Sensory Integration 1 Running Head: Sensory Integration Effects of Sensory Integration On Behavior Maintained by Automatic Reinforcement A Thesis Presented By

......................................................................5 Acknowledgments..............................................................3 Introduction.....................................................................6 Method........................................................................17 1. Participants 2. Setting and Materials 3. Response Measurement 4. Pre-Treatment Analysis 5. Sensory Diet Treatment Results............................................................................22 Discussion........................................................................23 References........................................................................26 Figures.............................................................................32 Sensory Integration 5 Abstract Sensory integration implies that a number of the characteristics of autism are caused by poor sensory integration. Sensory integration refers to theoretical integration and interpretation of sensory stimulation from the environment by the brain. It proposes that in order for an individual to be able to function properly in the environment, he has to be able to organize all the sensations he experiences. From a behavioral perspective, some sensory integration treatments may be effective by abolishing the reinforcing sensory consequences of responses maintained by automatic reinforcement (e.g., stereotypy and self-injury). One individual who exhibited vocal stereotypy was exposed to a form of sensory integration therapy referred to as the sensory diet. Prior to treatment a functional analysis was conducted in attempt to identify the function of the individuals’ behavior. The subject displayed a pattern of responding suggesting that his behavior was maintained by sensory consequences produced by the behavior. Following the assessment, the sensory diet procedure was implemented in an ABAB design to determine whether sensory integration would have an abolishing effect on vocal stereotypy. The data showed that the sensory diet had no effect on vocal stereotypy maintained by automatic reinforcement. The data do not support the notion that sensory integration has a positive impact on vocal stereotypy.Sensory integration implies that a number of the characteristics of autism are caused by poor sensory integration. Sensory integration refers to theoretical integration and interpretation of sensory stimulation from the environment by the brain. It proposes that in order for an individual to be able to function properly in the environment, he has to be able to organize all the sensations he experiences. From a behavioral perspective, some sensory integration treatments may be effective by abolishing the reinforcing sensory consequences of responses maintained by automatic reinforcement (e.g., stereotypy and self-injury). One individual who exhibited vocal stereotypy was exposed to a form of sensory integration therapy referred to as the sensory diet. Prior to treatment a functional analysis was conducted in attempt to identify the function of the individuals’ behavior. The subject displayed a pattern of responding suggesting that his behavior was maintained by sensory consequences produced by the behavior. Following the assessment, the sensory diet procedure was implemented in an ABAB design to determine whether sensory integration would have an abolishing effect on vocal stereotypy. The data showed that the sensory diet had no effect on vocal stereotypy maintained by automatic reinforcement. The data do not support the notion that sensory integration has a positive impact on vocal stereotypy. Sensory Integration 6 Effects of sensory integration on behavior maintained by automatic reinforcement The term automatic reinforcement has been applied in developmental research to describe behavior that is not dependent on social reinforcement. Vollmer (1994) wrote that Skinner referred to automatic reinforcement as a “general class of behavior that might include anything from shooting a basketball to scratching one’s head.” For Skinner (1957), automatic reinforcement can be either a primary or secondary reinforcer. A primary reinforcer, sometimes called an unconditioned reinforcer, is a stimulus that functions as a reinforcer even though the organism has had no particular learning history with that form of stimulation. Stimuli such as food, water, and sexual stimulation function as primary reinforcers. Primary reinforcers are readily identified and most people react similarly to the primary reinforcers when they are contingently arranged. Another type of reinforcer is the conditioned reinforcer, sometimes called a secondary reinforcer. Conditioned reinforcers are previously neutral stimuli that have acquired the capability to function as a reinforcer through pairing with one or more primary reinforcers or other conditioned reinforcers (Cooper, Heron, & Heward, 2007). A conditioned reinforcers are thought to lose their reinforcing efficacy when they are no longer paired with backup reinforcers. With individuals