Educational Articles - Summary

Focus on Autism and Other Developmental Disabilities 2015, Vol. 30(1) 13 –22 © Hammill Institute on Disabilities 2014 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1088357614525663 focus.sagepub.com

Article

Children diagnosed with Pervasive Developmental Disorders (PDD) often exhibit stereotypic behaviors such as rocking, spinning, hand flapping, and excessive move- ments (American Psychiatric Association, 2013). Professionals have used various procedures to eliminate these response classes, including differential reinforcement of alternative behaviors (e.g., Vollmer, Iwata, Zarcone, Smith, & Mazaleski, 1993), noncontingent reinforcement (e.g., Lalli, Casey, & Kates, 1997), and response blocking (e.g., McKerchar, Kahng, Casioppo, & Wilson, 2001).

Sensory Integration Therapy (SIT; e.g., Ayres, 1979) has been one approach utilized in the treatment of stereotypic behaviors of children with autism. “Sensory integration is a clinical frame of reference for the assessment and treatment of persons who have functional disorders in sensory pro- cessing” (Parham & Mailloux, 1996, p. 307). These “func- tional disorders” are believed to be manifested in overt behavioral challenges, such as problems with sensory dis- crimination, perception, proprioception, tactile discrimina- tion, visual perception, and vestibular processing (Parham & Mailloux, 1996). An unregulated sensory system may also inhibit the acquisition of positive skills, such as aca- demics and socialization competencies (e.g., O’Brien & Pearson, 2004).

Grandin (1992) posited that “deep pressure,” part of SIT, can provide a “calming effect” for persons with PDD, since some believe that persons with autism display high levels of

arousal (e.g., Hardy, 1990). Other methods of providing “sensory input” include adding weight to vests and back- packs (e.g., VandenBerg, 2001) and brushing parts of the body (e.g., Stagnitti, Raison, & Ryan, 1999).

Many researchers have attempted to experimentally examine the effect of sensory integration procedures (see reviews of this literature by Arendt, MacLean, & Baumeister, 1988; Baranek, 2002; Daems, 1994; Hoehn & Baumeister, 1994; Miller, 2003; Parham et al., 2007). The general opin- ion of these reviewers is that there is no consistent evidence demonstrating a causal relationship between SIT and posi- tive changes across numerous dependent variables (e.g., Parham et al. 2007; Shaw, 2002). For example, VandenBerg (2001) examined the effect of a weighted vest to increase on-task behavior of children diagnosed with attention-defi- cit/hyperactivity disorder (ADHD). Four children between the ages of 5 and 10 years participated in the study. All had a diagnosis of ADHD and were receiving school-based occupational therapy services due to a “sensory modulation problem” (p. 623), described as “excessive movement”

525663 FOAXXX10.1177/1088357614525663Focus on Autism and Other Developmental DisabilitiesSniezyk and Zane research-article2014

1Crossroads Center for Children, Schenectady, NY, USA 2Endicott College, Beverly, MA, USA

Corresponding Author: Carolyn J. Sniezyk, Crossroads Center for Children, 1136 North Westcott Road, Suite 100, Schenectady, NY 12306, USA. Email: [email protected]

Investigating the Effects of Sensory Integration Therapy in Decreasing Stereotypy

Carolyn J. Sniezyk, MS, BCBA1 and Thomas L. Zane, PhD, BCBA-D2

Abstract Sensory Integration Therapy (SIT) is a popular treatment for Pervasive Developmental Disorders that involves therapists using various strategies and manipulanda to provide sensory stimulation to improve behavioral dysfunctions. Although SIT is popular, the research literature demonstrates little experimental proof of effectiveness. Many published studies find little to no causal relationship between SIT and improvements in target behaviors. There are numerous internal and external validity threats that preclude confidence in a functional relationship for those studies that report positive changes. The current study attempted to evaluate the impact of different SIT techniques on the behavioral excesses of children diagnosed with autism, while using research designs that adhered to commonly accepted standards for internal and external validity controls. The results showed that there was no causal relationship between the sensory procedures and improvements in the targeted dependent variables. Thus, SIT remains an unproven treatment for autism.

