Functional Analysis and Noncontingent Reinforcement With Extinction in the Treatment of Perseverative Speech

Abstract

The present study evaluated the effectiveness of noncontingent reinforcement combined with extinction in the treatment of perseverative speech of an 11-year-old boy with autism spectrum disorder. Following a functional analysis that suggested perseverative speech was maintained by attention, treatment was introduced and consisted of delivering attention on a fixed-time schedule and placing perseverative speech on extinction. The intervention resulted in a 98.5% decrease in perseverative speech, which maintained over a 28-month period.

Noncontingent reinforcement (NCR) refers to the response-independent (or time-based) delivery of the reinforcers responsible for maintaining problem behavior and is an empirically supported behavioral treatment for various aberrant behaviors maintained by automatic (e.g., sensory stimulation, escape/avoidance) and social (e.g., attention, tangibles, escape/avoidance) reinforcement in individuals with developmental disabilities (Carr, Severtson, & Lepper, 2009; Richman, Barnard-Brak, Grubb, Bosch, & Abby, 2015).

NCR is most frequently used in the treatment of attention-maintained behavior (Carr et al., 2009). Buchanan and Fisher (2002) used an NCR intervention consisting of the fixed-time (FT) delivery of attention and music (identified as reinforcers from a prior functional analysis [FA]) for the treatment of disruptive vocalizations of two elderly dementia patients. For both participants, the NCR intervention resulted in a significant decrease in disruptive vocalizations.

Tomlin and Reed (2012) evaluated the effects of an NCR intervention on the disruptive (including verbal) behavior of four special-education classroom students. NCR consisted of the FT delivery of physical and/or verbal attention on an initially dense schedule of reinforcement, which was later thinned for each student. During NCR, students’ disruptive behavior was placed on extinction (EXT). For all students, the initial FT schedule resulted in a decrease in disruptive behavior, which maintained when the schedule of reinforcement was thinned.

Although NCR is effective in reducing problem behavior by itself, research indicates that when used as part of a treatment package or when combined with other behavior-reduction interventions such as EXT, treatment outcomes can be enhanced. More importantly, with respect to treatment relapse in the form of behavior resurgence, recent findings by Saini, Fisher, and Pisman (2017) comparing NCR with and without EXT showed greater behavioral resurgence in NCR without EXT (although immediate decreases in responding were observed with NCR without EXT), thus suggesting that NCR and EXT may be preferable to NCR alone.

The purpose of the present study was to evaluate the effectiveness of NCR with EXT for the treatment of perseverative speech suspected to be maintained by attention in a child with autism spectrum disorder (ASD). To the authors’ knowledge, only one study to date has investigated the effects of NCR as a treatment for attention-maintained perseverative speech (see Noel & Rubow, 2018).

Method

Participant, Setting, and Materials

Nathan was an 11-year-old male diagnosed with ASD by a neurologist at age 7. He communicated using full sentences and was able to follow multistep instructions. At the time of this study, Nathan was receiving in-home behavioral therapy based on applied behavior analysis (ABA) for the treatment of challenging behaviors and skill deficits. Therapy was provided after school (Nathan attended a public elementary school and was in a special education classroom), 5 days a week for 4 hr a day, by a Registered Behavior Technician (RBT), under the weekly supervision of a Board Certified Behavior Analyst (BCBA).

Approximately three months prior to the start of the study, Nathan began engaging in perseverative speech, which consisted of excessively talking about a restricted topic: “the Magical Land of Nathan,” an imaginary place where, according to Nathan, adults were not allowed. Previous interventions that were implemented included EXT, DRA, and a combination of the two. However, none were successful in reducing the behavior to a socially acceptable level (e.g., 80% reduction relative to baseline levels). Additionally, Nathan’s mother expressed concerns about implementing EXT (i.e., planned ignoring) with fidelity on her own due to the fact that Nathan’s perseverative speech often resulted in high-intensity disruptive behavior (e.g., screaming, object throwing, property destruction) when his mother ignored his behavior.

Sessions were conducted in Nathan’s home during the BCBA’s weekly supervision of his ABA-based therapy sessions, for 2 to 3 hr, once a week. Throughout the study, Nathan’s mother and Nathan’s RBT (both of whom had received training on all procedures and met the 90% fidelity criterion) interchangeably served as therapists and implemented procedures under the supervision of the BCBA (first author) during the FA and treatment evaluation. Materials used during sessions included various toys and leisure items (e.g., tablet, miniature toy animals, videos), digital timers, work tasks, and data sheets.

