RIRD vs. DRO: Vocal stereotypy 1 A Comparison of Differential Reinforcement of Other Behavior and Response Interruption and Redirection on Vocal Stereotypy A Thesis Presented By

Vocal stereotypy is a common problem behavior among persons with autism spectrum disorder and can prove to be a fairly difficult behavior to treat. One of the main reasons that this behavior is often challenging is that it is usually maintained by the sensory consequences that it produces. Response interruption and redirection and differential reinforcement of other behavior are two treatments which have been shown to be effective in reducing stereotypy. RIRD is often a staff intensive procedure which can be difficult to implement and may not always be done with 100% procedural integrity. While DRO is an easier procedure for staff to implement it is not always as effective when used on its own. This study is a comparison of RIRD and DRO in the treatment of vocal stereotypy using an ABAB design, alternating the treatments during the B phase. Three students with autism spectrum disorder that exhibit high rates of vocal stereotypy were chosen as participants. Stereotypy decreased with both treatments for all three participants. For one participant RIRD was more successful throughout the entire study, while for the other two DRO was at times more effective but not consistently. Results for both DRO and RIRD are effective treatments for decreasing stereotypy in persons with autism spectrum disorder with RIRD as a possibly more reliable treatment. RIRD vs. DRO: Vocal stereotypy 6 A Comparison Of Differential Reinforcement of Other Behavior and Response Interruption and Redirection on Vocal Stereotypy Vocal stereotypy, defined as any instance of non-contextual or non-functional speech, including singing, babbling, repetitive grunts, squeals and phrases unrelated to the present situation (Ahearn, Clark, MacDonald & Chung, 2007) is among the criteria for autism spectrum disorder. Individuals with autism often display some form of repetitive behavior (Lewis & Bodfish, 1998) that occurs frequently and at very high rates, and has been associated with impaired learning and social development (Kennedy, Meyer, Knowles, & Shukla, 2000). Such behavior is socially stigmatizing (Jones, Wint, & Ellis, 1990). Despite its prominent presence for this population it has received less attention in research then some of the other behaviors that are also on the diagnostic description of this disorder (Lewis & Bodfish, 1998). Researchers have assessed the consequences that may be maintaining stereotypy (Kennedy et al., 2000). For example, Durand and Carr (1987) and Mace and Belfiore (1990) found that stereotypy was maintained by escaping particular social situations. Kennedy et al. found that the function of stereotypy was multiply maintained by escape and attention and that stereotypy decreased from baseline levels when a functional communication response was taught. Numerous researchers who have conducted functional analyses found that stereotypy persisted in the absence of social mediation (e.g., Piazza, Adelinis, Hanley, Goh, & Delia, 2000; Rapp, Miltenberger, Galensky, Ellingson, & Long, 1999; Vollmer, Marcus, & LeBlanc, 1994) suggesting that it is automatically reinforced by the sensory consequences it produces. RIRD vs. DRO: Vocal stereotypy 7 Lovaas, Newsom, Litrownick, and Hickman (1987) found that participants in their study brought in a large variety of different types of high-rate, persistent selfstimulatory behaviors with near-identical forms across children from diverse cultures (Asia, Latin America, Europe). Because these behaviors continued to occur in the absence of social consequences the authors concluded that participants’ stereotypy were not maintained by social reinforcement. Because the behavior’s maintaining variables have implications for treatment it is important that the function of stereotypy is determined before treatment development. Iwata, Dorsey, Slifter, Bauman, and Richman (1982/1994) illustrated a functional analysis procedure determining the function of selfinjurious behavior. Automatically reinforced behavior is defined as behavior that is maintained by operant mechanisms independent of the social environment (Vaughn & Michael, 1982). Because problem behavior maintained by automatic reinforcement are not dependent on social reinforcement, it can be more difficult behavior to assess and treat (Vollmer, 1994). One of the difficulties facing researchers is that most treatments that have been informed by a functional analysis have been for behavior that is socially mediated. These interventions are highly effective but usually involve withholding the reinforcer that is maintaining the problem behavior (Vollmer). Automatic reinforcement is maintained by a variable that is not within the experimenter’s control to withhold. One idea is that behavior occurs in an environment where the participant is alone because the environment is bare and the consequences of the behavior are producing stimulation. In this theory, an enriched environment may reduce rates of behavior (Lovaas et al.). Another treatment option for automatically maintained problem behavior RIRD vs. DRO: Vocal stereotypy 8 is to provide an alternative source of stimulation that produces a similar sensory reinforcement (Vollmer). Differential reinforcement is one of the most often used treatments for stereotypy and other problem behaviors in persons with developmental disabilities (Marcus & Vollmer, 1996). A basic DRO contingency refers to reinforcement in the absence of a specific behavior after a specified interval of time. A DRO procedure may involve a reinforcer which is the maintaining variable for the problem behavior or it may involve an arbitrary reinforcer. Although DRO has been shown to be more effective if the maintaining reinforcer is used (Mazaleski, Iwata, Vollmer, Zarcone, & Smith, 1993), arbitrarily selected reinforcers are frequently used. There are two types of DRO; whole interval DRO and momentary DRO (Cooper, Heron, & Heward, 2007). In whole interval DRO the behavior must be absent for the entire interval in order for reinforcement to be delivered. In momentary