Keywords autism spectrum disorders, sensory integration, stereotypy

at UNIV OF NORTHERN COLORADO on March 19, 2015foa.sagepub.comDownloaded from

14 Focus on Autism and Other Developmental Disabilities 30(1)

characterized by constantly moving, touching body parts, being out of seat without authorization, laying and rolling on the floor, and demonstrating extreme difficulty in sus- tained attention and task completion. Experimenters mea- sured the total time on-task during fine motor activities in their classrooms. The intervention consisted of the children wearing a denim vest to which were added weights of up to 5% of the child’s body weight.

VandenBerg used a single-subject “AB” sequence of phases (Cooper, Heron, & Heward, 2007), with data first being taken without, and then with, the children wearing the vests. There were six sessions each of baseline and inter- vention. Results seemed to show an increase of on-task behavior for each child, based on comparisons of child per- formance to a celeration line indicating statistical signifi- cance, as determined by a standard deviation of two or greater. VandenBerg thus concluded that the weighted vest was an effective intervention to decrease sensory-based problem behaviors in children with ADHD.

However, there were several potential limitations of this study. First, the use of an AB sequence limits confidence in any claim of a functional relationship between the interven- tion and changes in subject behavior (e.g., Cooper et al., 2007). Second, although observers were trained to take the on-task data, this training took place prior to the beginning of the study, and there were no inter-observer reliability checks done during the baseline and intervention condi- tions. Thus, there is no confirmation of the accuracy of observations during the actual implementation of the weighted vest procedures. In addition, the observers were not blind to when the experimental condition was used, raising the possibility of “expectancy bias” as a threat to interval validity. Finally, the determination of significant results was based on statistical analysis, rather than a visual inspection of data for which there is a requirement of stabil- ity over time.

Fertel-Daly, Bedell, and Hinojosa (2001) examined the effect of a weighted vest on five young children ages 2 to 4 years, diagnosed with PDD. The researchers measured attention to task, duration of different stereotypic behaviors, and the number of distractions exhibited by the children during a brief activity. Using a reversal design (Cooper et al., 2007), the experimenters had the children alternate between wearing and not wearing the vest. Each of the five participants showed lower rates of attention in baseline, higher rates during intervention, and reduced rates in the second baseline phase. Only one subject showed the same reversal pattern for the duration of stereotypic behaviors; that is, there was a failure to replicate the higher baseline phase with four of the five subjects.

Although this study seemed to support the hypothesis that the weighted vest was functionally related to improve- ments in these particular behaviors, the possibility of con- founding variables influencing the results cannot be ruled

out. For example, Fertel-Daly et al. (2001) noted that increased attention after a weekend could have been due to the children being in a more calm and structured context (i.e., school) than over the weekend. Most disconcerting, however, is the fact that reliability data on the dependent variables, and procedural fidelity data were not reported, and the raters were not blind to the conditions being implemented.

Van Rie and Heflin (2009) conducted a study empirically investigating the causal relationship between sensory pro- cedures and positive changes in learning. The rationale for this study was the possible relationship between sensory dysfunction and learning. An individual engaging in stereo- typy and other repetitive behaviors may do so because of difficulty in processing sensory input as is typically done. An “optimal level of arousal” is needed for learning, and stereotypy may interfere with this. Thus, Van Rie and Heflin focused on using sensory procedures in an attempt to pro- duce this optimal state of arousal and assess whether the students then learned better. Four young children (aged 6–7 years) diagnosed with autism participated in the study. All these children exhibited behaviors that could be interpreted as being sensory based, such as bounding, looking away from ongoing tasks, jumping up from the floor or chairs and running away from instructional activities, walking on tip- toes, and engaging in verbal stereotypy. Using an alternat- ing treatments design, Van Rie and Heflin counterbalanced implementing a control condition of a teacher reading a story to the child, with sensory activities, of swinging or bouncing on a ball, during 5-min sessions. The authors measured the percentage of correct responses the children made on selected academic tasks (receptive or expressive identification of community helper and safety signs flash- cards) during instructional sessions immediately following the implementation of the independent variables.

