Abstract
This study examined four interventions targeted at decreasing multiply controlled vocal stereotypy for a 12-year-old boy diagnosed with autism spectrum disorder and a severe intellectual disability. These interventions included Noncontingent Music, Differential Reinforcement of Other Behaviors, Self-Recording, and Functional Communication Training (FCT). In addition to measuring vocal stereotypy during each condition, task engagement and challenging behavior were also monitored. Across conditions, vocal stereotypy did not vary significantly from baseline except in FCT, when it decreased significantly. Task engagement was higher in this condition as well. It is hypothesized that FCT provided an enriched environment by increasing social interaction and access to desired items as well as removal of less preferred activities. For these reasons, there was a decrease in the need for the participant to engage in vocal stereotypy and challenging behavior and increase in his ability to engage in a task.
Skinner (1957) coined the term verbal behavior to reflect the fact that human vocalizations are typically maintained by socially mediated consequences delivered by the listener. However, many people engage in vocalizations without the possibility of socially mediated consequences, such as humming or speaking aloud when alone. In most instances, such vocalizations do not have a detrimental effect on social interaction or engagement with appropriate activities; therefore, these are considered socially acceptable vocal behaviors. However, there are some topographies of vocalizations that do not produce socially mediated consequences, yet can be negative for the individual engaging in them. These vocal behaviors may be socially stigmatizing, interfere with daily functioning, or decrease opportunities for learning (Matson, Minshawi, Gonzalez, & Mayville, 2006). Children with autism spectrum disorder (ASD) typify this as the disorder is characterized by an increased display of restricted, repetitive behaviors, which includes vocal stereotypy (American Psychiatric Association, 2013).
Vocal stereotypy is typically defined topographically in terms of repetitive vocal behaviors (Ahearn, Clark, MacDonald, & Chung, 2007). This includes repetition of previously heard words or phrases (e.g., Mancina, Tankersley, Kamps, Kravits, & Parrett, 2000), repetition of nonword sounds such as musical instruments or animal noises (e.g., Falcomata, Roane, Hovanetz, Kettering, & Keeney, 2004), or bodily function nonword sounds such as moaning, grunting, or throat clearing (e.g., Ahearn, Clark, Gardenier, Chung, & Dube, 2003). However, some controversy exists regarding the definition of vocal stereotypy, as it only involves the topography of the behavior. Some argue that vocal stereotypy should be defined by the function maintaining it (Cunningham & Schreibman, 2008). Others go so far as to indicate that vocal stereotypy, by definition, must be maintained by nonsocial consequences, meaning it serves no function other than automatically reinforcing the individual (Ahearn et al., 2007; Rapp & Vollmer, 2005). There is no clear answer in the literature as to whether topography or function is the best definition for vocal stereotypy; regardless, research has forged ahead to develop treatment packages addressing inappropriate vocalizations, especially for children with ASD.
Lanovaz and Sladeczek (2012) provided an overview of behavioral interventions to treat vocal stereotypy among children with ASD. It was found that several antecedent- and consequence-based interventions have reported success in treating these behaviors. Antecedent-based interventions included a variety of formats using noncontingent access to reinforcement or similar environmental enrichment strategies (LeBlanc, Patel, & Carr, 2000). Frequently cited interventions included the use of noncontingent matched stimulation (Lanovaz, Fletcher, & Rapp, 2009; Lanovaz, Sladeczek, & Rapp, 2011; Love, Miguel, Fernand, & LaBrie, 2012). For example, music is considered a matched stimulation because it provides auditory feedback similar to that produced by vocal stereotypy. Studies have demonstrated that providing noncontingent access to music produced reductions in vocal stereotypy (e.g., Lanovaz et al., 2009). The two most frequently cited consequence-based interventions included response interruption and redirection (RIRD; for example, Ahearn et al., 2007; Love et al., 2012) and differential reinforcement of other behaviors (DRO; for example, Taylor, Hoch, & Weissman, 2005), both of which have demonstrated effectiveness in reducing different classes of challenging behavior.
