to a music therapist. [source: CBMT certification page (credential)] [synthesis (scope advice)]

Sources

Sources consulted:

Books and papers cited in the sources above (details as given there) [uncertain]:

  • Madsen, C. K., Cotter, V. A., & Madsen, C. H. Jr. (1968). A behavioral approach to music therapy. Journal of Music Therapy, 5(3), 69–71. doi:10.1093/jmt/5.3.69 (citation confirmed).
  • Madsen, C. K., & Madsen, C. H. Jr. (1968). Music as a behavior modification technique with a juvenile delinquent. Journal of Music Therapy, 5(3), 72–76.
  • Madsen, C. K. (1981). A behavioural music therapy guide for work with people with intellectual disabilities. National Association for Music Therapy (NAMT). Exact title unverified [uncertain] (the original title uses the outdated term of its time for this population).
  • Madsen, C. K., Greer, R. D., & Madsen, C. H. Jr. (1975). Research in music behavior. Teachers College Press.
  • Madsen, C. H. Jr., & Madsen, C. K. (1983). Teaching/discipline: A positive approach for educational development (3rd ed.). Contemporary Publishing.
  • Hanser, S. B. (1999). The new music therapist's handbook. Berklee Press. Cited via Árkosi (2024). A 3rd edition (2018) exists.
  • Barrett, B. H. (1962). Reduction in rate of multiple tics by free operant conditioning. Journal of Nervous and Mental Disease, 135.
  • Jeffrey, W. E. (1955). New technique for motivating and reinforcing children. Science.
← Previous · VIII-069Physio Acoustic TherapyNext · VIII-071 →Plurimodal music therapy