The authors reported mixed results. A functional rela- tionship between a sensory procedure and positive learning was demonstrated for two of the four children. One student showed better learning after the swinging intervention, and the other student improved after bouncing on the ball. For the other two children, no clear functional relationship could be shown. In addition to showing some positive results, Van Rie and Heflin provided an excellent model of state-of-the-art experimental methodology by building into their design commonly accepted research standards, such as the measurement of both dependent measure and proce- dural fidelity reliability, social validity assessments, a strong experimental design, random assignment of partici- pants to conditions, and rigorous data collection and interpretation.

Schaaf et al. (2013) conducted a recent test of sensory integration procedures with 32 children diagnosed with autism spectrum disorder ranging in ages between 4 and 8 years. Using a manualized sensory integration intervention,

at UNIV OF NORTHERN COLORADO on March 19, 2015foa.sagepub.comDownloaded from

Sniezyk and Zane 15

the researchers implemented the treatment 3 times per week during hour-long sessions, for 10 weeks. Dependent mea- sures included standardized assessments in the sensory and cognitive areas, and a general goal attainment assessment. The researchers reported significantly higher scores on the outcome variables after treatment, providing support for the sensory procedures used. A strength of this study involved the measurement of fidelity of implementation of the proce- dures. A potential limitation involved assessing progress through the use of standardized assessments instead of direct measurement of target behaviors hypothesized to be affected by the sensory procedures.

In summary, although sensory procedures are widely used clinically (e.g., Olson & Moulton, 2004), there is a lack of experimental literature using well-designed research protocols supporting the effectiveness of sensory integra- tion procedures on challenging behaviors exhibited by chil- dren with autism (see Smith, Mruzek, & Mozingo, 2005, for a review of this field). Dubouloz, Egan, Vallerand, and von Zweck (1999) noted that, in the past, some members of the occupational therapy field felt threatened by a focus on making treatment decisions based on sound research. Currently, there seems to be a clear trend in occupational therapy, at least in the number of articles published about “evidenced-based practice” (EBP), that this conceptual framework is becoming the guiding principle in practice (Ottenbacher, Tickle-Degnen, & Hasselkus, 2002). Professionals must demonstrate a cause-and-effect relation- ship between their techniques and improvements in their clientele (Goldstein, 2000). Ottenbacher et al. (2002) advo- cated for occupational therapists to pursue EBP more vigor- ously in an attempt to empirically verify the effectiveness of procedures in this field. They called occupational therapists to “develop and use the skills necessary to plan treatment strategies for individual clients based on what is supported by evidence” (p. 247).

The purpose of the current study was to experimentally evaluate the effect of various sensory techniques on the ste- reotypy of three children diagnosed with autism. Several specific questions were to be addressed. Will students engage in less stereotypy after being exposed to slow ves- tibular movements via the swing? Will students engage in less stereotypy after being provided deep pressure? Will the use of a sensory diet decrease rates of stereotypic behavior?

Method

Participants

Three children (two males and one female) attending a pre- school program for children with developmental disabilities served as participants. All students participated in speech and language testing using the Preschool Language

Scale–Fourth Edition (Zimmerman, Steiner, & Pond, 2002) and cognitive testing using the Battelle Developmental Inventory–Second Edition (Newborg, 2005). Tristan was 3 years, 5 months of age with receptive and expressive lan- guage standard scores (SS) of 59 and 58, respectively, and received an SS of 60 on measures of cognitive ability. He was non-verbal, used a Picture Exchange Communication System (PECS) to communicate, and engaged in stereo- typic behaviors and active non-compliance to the extent that performance of academic and social tasks were interrupted.

CJ was also non-verbal and was learning to utilize PECS for communication. He was 2 years, 10 months of age and earned SS of 50 and 58 for receptive and expressive lan- guage skills, respectively, and functioned in the Mild Developmental Delay range (SS of 71) of cognitive ability. CJ engaged in self-injury, active non-compliance, and ste- reotypic behaviors that interfered with attention to task and disrupted group activities.