Although Lanovaz and Sladeczek (2012) identified intervention practices with support for effectiveness in the literature, they also reported that few studies directly compared existing interventions. As a result, the ability for practitioners to select one treatment over another is limited, which may increase the amount of time vocal stereotypy persists while exhausting a list of intervention options and decreasing academic and social learning opportunities. Since the publication of the Lanovaz and Sladeczek (2012) review, some studies have compared two or more treatments to further refine the treatment selection process (Lanovaz et al., 2014; Rapp et al., 2013; Saylor, Sidener, Reeve, Fetherston, & Progar, 2012). However, comparative studies for vocal stereotypy interventions are relatively limited, prompting the design of this study, which was developed in an effort to better assist practitioners in making informed decisions about treatment options, including the use of Functional Communication Training (FCT). The effectiveness of four interventions was evaluated in one child with ASD including (a) Noncontingent Music, (b) DRO, (c) Self-Recording, and (d) FCT.
Method
Participant
Elton was a 12-year-old Caucasian male diagnosed with ASD and a severe intellectual disability. He communicated using a speech generating device (SGD) application on an iPad® and was capable of manding for two reinforcers using this application. Elton frequently engaged in vocal stereotypy, which was a loud, distinctive whining or grunting sound. In addition to stereotypy, Elton displayed aggressive and self-injurious behaviors that resulted in weekly visits to a physician.
Setting
All sessions were conducted at a university-based applied behavior analysis (ABA) clinic, where Elton received therapy twice weekly. Sessions were conducted in typical therapy rooms that consisted of a table, chair, and experiment-specific materials, including work tasks (i.e., puzzles, blocks), reinforcers, two iPads®, and a BIGmack SGD. Two to three experimenters were present in the room during sessions to implement the interventions and collect data.
Experimental Design
A modified reversal design (i.e., ABACADAEAE) was implemented in this study of which the following conditions were evaluated for Elton: (a) Baseline, (b) Noncontingent Music, (c) DRO, (d) Self-Recording, and (e) FCT.
Measurement and Interobserver Agreement
Data were collected on vocal stereotypy, task engagement, and challenging behavior. Vocal stereotypy was defined as any nonword vocalization above typical speaking volume or pitch. Task engagement was defined as sitting at the table and having physical contact with work task items (e.g., puzzle piece, matching cards, etc.). Challenging behavior included pinching, hitting, or kicking objects, self, or others as well as spitting and throwing objects. Data were collected on all three classes of behavior using a 10-s partial interval procedure across all conditions of the study.
Interobserver agreement (IOA) was calculated using an interval-by-interval method. The number of intervals in which both observers agreed (occurrence plus nonoccurrence) was divided by the total number of intervals (agreements plus disagreements) and multiplied by 100 to report a percentage. IOA was conducted on 63% of the functional analysis. Mean IOA was 94% (range = 87%-100%). IOA was measured on 27% of sessions comparing the four interventions under investigation as well as baseline. Mean IOA was 91% (range = 80%-98%) for vocal stereotypy, 93% (range = 83%-100%) for task engagement, and 94% (range = 85%-100%) for challenging behavior.
Procedure
Functional behavior assessment
To identify the function of vocal stereotypy, a functional behavior assessment was conducted with Elton. First, the Questions About Behavioral Function (QABF; Paclawskyj, Matson, Rush, Smalls, & Vollmer, 2000) was administered to his parents for information related to Elton’s vocal stereotypy. Following this, a functional analysis was conducted using procedures similar to Iwata, Dorsey, Slifer, Bauman, and Richman (1982/1994) to further verify the function maintaining Elton’s vocal stereotypy.
Intervention comparison
To compare treatments, the participant was exposed to five conditions: (a) Baseline, (b) Noncontingent Music, (c) DRO, (d) Self-Recording, and (e) FCT. Each condition lasted 10 min in duration.
Baseline conditions represented typical ABA therapy sessions Elton engaged in each week. Elton’s therapy sessions aimed to teach him pre-academic and academic skills, such as sorting, matching, and puzzle completion. Therefore, his baseline represented the presentation of three work tasks per learning trial with least-to-most prompting (Duker, Didden, & Sigafoos, 2004) used immediately following the discriminative stimulus. Verbal praise was provided contingent on task completion. In addition to verbal praise, access to his highest preferred item, a hand towel, was provided on a fixed ratio (FR) 3 schedule for task completion. All vocal stereotypy was ignored.