Dependent Variable, Response Measurement, and Interobserver Agreement

Perseverative speech was defined as any conversation related to “the Magical Land of Nathan,” and included any of the following verbalizations: “Do you know about the Magical Land of Nathan?” “Let me tell you about the Magical Land of Nathan,” or “Can I tell you about the Magical Land of Nathan?” The onset of perseverative speech was defined as the first instance of Nathan saying the words “Magical Land of Nathan,” and the offset was defined as 5 s without saying those words. Data were collected on perseverative speech using a 1-min and 2-min partial-interval recording method during the FA and treatment evaluation sessions, respectively. Data were reported as the percentage of intervals with perseverative speech. During the treatment evaluation, the session timer was paused during reinforcement delivery, and perseverative speech was not recorded (Nathan was allowed to engage in perseverative speech during the reinforcement period). The session timer resumed once Nathan stopped engaging in perseverative speech for 3 s.

A second observer (RBT) independently but simultaneously recorded the occurrence and nonoccurrence of perseverative speech in each interval during 50% and 100% of FA and treatment evaluation sessions, respectively. Interobserver agreement (IOA) was assessed using an interval-by-interval method and calculated by dividing the total number of intervals with agreement by the total number of intervals with agreement plus intervals with disagreement, then multiplying the result by 100%. IOA averaged 97.28% and 98.66% for FA and treatment evaluation sessions, respectively.

Functional Analysis

An FA based on procedures described by Iwata, Dorsey, Slifer, Bauman, and Richman (1994) was conducted to determine whether perseverative speech was maintained by social reinforcement, specifically attention. (Prior descriptive assessments and direct observations suggested that Nathan’s perseverative speech was maintained by social reinforcement in the form of attention.) Our FA consisted of twelve 5-min sessions and included the following four conditions conducted using a fixed-sequence multielement design: contingent attention, no interaction, free play, and contingent escape. Sessions were divided into 1-min intervals each. A 2-min transition period was added between sessions. There were no programmed consequences for perseverative speech during transition periods. A no-interaction condition was used in lieu of an alone condition given that Nathan was never left alone at home. Table 1 provides a detailed summary of each FA condition and its environmental manipulations.

Table 1.

Summary of Functional Analysis Conditions

FA Conditions	Arrangement/Manipulation of Antecedent and Consequence Variables
Contingent attention	• Nathan was provided with unrestricted access to moderately preferred toys. • The session started with the therapist stating, “I have some work to do,” turning away from Nathan, and pretending to write. • Contingent on perseverative speech, the therapist delivered verbal attention: “Not right now. You just told me about it” or “I don’t want to talk about the Magical Land of Nathan anymore.” • There were no programmed consequences for other problem behavior.
No interaction	• No toys were present, but Nathan was provided with restricted access to leisure items. • The session started with the therapist pretending to talk on the phone; there was no interaction with Nathan. • No demands were placed. • No attention was provided. • There were no programmed consequences for perseverative speech or other problem behavior. • The session was terminated if Nathan engaged in high-intensity disruptive behavior (e.g., object throwing, property destruction) with the mother as therapist.
Contingent escape	• No toys/leisure items were present. • The session started with the therapist stating, “Time to work,” and presenting task materials (e.g., math questions, reading). • Least-to-most three-step prompting was used. • Neutral, brief verbal praise followed compliance with a verbal or model prompt. • Contingent on perseverative speech, the therapist delivered a 30-s break and stated, “OK, let’s take a little break.”
Free play	• Nathan was provided with unrestricted, noncontingent access to highly preferred toys/leisure items (e.g., tablet, videos). • The session started with the therapist stating, “Time to play.” • No demands were placed. • Verbal and physical attention was delivered every 15–20 s. • Perseverative speech and all other problem behavior were ignored.

Treatment Evaluation

Baseline and treatment sessions were 10 min in duration, and each session was divided into 2-min intervals (a 2-min interval duration was selected for practical reasons due to the RBT’s prior experience collecting partial-interval data on former target behavior using this interval duration). A total of 12 sessions were conducted a day. Nathan’s mother assumed the therapist role and implemented the NCR procedure during 31% of treatment sessions. An A-B-A-B reversal design was used to evaluate treatment effects.

Baseline

Sessions were identical to those in the contingent-attention condition of the FA (i.e., contingent on every occurrence of perseverative speech, the therapist delivered verbal attention in the form of comments, and all other problem behavior was ignored).

