Bi-Directional Translational Research: Behavioral Momentum Theory and Treatment Relapse

2. Bi-Directional Translational Research:Behavioral Momentum Theory and Treatment Relapse F. Charles Mace Nova Southeastern University Mailman Segal Center Autism Institute

3. Overview • Traditional concept of response strength • Overview of BMT • Treatment relapse and factors leading to it • BMT inspired models of treatment relapse • Clinical demonstration of two relapse models • Basic and clinical research demonstrating two possible solutions to treatment relapse

4. Traditional Concept of Response Strength Skinner (1938) • Reinforcement ‘strengthens’ behavior—extinction ‘weakens’ behavior • What is observed? • Response rate during reinforcement • Changes in response rate during extinction • The problem—baseline response rate and resistance to extinction are not highly correlated • Implication—both terms are related to reinforcement but not entirely explained by rates of reinforcement

5. Two Distinct Aspects of Behavior Nevin (JEAB, 1974) • BL multiple VI 1-min, VI 3-min schedule • Different rates of dark-key food presentations between components • Changes in response rate over successive periods of EXT • Resistance to EXT • Expressed as: • Absolute response rates • Proportion of BL response rates • What is represented by the area under the curve in the left panel?

6. Behavioral Momentum • Metaphor developed by John A. (Tony) Nevin • Refers to the tendency for a pattern of behavior, once established, to persist despite some opposition to the response-reinforcer relationship • Physical momentum is the tendency for an object in motion to continue at the same velocity unless opposed by a physical force. • Momentum = mass x velocity

7. Behavioral Momentum • In the behavioral momentum metaphor, behavioral momentum is the product of behavioral mass and behavioral velocity • Behavioral velocity is equal to (baseline) response rate • Behavioral mass is the resistance of the baseline response rate to change when the response-reinforcer relationship is disrupted

8. Response-Reinforcer Disruptors • Extinction • Satiation • Dark key food (periods between components) • Punishment • Alternative reinforcement • DRA • FT or VT reinforcer deliveries • Distraction

9. Nevin, Tota, Torquato & Shull (1990) Purpose? • To determine whether resistance to change (behavioral mass) is a function of R-S or S-S contingencies • Contingent baseline reinforcement supplemented by 2 forms of alternative reinforcement • Added time-contingent reinforcer deliveries • Added response-contingent reinforcer deliveries • Arrangement permits direct tests of the R-S vs. S-S hypotheses

10. Foods/hr 45 Foods/hr 15 Overall Foods/hr Nevin et al. JEAB, 1990 60 Condition A Experiment 2 • Baseline—multiple concurrent schedule of reinforcement • Hypotheses: • If RTC ƒ of R-S, then A = B < C • If RTC ƒ of S-S, then A = C > B L R Ø 15 15 B L R Ø 60 60 C L R

11. Foods/hr 45 Foods/hr 15 Overall Foods/hr Nevin et al. JEAB (1990) 60 Condition A Experiment 2 Findings • If RTC ƒ of R-S, then A = B < C • If RTC ƒ of S-S, then A = B > C • Results: A = B > C • Conclusion: RTC ƒ of S-S L R Ø 15 15 B L R Ø 60 60 C L R

12. Enduring Limitations of ABA Interventions Osnes & Leiblein (2003); Wacker (2008) • Treatment Relapse • Failure to Generalize

13. Treatment Relapse • Recurrence of problem behavior following successful treatment • Reflects a failure to maintain or generalize treatment effects • A commonly reported phenomenon that is under-assessed (Nevin & Wacker, in press; Osnes & Lieblein, 2003)

14. Factors Related to Treatment Relapse • Lapses in Treatment Integrity • Resumption of reinforcement of problem behavior • Failure to maintain reinforcement of alternative behavior • Failure to maintain schedule of noncontingent reinforcers • Reintroducing client to the context previously correlated with reinforcement of TB • Introducing a client to a context not correlated with treatment

15. Behavioral Momentum Theory and Treatment Relapse • Treatment relapse reflects the persistence of problem behavior when conditions change • General and robust findings: • Persistence is a function of the total reinforcement correlated with a specific context • higher context-reinforcer rate greater persistence • BMT may stimulate effective technologies to reduce or avoid treatment relapse

16. Models of Treatment Relapse Inspired by Behavioral Momentum Theory • Resurgence • Reinstatement • Renewal

17. Models of Treatment Relapse Inspired by Behavioral Momentum Theory Resurgence Paradigm • A target response is established • It is then placed on EXT while reinforcing an alternative response (analogous to DRA with EXT) • If reinforcement of the alternative is subsequently diminished, there is a resumption of the target response