During the Noncontingent Music condition, the exact procedures of baseline were repeated with one exception; music was played at a low volume on an iPad® placed on a shelf in the back of the room. The volume was loud enough so that it was easily heard by the participant and experimenters, but was not enough to interrupt typical conversation. The volume on the iPad® was set to approximately 50% of total output capability. Songs from a popular children’s television show were played during Elton’s sessions.
Music was selected via parent and therapist interviews and information from preference assessments conducted previously with Elton. Although research indicates individuals with developmental disabilities can successfully participate in preference assessments including the selection of preferred music (Horrocks & Higbee, 2008), Elton engaged in significant challenging behavior when administered preference assessments in the past. In addition, very few reinforcing stimuli had been identified in the several years he had been receiving ABA therapy. Therefore, the music selected for this study was based on parent and therapist reports that he appeared to enjoy the music from a popular children’s television show as compared with other pieces of music.
The DRO condition followed the same procedures as baseline including access to the same tangible reinforcer (i.e., a hand towel) on the same FR 3 schedule. However, Elton received additional tangible reinforcement contingent on the absence of vocal stereotypy. The duration of the DRO schedule was fixed at 20 s, which was selected given that his baseline levels of vocal stereotypy occurred approximately 2.5 times per minute. The DRO-specific reinforcer was selected via pairwise preference assessment (Fisher et al., 1992) conducted previously outside of this study. After a period of 20 s without vocal stereotypy, Elton was given 20 s of access to the preferred pompom and told, “Good job being quiet.” When vocal stereotypy occurred during the interval, the timer was reset to 0 s and restarted.
During Self-Recording, the procedures of baseline were repeated with one exception. A pre-recording of Elton’s vocal stereotypy was played at a low volume on an iPad® placed on a shelf in the back of the room. The volume of the pre-recording was the same as in the Noncontingent Music condition. The recording was played on a loop so that there was no lapse in play.
During FCT, Elton was taught to request a free-time break using a BIGmack SGD through procedures outlined by Carr and Durand (1985). The SGD was programmed to emit the pre-recorded message, “no.” To begin the session, a work task was placed on the table in front of Elton, and the experimenter prompted him to begin the task. Immediately, progressive time delay was utilized to teach Elton to emit the pre-recorded message. During initial training, there was a 0-s delay in which a full physical prompt was used. The duration of the delay was systematically increased by 3 s after every fourth trial. If an error occurred, the delay returned to 0 s. Contingent on prompted and independent SGD use, the work task was removed, and Elton was given access to the same preferred item used in baseline sessions for 15 s. All other procedures were identical to baseline, and all vocal stereotypy was ignored.
RIRD was slated for inclusion in this study, as it is commonly used to treat vocal stereotypy (e.g., Ahearn et al., 2007; Love et al., 2012). One session was conducted in which the same procedures were followed as baseline, except, contingent on the display of vocal stereotypy, the response was interrupted by a disapproving statement from the experimenter, followed by the experimenter prompting the participant to complete a specific task. When Elton engaged in vocal stereotypy, the experimenter interrupted the sound and said, “Elton, that sound is not okay” and then prompted Elton to complete two gross motor imitation responses, utilizing least-to-most prompting (Duker et al., 2004) as necessary. Verbal praise was provided contingent on the completion of motor imitation. Redirection tasks were selected based on skills within the participant’s repertoire. Although the experimenters wished to make the redirection skills as relevant to the topography of the vocal stereotypy and activity at hand as possible (e.g., requiring Elton to repeat a phrase relevant to the activity), Elton’s limited repertoire of skills, especially verbal behavior, including echoics, narrowed the selection of redirection activities. Elton’s vocal stereotypy during this condition was within the range displayed during baseline; however, the severity of his challenging behavior in response to the experimenter’s redirection markedly increased, posing significant risk of injury to self and others. To maintain the safety of the participant and experimenters, no other RIRD sessions were conducted. Because there was only one datum point, the RIRD session was not graphed.
Results
Functional Behavior Assessment
The results of the QABF (Paclawskyj et al., 2000) not only suggested that Elton’s vocal stereotypy indicated an automatic function but also suggested multiple sources of control as stereotypy was elevated across all measured functions. The Nonsocial function received the highest score followed by Escape.