Treatment (NCR + EXT)

Treatment consisted of fixed- and variable-time (FT and VT) delivery of reinforcement with schedule thinning and EXT. The initial schedule of reinforcement was selected based on the mean interresponse time for perseverative speech during baseline (2 min), and the schedule of reinforcement was thinned using a fixed-increment schedule. Thus, the initial schedule was FT 2 m and subsequent schedules were FT 5 m, FT 10 m, and VT 10 m. The criterion for schedule thinning was a minimum of five consecutive sessions with no more than 20% of intervals with perseverative speech. The NCR procedure consisted of the therapist delivering 90 s of verbal (i.e., conversational) attention; in other words, Nathan had 90-s access to the therapist’s undivided attention, which included uninterrupted conversations about the restricted topic. At the beginning of the reinforcement delivery period, the therapist said to Nathan, “OK, Nathan, (tell me / time to talk) about the Magical Land of Nathan,” and allowed him to talk about the restricted topic for 90 s. At the end of the reinforcement period, the therapist stated, “OK, we’re done talking about the Magical Land of Nathan for now,” and placed perseverative speech on EXT. EXT was implemented during each session, except during reinforcement delivery. Moderately preferred items and work tasks were available during treatment sessions; however, the therapist never placed any demands on Nathan and did not interact with Nathan unless Nathan initiated task-related interactions (e.g., requested help with a work task, requested a different task).

Results

The top panel of Fig. 1 shows the results of the FA. In the contingent-attention condition, perseverative speech occurred during 100% of intervals. In the no-interaction condition, perseverative speech occurred during a mean of 6.66% of intervals. In the contingent-escape and free-play conditions, perseverative speech occurred during means of 13.33% and 53.33% of intervals, respectively. The percentage of intervals with perseverative speech was highest in the contingent-attention condition and lowest in the no-interaction condition, thus confirming our hypothesis that the behavior was maintained by social reinforcement, specifically attention.

Fig. 1.

Percentage of 1-min intervals with perseverative speech during the FA (top), and percentage of 2-min intervals with perseverative speech during the NCR + EXT treatment evaluation (bottom).

The bottom panel of Fig. 1 shows the results of the treatment evaluation. During baseline, the mean percentage of intervals with perseverative speech was 97.14%. When treatment (NCR-FT 2 min + EXT) was introduced, the mean percentage of intervals with perseverative speech decreased to 22.66%. During the reversal to baseline, the mean percentage of intervals with perseverative speech increased to 60%. When treatment (NCR-FT 2 min + EXT) was reintroduced, perseverative speech occurred during a mean of 8.57% of intervals. When the schedule of reinforcement was thinned to FT 5 min, followed by an FT 10-min schedule, perseverative speech further decreased to mean occurrences of 3.33% and 3.33% of intervals, respectively. Finally, during the final schedule-thinning condition, when attention was delivered on a VT 10-min schedule, perseverative speech occurred during a mean of 1.42% of intervals. Overall, the NCR + EXT intervention resulted in a 98.53% reduction in perseverative speech, relative to baseline.

Discussion

The present study evaluated the effectiveness of an NCR + EXT intervention for the treatment of perseverative speech maintained by attention in an 11-year-old boy with ASD. NCR combined with EXT was effective at decreasing perseverative speech to near-zero levels. Our findings corroborate previous research on NCR as an empirically supported treatment for problem behavior maintained by social reinforcement (Buchanan & Fisher, 2002; Carr et al., 2009; Richman et al., 2015; Tomlin & Reed 2012). Our findings are also consistent with Noel and Rubow’s (2018) findings on NCR as an effective intervention for reducing perseverative speech.

The present study also emphasizes three key considerations with regard to the effectiveness of NCR as a behavior-reduction intervention. The first consideration is with respect to identification—and subsequent response-independent delivery—of known reinforcers. Research has shown that NCR is more effective when the reinforcer responsible for maintaining the problem behavior is identified using empirical methods (i.e., FAs). This was accomplished in this study by conducting an FA.

The second consideration is combining NCR with other interventions. Research indicates that NCR can be combined with other interventions to enhance treatment efficacy. In the present study, NCR + EXT resulted in a significant decrease in perseverative speech. Prior to this intervention, previous interventions used to reduce perseverative speech consisted of EXT alone, DRA alone, and a combination of EXT + DRA, none of which resulted in the significant decrease achieved using NCR + EXT. Furthermore, although a verbal rule (to end perseverative speech) was provided to Nathan at the end of the reinforcement period, it should be noted that it is unlikely that the rule itself contributed to the enhanced treatment efficacy given that verbal rules were also included in the previously attempted (and unsuccessful) DRA, EXT, and DRA + EXT interventions.