with autism repetitive selfstimulatory behaviors such as hand flapping, head rolling, body rocking, and hair twirling may produce sensory stimulation that functions as automatic reinforcement. The response product that functions as automatic reinforcement is likely an unconditioned reinforcer (Smith, Michael, & Sundberg, 1996). Sensory Integration 7 For some individuals, problem behavior appears to be maintained by automatic reinforcement (Vollmer, 1994). The term has been used to describe reinforcement that is produced independent of the social environment (Vaughan & Michael, 1982). The term automatic does not specify whether behavior is maintained by positive or negative reinforcement contingencies (Vollmer). That is, automatic reinforcers can either serve as positive or negative reinforcers. Such responses develop independent of social consequences and are referred to as automatic because the behavior is maintained by the sensory consequence produced by engaging in the response. The specific source of reinforcement is often difficult or impossible to identify, manipulate, or control (Vollmer). When reinforcers maintaining a behavior are not within the control of the experimenter, behavior is particularly difficult to assess and treat. The development of treatment is often difficult because many behavioral interventions involve manipulations of the social environment, an approach that may be functionally irrelevant in the case of automatic reinforcement. Vollmer (1994) indicated that behavior maintained by automatic reinforcement must follow certain criterion such as: the behavior should persist in the absence of other stimulation (e.g., the alone condition of a functional analysis), social consequences do not maintain the behavior, access to the behavior should function as a reinforcer, and, when the automatically reinforced behavior is blocked it will result in response suppression. Furthermore, Vollmer stated that if functional analysis shows that a behavior occurs more often in the alone condition, it may be concluded that such behavior may be decreased by simply enriching the environment that the subject is in. Iwata, Dorsey, Slifer, Bauman, and Richman (1982) stated that when barren environment functions as an establishing operation, we could conclude that the Sensory Integration 8 consequences of the behavior serve the function of automatic positive reinforcement because the behavior is producing stimulation of certain type. A considerable amount of research conducted over the past years implies that stereotypic behavior tends to be maintained by the sensory consequence produced by engaging in the response. Though it has been suggested that stereotypy serves no function (Bodifish, Symons, Parker & Lewis 2000), a considerable amount of research conducted over the past years implies that stereotypic behavior tends to be maintained by the sensory consequence produced by engaging in the response (Rincover, 1978; Iwata 1994; Rapp, 2007). Previous research implies that stereotypic behavior tends to be maintained by automatic reinforcement (Vollmer, 1994). That is, the sensory consequence produced by engaging in the response is the functional reinforcer. Stereotypic behavior consists of repetitive behavior that has no apparent function in the environment. Examples of such behavior include‘ rocking, hand waving, and head waving (Kaufman & Levitt, 1964), mouthing or rubbing parts of one’s body (Berkson & Mason, 1964), and mouthing or spinning objects (Hutt & Hutt, 1965). Ahearn, Clark, MacDonald and Chung (2007) examined the non-contextual vocalizations of four children diagnosed with autism. For each of the participants, it was found through functional analysis that vocal stereotypy was likely maintained by sensory consequences produced by the behavior. Following the functional assessment, a response redirection procedure was used to determine whether stereotypy could be successfully redirected. The procedure consisted of issuing a series of vocal demands contingent on any vocal stereotypy, and continuing these vocal demands until the child successfully complied with three of them while not engaging in vocal stereotypy. For each individual the response/redirection procedure Sensory Integration 9 produced levels of vocal stereotypy that were lower than those in baseline. Also an increase in appropriate communication was detected for 3 of the 4 participants. A large number of studies have been conducted to identify stimuli that effectively compete with the problem behavior (e.g., Ahearn et al., 2005., Vollmer, Marcus, & LeBlanc, 1994). Piazza, Adelinis, Hanley, Goh and Delia (2000) evaluated the effects of providing access to matched stimulation non-contingently on problem behavior that was maintained by automatic reinforcement. Three children participated in the study. All three exhibited problem behavior that was automatically reinforced. Preference assessments were used to identify preferred leisure items that did or did not match the hypothesized sensory consequences o the problem behavior. Each participant was given non-contingent access to either highly-preferred matched or highlypreferred unmatched items. Results suggested