DRO the behavior need only not occur at the moment that the interval ends. Repp, Barton, and Brulle (1983) found in a comparison of interval DRO and momentary DRO that interval DRO was more effective in reducing problem behavior than momentary DRO. Barton, Brulle, and Repp (1986) found that after behavior was successfully reduced using whole interval DRO, low levels could be maintained using momentary DRO. One possible alteration to the DRO contingency is to use a variable interval instead of a fixed interval. In the variable interval the time of the interval changes throughout a treatment session. Other types of differential reinforcement can include DRA (differential reinforcement of alternative/incompatible behavior) or DRL (differential reinforcement of low rates of behavior). Iwata and Pace (1990) found that DRO was effective in reducing self injurious behaviors without being combined with RIRD vs. DRO: Vocal stereotypy 9 other treatments and rates of behavior were reduced as effectively as when those responses were exposed to punishment. The effectiveness of differential reinforcement may depend on several different variables. Some of these variables are the amount and immediacy of the reinforcement that is being delivered and whether or not the reinforcer is also the maintaining reinforcer or an arbitrary one (Lerman, Kelley, Vorndran, & Van Camp, 2003). Differential reinforcement can also be used in combination with other treatments, and has been found in some cases to be more effective this way. Fellner, Laroche, & SulzerAzaroff (1984) found that DRO/DRI was considerably more effective in lowering rates of stereotypy when interruption was added. Barton, Repp, and Brulle (1985) found that when combining DRO and momentary restraint on four individuals with stereotypic behaviors, rates of stereotypy decreased rapidly. Sensory extinction is another treatment that has been fairly successful in reducing rates of automatically maintained behavior (Rincover, 1978). Sensory extinction consists of disrupting the contingency between the response and the product it produces (Rincover). If perceptual consequences reinforce self-stimulatory behaviors, the behaviors previously maintained by such consequences should decrease in strength when the consequences are removed. That is, the behavior should exhibit extinction (Lovaas, Newsom, Litrownik, & Hickman, 1987). Response blocking is in some cases is considered a type of sensory extinction, although the response is prevented from continuing as opposed to no longer producing the stimulation that it once did (Lalli, Livezey, & Kates, 1996). RIRD vs. DRO: Vocal stereotypy 10 Response blocking has been successful in reducing problem behavior which is not socially mediated. Reid, Parsons, Phillips, and Green (1993) found that self injurious hand mouthing was significantly reduced in two individuals with profound disabilities when blocking was introduced. Haerris and Wolchick (1979) found that when comparing DRO, timeout and overcorrection in 4 individuals with developmental disabilities for the treatment of stereotypic behavior, overcorrection had the most dramatic reduction in problem behavior for all four individuals. A potential side effect of response blocking is that it can induce aggression or responses that are in the same class as the one being blocked (Lerman et al., 2003). Lerman et al. found that when response blocking was used as a treatment there were undesirable side effects that arose in the participant’s behavior. Hagopian and Adelinis (2001) conducted a study in which pica was blocked with and without redirection. During the blocking and redirection condition, the participant was redirected to eat popcorn contingent on pica. During the combined condition, pica decreased to lower levels and aggressive behavior was observed. By contrast, higher levels of aggression occurred when only blocking was conducted. It is also possible that response blocking reduces behavior as a form of punishment. Lerman and Iwata (1996) evaluated response blocking when it was applied intermittently. For instance, in one phase behavior was blocked every fourth response. If blocking was reducing behavior by extinction it would be on a fixed ratio reinforcement schedule and therefore would not decrease in levels. It is also possible that blocking functioned as a punisher and if so, blocking every fourth response would decrease behavior. Lerman and Iwata found that for this particular participant levels of behavior RIRD vs. DRO: Vocal stereotypy 11 decreased to near zero levels even when they only blocked a portion of the responses, suggesting that blocking functioned as punishment as opposed to extinction. Vocal stereotypy presents a problem for clinicians who want to use response blocking because vocalizations cannot be physically blocked. Ahearn et al. (2007) interrupted vocal stereotypy and redirected the participant to emit appropriate vocalizations. Four participants were used in this study and were chosen for their high rates of vocal stereotypy. Each instance of vocal stereotypy was interrupted with a series of vocal demands until the participant complied with three in the absence of stereotypy. Data was taken on percentage of vocal stereotypy and percentage of appropriate vocalizations. An ABAB design was used and results showed that for all four participants stereotypy decreased to near zero levels during treatment and returned to baseline levels when treatment was withdrawn. For three out of the four participants appropriate vocalizations increased from baseline levels as well. Response interruption and redirection has been shown to be effective in reducing stereotypy in persons with an autism spectrum disorder; however it is an extremely staff and labor intensive procedure. DRO is a less staff intensive procedure to implement than RIRD, however it has not been proven to be as effective in reducing stereotypy. The purpose of this study is to replicate the study conducted by Ahearn et al. (2007), as well as compare the efficacy of differential reinforcement of other with response interruption and redirection in reducing rates of vocal stereotypy in persons with autism spectrum