Last, previous NCR studies have shown that decreases in problem behavior as a result of denser FT schedules of reinforcement can be effectively maintained during and after schedule thinning, which our findings corroborate (the reductive effects of the FT schedule maintained during schedule thinning and across the RBT and parent). We continued maintenance checks for approximately 12 months after the end of the study, during regular ABA-based therapy sessions, and perseverative speech remained at near-zero levels (those data are not displayed on the treatment evaluation graph). We also assessed social validity 28 months after the end of the study to evaluate the extent to which Nathan’s mother found the treatment to be effective and socially significant (see Table 2). Social validity was assessed via a 15-question rating scale ranging from strongly disagree to strongly agree within the following three areas: the social significance of goals, the social appropriateness of procedures, and the social importance of treatment effects (Ferguson et al., 2019).

Table 2.

Social Validity Assessment

Social Significance of Goals	Caregiver Response
The behavior was problematic enough to warrant immediate behavioral intervention.	Strongly agree
Decreasing the frequency of the behavior was important to my family and me.	Strongly agree
The behavior was easy to manage on my own.	Strongly disagree
I was concerned about the negative social effects of the behavior.	Strongly agree
I found the behavior to be disruptive to our family’s quality of life.	Strongly agree
Social Appropriateness of Procedures (NCR + EXT)	Caregiver Response
I found the intervention to be acceptable.	Strongly agree
I found the intervention to be easy to implement.	Strongly agree
I would have preferred a different intervention from the intervention used.	Disagree
The intervention caused major disruptions to my family’s daily routine.	Disagree
I would recommend this intervention to other caregivers.	Strongly agree
Social Importance of Observed Effects	Caregiver response
The NCR + EXT intervention resulted in negative side effects.	Strongly disagree
I am satisfied with the outcome of the NCR + EXT intervention.	Strongly agree
There has been a significant decrease in the frequency of the behavior as a result of the NCR + EXT intervention.	Strongly agree
Decreasing the behavior has had a positive impact on my family’s daily quality of life.	Strongly agree
I am better equipped to manage the behavior as a result of treatment.	Strongly agree

Although the present study demonstrated the effectiveness of NCR + EXT in the treatment of perseverative speech, three main limitations are worth noting. First, although our study demonstrated the enhanced treatment effects of combining NCR with EXT (as opposed to using EXT alone, as was done prior to this intervention), the results presented in this study cannot determine with empirical certainty whether NCR enhanced the reductive effects of EXT or vice versa, due to the fact that the present study did not actually evaluate the efficacy of NCR both with and without EXT. Nevertheless, our findings are consistent with Saini et al.’s (2017) research findings suggesting that it may be preferable to implement NCR + EXT.

Second, although Nathan did have a behavioral acquisition component in place in his behavior intervention plan (which was implemented during his ABA-based therapy sessions, separate from the study), we did not formally teach Nathan a replacement behavior once perseverative speech decreased to near-zero levels. Given that DRA is also an empirically supported intervention in reducing aberrant behavior while reinforcing alternative desired behavior, we advise clinicians to consider combining DRA with NCR to enhance treatment effects.

Last, the type of attention delivered in the contingent-attention condition of the FA was slightly different from that used during the treatment evaluation. Function-based interventions (e.g., NCR, EXT, differential reinforcement) for the treatment of problem behavior require that the function of the problem behavior be identified, and that the consequence found to be maintaining the problem behavior be delivered (i.e., NCR) or withheld (i.e., EXT). In the present study, the consequence delivered in the contingent-attention condition of the FA was attention in the form of adult refusal to engage in perseverative speech. However, the consequence delivered noncontingently during the treatment evaluation was attention in the form of access to conversations about the restricted topic. Given that one of the mechanisms underlying the effectiveness of NCR is satiation (which can be achieved via repeated and abundant exposures to the reinforcer), we hypothesized that attention in the form of access to conversations about the restricted topic would be a more potent reinforcer than adult refusal, and its noncontingent delivery would function as an abolishing operation, thus decreasing its potency as a reinforcer (DeLeon, Williams, Gregory, & Hagopian, 2005). Nevertheless, we recommend that practitioners carefully identify which consequences should be tested in an FA and whether or not the same consequences should be used in their function-based treatment.

Despite these limitations, we encourage practitioners to explore NCR as a practical intervention for attention-maintained problem behavior such as perseverative speech.

Compliance with Ethical Standards

Conflict of interest

The authors declare they have no conflict of interest.

Ethical approval

Ethical approval was waived by the Chicago School of Professional Psychology’s Institutional Review Board in view of the retrospective nature of the study, and all the procedures performed were part of the participant’s routine applied behavior analysis–based behavioral therapy.

Informed consent

Written informed consent was obtained from the participant’s legal guardian.