The results of the functional analysis are presented in Figure 1. Vocal stereotypy was high and varied across all conditions. The mean percentage of intervals with vocal stereotypy was 76, 60, 57, 48, and 17 across demand, attention, tangibles, ignore, and play conditions, respectively. The levels of vocal stereotypy exhibited across conditions led to the conclusion that it was multiply maintained with both social and nonsocial functions.
Results of the functional analysis for Elton’s vocal stereotypy.
Evaluation of Interventions
Vocal stereotypy
Figure 2 shows Elton engaged in high and variable levels of vocal stereotypy across baseline conditions (M = 42%; range = 22%-67%). Vocal stereotypy remained at similar levels (M = 40%; range = 23%-53%) during the Noncontingent Music sessions. It was anecdotally noted that Elton did respond to the music in that his facial expression changed when the music was first turned on, and he walked around the room searching for the origin of the sound. However, the iPad® was placed on a shelf out of sight due to previous instances in which Elton would engage in challenging behavior when the iPad® was not available for free play. Vocal stereotypy during the DRO condition was also similar to that of the Baseline condition, with a slightly lower mean (30%) but similar range (22%-43%). However, vocal stereotypy was slightly reduced during the Self-Recording condition (M = 18%; range = 8%-32%). Finally, vocal stereotypy during the first FCT condition initially demonstrated a slight increase relative to the immediately previous baseline but then exhibited a decreasing trend in vocal stereotypy (M = 29%; range = 3%-60%). In the second condition of FCT, vocal stereotypy was markedly lower than all previous baseline and intervention conditions (M = 5%; range = 2%-8%).
Elton’s vocal stereotypy across conditions.
Task engagement
From Figure 3, it is apparent that Elton engaged in moderate and variable levels of task engagement in all baseline conditions (M = 40%; range = 16%-60%). Engagement slightly decreased (M = 35%; range = 25%-42%) during the Noncontingent Music sessions. This slight reduction may have been a result of Elton’s occasional attempts to locate the source of the music. Task engagement during DRO was also similar to that of baseline conditions (M = 37%; range = 23%-48%). However, task engagement was slightly higher during the Self-Recording condition (M = 46%; range = 31%-60%). Finally, task engagement during the first FCT condition was noticeably higher than all previous conditions (M = 46%; range = 31%-60%). In the second condition of FCT, task engagement was again higher than previous conditions and also demonstrated stability (M = 79%; range = 78%-80%). It is worth noting that task engagement was higher during FCT conditions despite frequent task breaks provided contingent on the target communicative response.
Elton’s task engagement across conditions.
Challenging behavior
Figure 4 shows Elton engaged in varied levels of challenging behavior across all baseline conditions (M = 34%; range = 5%-65%). Challenging behavior notably decreased, relative to the first Baseline condition, during the Noncontingent Music condition (M = 19%; range = 8%-27%). Challenging behavior was similar to baseline conditions in both the DRO condition (M = 20%; range = 5%-27%) and the Self-Recording condition (M = 23%; range = 12%-52%). Finally, challenging behavior demonstrated a steady decrease during the first FCT condition (M = 15%; range = 0%-15%). In the second condition of FCT, challenging behavior was again lower than previous conditions and also demonstrated stability (M = 2%; range = 0%-6%).
Elton’s challenging behavior across conditions.
Discussion
The present study found that vocal stereotypy was moderated to a large extent by teaching functional communication skills. Furthermore, it filled a gap in the literature by evaluating several interventions in a child with ASD whose vocal stereotypy had multiple maintaining contingencies. Although three of the four interventions did not decrease Elton’s vocal stereotypy, matched stimulation (Lanovaz et al., 2009; Lanovaz et al., 2011) and DRO (Taylor et al., 2005) have been found efficacious in various contexts. It took a considerably long period of time for the utility of FCT to be demonstrated, which suggests that the other interventions may not have been implemented long enough to see an overall decrease in vocal stereotypy. Perhaps Elton was not exposed to the contingencies associated with each of the conditions for a sufficient duration, accounting for the sustained levels of vocal stereotypy across those conditions. However, there may be a variety of other factors that influenced the outcomes of this research.