that providing access to items that matched the hypothesized sensory consequences of aberrant behavior may be more effective than providing access to unmatched stimuli. Falcomata, Roane, Hovanetz, Kettering, and Keeney (2004) examined the utility of a procedure consisting of noncontingent reinforcement (NCR) with and without response cost in the treatment of inappropriate vocalizations maintained by automatic reinforcement. Following the functional analysis a treatment phase was implemented. The treatment consisted of comparison of baseline and two treatment conditions, NCR and NCR plus response cost. Noncontingent reinforcement is a procedure where stimuli with known reinforcing properties are presented independent of behavior (Cooper, Heron & Heward, 2007). In the study a functional analysis was conducted first to identify the reinforce that maintained inappropriate vocalizations. Following the functional analysis, a phase of consecutive alone sessions was implemented to evaluate further the occurrence of the target behavior. The treatment analysis consisted of comparison of baseline and two treatment conditions, NCR and NCR plus response Sensory Integration 10 cost. Baseline sessions were similar to the alone sessions for functional analysis, except that the therapist was present and a preferred item was placed in the room. During baseline the participant was not allowed to use the preferred item even though it was present. During NCR condition, the participant was given continuous access to the preferred item, and the therapist provided no consequences for occurrences of the target behavior. Although levels of inappropriate vocalizations decreased below baseline levels during NCR, the observed reduction was not clinically significant. Therefore, a response cost component was introduced to the NCR condition. In this condition the participant lost access to the preferred item for 5 s contingent on inappropriate vocalizations. In this study the use of response cost reduced the occurrence of the automatically reinforced behavior. The results of this study are similar to those in a previous study done by Mason and Iwata (1990) and suggested that for some individuals the loss of preferred stimuli may compete with engagement in automatic reinforced problem behavior. Researchers have also investigated sensory extinction for automatically reinforced behaviors and they have explored the role environment plays in self-stimulation (Rincover, Cook, Peoples, & Packard, 1979). In the Rincover et al. study, the topography of problem behavior for each child was observed in order to develop a hypothesis about sensory consequences that could be responsible for maintaining the behavior. Then, a sensory extinction procedure was created based on this hypothesis wherein preferred objects were removed in order to see the participant’s reactions. Afterward the participants played with toys that had similar oconsequences (which were observed to be highly reinforcing during sensory extinction) to the toys used during the sensory extinction procedure, and once again behavior was evaluated. It was reported that the first participant was stimulated by sounds during assessment. When auditory feedback was removed when carpeting was installed atop the table, stereotypy was near Sensory Integration 11 zero for 17 consecutive sessions. The participant played with a music box during treatment at an average of 81% of the time and ignored other toys. The second participant in this study was stimulated by touch. A small vibratory mechanism was taped to the back of a child’s hand, generating a low intensity, high frequency pulsation. When proprioceptive feedback was removed; stereotypy was immediately reduced to zero. The participant played with toys that provided proprioceptive stimulation at an average of 77% of the time , and ignored other toys. Stereotypy was maintained at low rates of behavior that averaged 18% across seven sessions. The third participant was stimulated by tangible items. The participant constantly picked feather, lint or a small string and then threw it in the air and waver hands vigorously trying to keep it afloat. This suggested that visual consequences may have reinforced the hand movement. When visual feedback was removed, stereotypy decreased from a 68% to near zero rates. The participant played with bubbles during treatment and ignored other toys. Stereotypy was maintained near zero rates. The forth participant was stimulated by multiple sensory events. The self-stimulation consisted of finger flapping. When visual feedback was removed, stereotypy decreased to an average of 77%. When visual and proprioceptive feedback were removed, stereotypy decreased to a mean of 34%. Rincover et al. suggested that better methods and treatments may be necessary to design effective interventions for individuals whose stereotypy behavior is supported by multiple senses. Overall, this study suggests that stereotypic behaviors can decrease if programming is created on the basis of sensory reinforcement. Vollmer (1994) states that manipulating motivating operations can be an effective option for treatment for automatically-reinforced behavior. According to