18. Models of Treatment Relapse Inspired by Behavioral Momentum Theory Reinstatement Paradigm • A target response is established • It is then eliminated with EXT or DRA • Subsequent VI/VR reinforcement of the target response results in greater relapse with high rate DRA vs. low rate DRA • Subsequent VT/FT deliveries of the maintaining reinforcer results in recovery of the target behavior despite ongoing EXT

19. Models of Treatment Relapse Inspired by Behavioral Momentum Theory Renewal Paradigm • A target response is eliminated in a context different from the baseline context • Recovery of the target response is observed with the reintroduction of the baseline context despite ongoing EXT

20. Clinical Implications of Behavioral Momentum Theory Does alternative reinforcement (DRA/FCT) increase the persistence of problem behavior?

21. Clinical Implications of Nevin et al. (1990) Foods/hr 45 Foods/hr 15 Overall Foods/hr 60 Condition A Problem Behavior Communication L R Ø 15 15 B L R Ø 60 60 C L R

22. Mace, McComas, Mauro, Progar, Taylor, Ervin & Zangrillo (JEAB, 2010) • Clinical demonstration of the persistence-strengthening effects of DRA • An animal model of a possible solution to the problem • Clinical demonstration of one possible solution

23. Can DRA Increase the Persistence of Problem Behavior?A Test of the Resurgence Model Participants and Target Behaviors • Tom—7 yr, Down’s Syndrome—food stealing • Jackie—4 yr, severe MR—aggression • Andy—7 yr, autism & severe MR—hair pulling Baseline Reinforcement of Problem Behaviors • Tom—access to stolen food—CRF schedule • Jackie—access to edibles—VI 60-s schedule • Andy—access to attention—VR3 schedule DRA • Baseline rates of reinforcement for problem behaviors remained in place • Tom—food contingent on appropriate eating—VI 300 s schedule (165% of BL) • Jackie—food contingent on appropriate toy play—VR 2 schedule (185% of BL) • Andy—attention contingent on appropriate toy play—VI 30 s schedule (195% of BL) Extinction • Response blocking and withholding food (Tom/Jackie) and attention (Andy)

24. Can DRA Increase the Persistence of Problem Behavior?A Test of the Resurgence Model Participants and Target Behaviors • Andy—7 yr, Autistic Disorder—hair pulling • Tom—7 yr, Down’s Syndrome—food stealing • Jackie—4 yr, Autistic Disorder—aggression Baseline Reinforcement of Problem Behaviors • Andy—access to attention • Tom—access to stolen food • Jackie—access to edibles Functional Communication Training (FCT) • Andy—attention contingent on pointing to FCT card (195% of BL) • Tom—food contingent on pointing to FCT card (165% of BL) • Jackie—food contingent on pointing to FCT card (185% of BL)

25. Can DRA Increase the Persistence of Problem Behavior?A Test of the Resurgence Model Participants and Target Behaviors • Andy—7 yr, Autistic Disorder—hair pulling • Tom—7 yr, Down’s Syndrome—food stealing • Jackie—4 yr, Autistic Disorder—aggression Baseline Reinforcement of Problem Behaviors • Andy—access to attention • Tom—access to stolen food • Jackie—access to edibles Functional Communication Training (FCT) • Andy—attention contingent on pointing to FCT card (195% of BL) • Tom—food contingent on pointing to FCT card (165% of BL) • Jackie—food contingent on pointing to FCT card (185% of BL) Extinction • Response blocking and withholding attention (Andy) and food (Tom/Jackie)

28. Can anything be done to avoid or reduce the persistence-strengthening effects of FCT? General Strategy • Avoid adding reinforcement to the context in which problem behavior has or will be reinforced

29. An Animal Model for Reducing Persistence Subjects • 4 experimentally naïve Charles River rats Baseline • 3-component multiple concurrent schedule • 3 cycles of 6-min components • C1—EXT (left), 24 hr (right) • C2—96 hr (left), 24 hr (right) • C3—96 hr (left), EXT (right) Hypothesis • Can training the alternative response in a separate setting reduce the persistence-strengthening effects of DRA?