The first of these factors relates to the nature of the intervention itself. To the authors’ knowledge, no other study has examined the effects of FCT on vocal stereotypy. One hypothesis as to the success of the intervention for Elton is that FCT may have provided him a type of enriched environment by which there were increases in social interaction, access to desired items, and removal of less preferred activities. There is additional evidence to support this notion when considering the decrease in his challenging behavior. However, it is possible that FCT was effective because Elton’s vocal stereotypy also served a socially mediated function, including escape from work demands. There is a considerable body of research demonstrating the reduction of socially mediated challenging behaviors when functional communication is taught (Carr & Durand, 1985). However, it is noteworthy that although Elton requested a break frequently, he was more engaged in work tasks than in other conditions. In other words, Elton’s work did not actually decrease during FCT; in fact, engagement increased. Perhaps both hypotheses have merit considering that FCT reduced vocal stereotypy that appears to have socially and nonsocially mediated functions.
A second factor to consider in relation to the results of this study is the definition of vocal stereotypy. Notably, there is debate as to whether vocal stereotypy is defined by the topography of the behavior or the function (Cunningham & Schreibman, 2008; Rapp & Vollmer, 2005). To address this discrepancy, the topography of Elton’s vocal stereotypy was scrutinized to develop a comprehensive operational definition for use throughout this study, and a functional behavior assessment was completed. The results of the functional analysis suggested multiple maintaining functions, both socially and nonsocially mediated, for Elton’s vocal stereotypy. However, it may have been the case that Elton was unable to discriminate between the contingencies of the functional analysis conditions, skewing the data and subsequent analysis. It is evident that the topography of Elton’s vocal stereotypy appeared consistent with the literature in its manifestation, although the function of vocal stereotypy remains less than clear.
A third and final factor affecting the results of this study is Elton. In addition to an ASD diagnosis, he had a severe intellectual disability that impaired his functioning to a significant degree. He needed very distinct cues and considerable repetition to learn the contingencies of conditions. Furthermore, Elton’s vocal stereotypy and challenging behaviors often interfered with learning opportunities. For example, Elton historically refused work tasks and engaged in high levels of aggression when presented with nonpreferred tasks. It is not surprising then that the one session of the RIRD condition conducted resulted in higher intensity aggressive behaviors in comparison with other conditions. In hindsight, it is obvious that this type of intervention would result in this behavior based on his past performance in academic and clinical settings. However, this was not considered prior to implementing the intervention, and the safety concerns this caused were alarming enough to end the RIRD condition after one session. Yet other researchers have successfully used RIRD to decrease and eliminate vocal stereotypy (Ahearn et al., 2007; Love et al., 2012). Considering individual differences during assessment and program development may be particularly salient for clinicians working with children who engage in vocal stereotypy, specifically those with multiple sources of control.
Although this research provides initial groundwork for the evaluation of interventions in treating multiply controlled vocal stereotypy, there are limitations that need to be addressed. First, this study was conducted with one child, and FCT has not been used to treat vocal stereotypy before. Additional research is needed to determine the effects of FCT and the hypothesized enriched environment. Second, it is unclear as to the operant mechanisms responsible for the decrease in vocal stereotypy during FCT conditions, as this intervention may not be function-based. It is possible that FCT was indeed a function-based intervention (i.e., escape from a demand), but it remains unclear given results of the functional analysis. Future research examining alternative methods of identifying the function of vocal stereotypy may be prudent.
Third, a preference assessment to determine preferred items used during conditions was not conducted as part of this study. Parent and therapist reports as well as previously conducted preference assessments were used to inform the selection of reinforcers and music. Fourth, this study did not evaluate vocal stereotypy in a condition with noncontingent access to preferred stimuli, and future research should consider a comparison of noncontingent reinforcement and FCT as interventions. Fifth, although FCT worked to decrease Elton’s vocal stereotypy, children and adults with other disabilities and levels of functioning may benefit from one of the other interventions or combinations of interventions tested, specifically those individuals with an identifiable function for vocal stereotypy. Therefore, it is suggested that future research examine treatment selection with individual differences in mind. FCT may have a novel application in treating multiply maintained or escape-maintained vocal stereotypy, yet more research is needed to determine the features of the intervention that do so.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.