Vollmer, altering motivating operations can be accomplished by response competition, enriching the environment or by matched forms of stimulation. Establishing operation refers to an event that alters the Sensory Integration 12 reinforcing efficacy of a stimulus as reinforcement (Michael, 1982). Satiation and deprivation are two examples of motivating operations. Satiation is the opposite of deprivation. The more frequently a person has received a particular reinforcement in the recent past, the more satiated she is. Deprivation is the frequency with which a person has received a particular reinforcement. The less frequent the reinforcer, the more deprived the person is. Vollmer and Iwata (1991) illustrated in their study that food, music and social praise were more potent as positive reinforcement when a short period of deprivation preceded skill acquisition training session. According to Vollmer, differential reinforcement procedures can be used to reduce automatically-reinforced problem behavior. When differential reinforcement is used reinforcement is delivered contingent on appropriate behavior (Vollmer 1994). Kennedy et al. (2000) studied behavioral functions associated with stereotypical responses for students with autism. In Study 1, functional analyses were conducted for 5 individuals. For all participants, stereotypy occurred across all assessment conditions, suggesting that the behavior is maintained by automatic reinforcement. In the second study the stereotypy of one of the individuals was further analyzed on a function by function base. Alternative responses were taught to be student using functional communication training (FCR). The results of the study showed that similar topographies of stereotypy were reduced only when differential reinforcement contingences for alternative form of communication were implemented for specific response-reinforcer relation. Sensory Integration It has been suggested that individuals with autism display problem behavior due to poor sensory integration (Ayres, 2005) The child with autism shows many of the symptoms of poor sensory processing that are seen in the child with minimal brain dysfunction and his interactions Sensory Integration 13 with the physical environment are corresponding poorly (Ayres, 2005). However, a child with autism has additional problems. A child with autism is often described as being “in a world of his own”, and he usually does not want others to intrude. If he learns to speak, his speech tends to be limited. Some children with autism display too little emotion. Others become very emotional, have tantrums and become aggressive. According to Ayers et al. (2005), there are three aspects of poor sensory processing that occur in children with autism. One, sensory input is not being registered correctly in the child’s brain, and so he pays little attention to his environment, while at other times he overreacts. Two, he may not modulate sensory input well and so he may be gravitationally insecure or tactilely defensive. Three, the part of his brain that makes him want to do things, especially new or different things, is not operating normally, and so he has little interest in doing things that may be useful in his life. The objective therapy for the autistic child is to improve the sensory processing so that more sensations will be more effectively “registered” and modulated, and to encourage the child to form simple adaptive responses as a means of helping him to learn to organize his behavior (Ayers). According to Ayers et al. (2005), the central idea of the sensory integration therapy is to provide and control sensory input in a way where the child spontaneously forms the adaptive responses that integrate those responses. Sensory integrative therapy is a specialty of occupational therapy, a profession that has emphasized understanding of human behavior from a neurobiological viewpoint. Most activities in sensory integration therapy are said to be purposeful, since the child has a goal in doing them. Doing such activities is the best way to improve human functioning when the problem lies in the way the brain is working. In order for the therapy to be effective a skilled therapist and large room with special equipment is required. Sensory Integration 14 The therapy is most effective when the child directs his own actions while the therapist unobtrusively directs the environment. Integration most often occurs when the child wants the stimulus and initiates an activity to get those sensations. Typically, sensory integration therapy is delivered in one-to-one sessions over extended periods of time (Ayres, 1972). Furthermore, Ayers explains why sensory integrative therapy (SI) is putatively effective. She says that because child brain is young it is flexible and capable of natural change. Sensory diet is a common SI treatment method where exposure to sensory items is provided during the day (Wilbarger, 1991). Sensory integration occurs in the nervous system and the main task of our central nervous system is to integrate the senses. In order to assume that this therapy alone was responsible for any apparent changes in the child, the changes would first need to be documented over