Rosie was 2 years, 9 months of age at the onset of the study. Testing revealed that her cognitive ability fell within the Mild Developmental Delay range (SS of 73) with recep- tive and expressive language scores being 3 standard devia- tions below the mean (SS of 50 on both measures). She engaged in stereotypy and tantrums that disrupted perfor- mance on tasks and interrupted social exchanges, had no functional language, and was working on verbal imitation skills at the onset of the project. All children were diag- nosed with autistic disorder by a developmental pediatri- cian prior to participating in the study.

Program Description

All participants attended a preschool program that utilized a treatment philosophy based on applied behavior analysis. The program operated 5 days per week for 5½ hr each day. The classrooms were organized by level of student func- tioning rather than age range. Children placed in this pro- gram received from 0 to 2½ hr of discrete trial training per day depending on their level of cognitive functioning, speech and language skills, socialization skills, and behav- ioral excesses. They also received speech therapy, occupa- tional therapy, and physical therapy as needed based on their initial evaluation.

Target Behavior and Response Measurement

Stereotypy was chosen as the target behavior for measure- ment for all three participants. Staff believed that this response class negatively affected classroom performance and learning, and was an important intervention target. Stereotypy was defined as a rhythmic, repetitive, exagger- ated or unusual tightening of the muscles of either a particu- lar body part or entire body, in the absence of an appropriate

at UNIV OF NORTHERN COLORADO on March 19, 2015foa.sagepub.comDownloaded from

16 Focus on Autism and Other Developmental Disabilities 30(1)

environmental context. Functional assessments of behavior were conducted for all three participants. In each case sen- sory/automatic reinforcement was identified as the primary function. The data collection system chosen for the study was 10-s partial interval recording.

Research Design

A within-subjects reversal design (Cooper et al., 2007) was chosen. First steady state responding during an initial base- line period was established, and then SIT (in whatever form it was prescribed by the occupational therapist for each child) was implemented. A return to baseline phase would have been conducted if, during the treatment phase, there was a visually clear pattern of the target behaviors changing in a positive way from the trend in the initial baseline phase. If, during the second baseline phase, it was determined that the behavioral trend matched the trend in the initial baseline phase, the treatment would have been implemented again.

Settings and Procedures

The investigators met with the occupational and physical therapy (OT/PT) staff to discuss the target behaviors and possible interventions. In all cases, the OT/PT staff hypoth- esized that the participants were over-stimulated and engaged in the behavior as a means to “calm their sensory systems.” Three different treatment plans were proposed by the OT/PT department to reduce stereotypy. All plans were chosen based on their hypothesized calming effects.

Tristan. Snack time was chosen as the target setting for Tristan. A net swing was used to treat his stereotypy. This technique was chosen based on the observation by the OT/ PT staff that the net swing seemed to produce a calming effect for Tristan. During baseline, on a daily basis, a staff member brought Tristan to the OT/PT room prior to partici- pating in snack time; however, he was not placed on the net swing. The staff simply sat on the floor with him and sang to him.

During treatment, a staff member brought him to the OT/ PT room at which time he was placed in the net swing and gently pushed back and forth by the staff. The staff placed him in the net swing on a pillow and slowly and rhythmi- cally swung him from side to side for 3 to 5 min while sing- ing to him. He was then removed from the swing and brought back to the classroom to participate in snack. This procedure was conducted once per day.

CJ. Circle time was chosen as the target setting for CJ. He was provided with “deep pressure” in an effort to reduce ste- reotypy during the activity immediately following the sen- sory technique. The baseline condition involved observing CJ during circle using a 10-s partial interval recording sys- tem. For the intervention, prior to the daily onset of circle

time, a staff member provided CJ with “deep pressure” by prompting CJ to sit down in a chair, sitting behind CJ, placing her hands on CJ’s shoulders (left hand on his left shoulder and right hand on his right shoulder), pushing down for 2 s and releasing for 2 s. The staff member repeated this 5 times before moving to the next body part (e.g., bicep). The staff member applied pressure to each bicep for 2 s and released for 2 s, moving down the arm 2 to 3 in. at a time. The staff member repeated the same procedure starting at CJ’s thighs and moving down to his ankles (left hand on left thigh and right hand on right thigh). Finally, the staff member provided “deep pressure” to CJ’s trunk (left hand on left side of trunk and right hand on right side of trunk) and moved down to his waist. This procedure took approximately 2 min to complete and was conducted once per day.