30. An Animal Model for Reducing Persistence Extinction • Single session of 6 blocks of the mult conc schedule • C3 compound stimulus • Presents stimuli correlated with the same rft rates as C2 (DRA) • Analogous to introducing FCT to the setting in which problem behavior has been reinforced

32. Clinical Demonstration: FCT in a Separate Setting Reduces Persistence Participants and Target Behaviors • Paul—21 yr, severe MR—severe disruptive beh • Terry—6 hr, autism—disruptive behavior • FA showed disruption maintained by escape from tasks Baseline Reinforcement • Yellow gowns—reinforcement of disruption only • Blue gowns—reinf of communication and disruption • White gowns—reinf of communication—separate setting Functional Communication Training (FCT) • Baseline rates of reinforcement for problem behaviors remained in place • Prompts to communicate with FC cards presented every 20 s schedule • Reinforcement of communication was 375% of reinforcement of disruption

33. Clinical Demonstration: FCT in a Separate Setting Reduces Persistence Participants and Target Behaviors Paul—21 yr, severe MR—severe disruptive beh Terry—6 hr, autism—disruptive behavior FA showed disruption maintained by escape from tasks Baseline Reinforcement Yellow gowns—reinforcement of disruption only Blue gowns—reinf of communication and disruption White gowns—reinf of communication—separate setting Functional Communication Training (FCT) Baseline rates of reinforcement for problem behaviors remained in place Prompts to communicate with FC cards presented every 20 s schedule Reinforcement of communication was 375% of reinforcement of disruption

34. Clinical Demonstration: FCT in a Separate Setting Reduces Persistence Baseline Reinforcement Yellow gowns—reinforcement of disruption only Blue gowns—reinf of communication and disruption White gowns—reinf of communication—separate setting Functional Communication Training (FCT) Baseline rates of reinforcement for problem behaviors remained in place Prompts to communicate with FC cards presented every 20 s schedule Reinforcement of communication was 375% of reinforcement of disruption Extinction Task prompts presented every 10-s regardless of disruptive behavior White--Yellow stimulus compound—White therapist provides task instruction in the Yellow setting Response to mands—’I’m glad you asked for a break, you can have one after you finish your work’

35. Reducing the Magnitude of Treatment Relapse Using Low-Rate Reinforcement F. Charles Mace Nova Southeastern University Duncan Pritchard and Marguerite Hoerger Bangor University and Aran Hall School Brian Harris, Lisa Kovacs and Heather Penny Aran Hall School

36. Study Hypotheses Comparatively low-rate VI VT schedules will reduce the magnitude of treatment relapse following: 1. Reinstatement of baseline reinforcement (reinstatement model) 2. Abatement of VI VT treatment (EXT) (resurgence model)

37. Method Participant and Setting • 17 yr-old male with Autistic Disorder and Severe Mental Retardation Target Behaviors and Data Collection • Aggression—hitting, pinching, kicking, biting, spitting, forceful pushing, throwing objects at the therapist • Disruption—throwing objects, tearing paper • Prompted mands for attention—pointing to a PECS card following a combined vocal/gestural prompt • Unprompted mands for attention—single-word initiations of interaction

38. Method—Experimental Phases and Conditions Functional Analysis • Attention—VI 60-s attention for aggression and disruption delivered by Therapist 1 and Therapist 2 • Play—near continuous noncontingent attention delivered by Therapist 1 and Therapist 2 VI VT Attention at Different Rates (plus EXT) • Therapist 1—On a quasi-random variable 30-s schedule, therapist approached participant and prompted use of PECS card. If no response, therapist praised engagement in a pro-social activity or provided noncontingent attention • Therapist 2—same as Therapist 1 but on VI VT 120-s schedule Equal Reinforcer Reinstatement • Both therapists reinforced aggression and disruption on VI 60-s schedules Extinction (Reinforcer Abatement) • No attention plus response blocking as needed

41. Specific Conclusions • Treatment relapse and generalization failure are enduring limitations of Applied Behavior Analysis • Basic behavioral research has developed models of treatment relapse • Evidence suggests these models can be translated into new clinical practices that can improve outcomes

42. Specific Conclusions • Basic BMT research predicted that DRA would increase the persistence of PB in clinical populations • An animal model suggests that training an alternative response in a separate context can avoid or reduce the persistence-strengthening effects of DRA • The animal predicted that FCT training in a separate context before introducing FC to the context in which PB occurs can avoid/reduce persistence of PB • Reductions in treatment relapse achieved by comparatively low-rate reinforcement is counter-intuitive

43. General Conclusions • Translational research is vital to the ‘health’ of basic research and technology development • Consideration of basic research can stimulate new and more effective technologies • Consideration of socially relevant human problems can stimulate basic research with relevance and greater value to the culture • Collaborative and bi-directional translational research may be especially profitable