time by objective methods. It would be also necessary to conduct controlled experiments to rule out several plausible alternatives to the possibility that SI per se produced the changes: placebo effects, maturation; the potential reinforcing effects of sensory stimulation, motor activities, and adult attention; escape from demands; and other potential treatments. However, observation that a child likes an activity is not sufficient evidence that the activity is producing significant, lasting changes in the child’s brain and or behavior, or any short-term benefits other than momentary pleasure. Most studies reporting that SI is effective, lack necessary controls to permit any alternative explanation to be ruled out (Bright, Bittick, & Fleeman, 1981; Lemke 1974; Wells & Smith, 1983). Many lack operational definition of the critical variables as well as objective measurement procedures and controls for observer bias. Lemke (1974) suggested that stereotypic behavior and self-injurious behavior seen in developmentally disabled individuals may be a reflection of poor sensory-motor integration. To Sensory Integration 15 support this statement Lemke presented an uncontrolled case study where multiple forms of stimulation were applied to an individual who exhibited self-injurious behaviors. The multiple forms of stimulation included: finger massage and ice to the mouth, quiet talk, tooth brushing, towel massage to the arms, patty cake, feet slapping, and body rolling. Unfortunately no quantitative data were presented, however the author noted that he subject was freed from restraint and learned to hold a toy in each hand, (one toy was mouthed the other was banged). In another case study, Bright et al. (1981) administered tactile, vestibular, and social stimulation. It was reported that their subject’s frequency of SIB decreased during treatment sessions. Physiological stimulation was presented concurrent with other events, most notably noncontingent social stimulation, that alone may have been responsible for reported changes in behavior. Wells and Smith (1983) studied the use of sensory integration to reduce SIB with four participants. SIB was assumed to be maintained by automatic reinforcement, however no functional analysis was conducted. Sensory integration treatment consisted of rocking in a hammock or chair, vibrator or massage, rolling on therapy ball, rolling large bolster over legs, back and shoulders. Results showed large decrease in SIB for all four participants. Although these results are good, this study, like many others lacked scientific rigor. Well controlled studies conducted by Dura, Mulick, and Hammer (1988), and Mason and Iwata (1990) found that SI was ineffective or no more effective than other treatments. Mason and Iwata (1990) studied the effects of sensory integration therapy on the self-injurious behavior of three children with profound disabilities. Prior to the treatment, a functional analysis baseline was conducted to identify what maintained the participants’ problem behavior. Three individuals who exhibited self-injurious behavior (SIB) were exposed to SI. One subject's SIB appeared to Sensory Integration 16 be an attention-getting response (maintained by positive reinforcement), which varied subsequently as a function of attention being either withheld or provided noncontingently during SI. The second subject displayed a pattern of responding suggestive of stereotypic SIB (maintained by automatic reinforcement), which paradoxically increased during SI. The third subject's SIB appeared to function as an escape response (maintained by negative reinforcement), and his behavior during SI was similar to that observed during baseline sessions in which demands were not present. The SIB of all 3 subjects later was reduced when functionallyappropriate behavioral interventions were applied. The data presented questions about the active components of SI and the functional types of SIB for which it might be appropriate. These questions that were presented are; , what are the effects of SIT as a treatment for SIB? Is it limited to the stereotypic function of SIB for which SIT was design? What are the active components of therapy? How would these components affect SIB that is maintained by environmental events? From a behavioral perspective, SI may be effective by abolishing the reinforcing sensory consequences of automatically reinforced behavior (Laraway, Snycerski, Michael, & Poling, 2003). Abolishing operations make reinforcement and punishers less effective. For example; food ingestion decreases the effectiveness of food as reinforces. The purpose of the present study was to identify behaviors maintained by automatic reinforcement and determine if sensory integration had any impact on behaviors that are maintained by sensory consequence. It is important to evaluate if Sensory Integration Therapy is effective for treating automatically maintained stereotypy. In this study an individual was exposed to a sensory diet where items that were shown to compete with stereotypy were Sensory Integration 17 provided at specific times during the school day. Vocal stereotypy was measured twice a day; in the morning and in the afternoon.