Rosie. Circle time was the target setting chosen for mea- surement; however, Rosie was exposed to the sensory diet throughout the entire day. To establish baseline levels of the target behavior, Rosie was observed during circle time using a 10-s partial interval recording system. Four activi- ties comprised the sensory diet and staff exposed her to one of the four activities on a half-hour rotating schedule. The four activities were “deep pressure” via heavy work activi- ties, “deep compression” via the therapy ball, joint com- pressions, and the “meatball squeeze.” It was hypothesized that this type of schedule would reduce stereotypy during the entire day.

“Deep pressure” via heavy work activities involved the therapist placing three gallon-sized laundry detergent bot- tles filled with water on small utility cart/stroller. The thera- pist prompted Rosie to push the cart with both hands up and back the entire length of the hall in the school. This proce- dure took approximately 3 to 5 min to complete.

Staff provided Rosie with “deep compression” via the therapy ball by prompting Rosie to lie down on her stom- ach, placing the therapy ball on Rosie’s back and applying gentle pressure down onto the ball. The therapist rolled the ball from Rosie’s shoulders, down her back, over her but- tocks, and down to her calves. They then reversed the motion of the ball to return back to her shoulders. This pro- cedure was repeated 15 times and took approximately 2 min to complete.

The therapist provided Rosie with joint compressions on the following joints: elbow, wrist, fingers, hip, knee, and ankle (both left and right sides of the body). The therapist placed one hand just below the joint and one hand above the joint on Rosie, gently compressed the joint together, and then released compression. The compression was repeated at all joints 10 times at a rate of 1 compression per second. Joint compressions could be applied to all the above men- tioned joints in approximately 2 min.

To implement the “meatball squeeze,” the therapist sat Rosie in front of her (therapist and Rosie facing the same direction), bent Rosie’s knees up and crossed her arms in

at UNIV OF NORTHERN COLORADO on March 19, 2015foa.sagepub.comDownloaded from

Sniezyk and Zane 17

front of her chest and knees (fetal position). The therapist wrapped her arms around Rosie with the therapist’s hands at Rosie’s shins while she was in this position and gently squeezed Rosie for 10 s and then released. The therapist repeated the squeeze 10 times. The “meatball squeeze” took approximately 3 min to complete.

Inter-Observer Agreement (IOA) and Procedural Fidelity

We obtained dependent measure and procedural reliability checks on a minimum of 25% of all sessions across the par- ticipants. A second independent observer viewed the ses- sion and collected data on the occurrence of the target behaviors. IOA was calculated by dividing the number of 10-s intervals with agreements by the number of 10-s inter- vals with agreements plus disagreements and multiplying by 100. The mean IOA was 97.3% (range, 96.5%–97.8%).

To determine procedural fidelity, the OT who designed the intervention trained all staff to implement the proce- dures and signed off on their competence. She then viewed a minimum of 25% sessions for Tristan and CJ via video- tape and observed 98.8% (range, 97.5%–100%) accuracy in implementing the procedures, using the attached procedural fidelity checklist (see appendix). Due to the high frequency with which techniques were implemented with Rosie (rotat- ing through 11 opportunities throughout each day for sev- eral months), it was not feasible for the OT to observe 25% of sessions. However, she viewed each intervention twice. The fidelity of implementing the sensory procedures ranged from 83% to 100% (M = 92.5%).

Results

Tristan engaged in the target behavior during an average of 29% (range, 26%–31%) of intervals during baseline (see Figure 1). The average number of intervals during which he engaged in the target behavior during intervention was 40% (range, 31%–62%). A return to baseline condition was not conducted because the levels of the behavior increased rather than decreased.

Data for CJ are displayed in Figure 2. CJ engaged in the target behavior during an average of 36% of intervals dur- ing baseline (range, 22%–58%). The average number of intervals during which he engaged in the target behavior during intervention reduced to 28% (range, 24%–25%). We then returned to baseline conditions and the occurrence of the target behavior continued to trend downward (M = 9%; range, 4%–10%).

Figure 3 depicts the percentage of intervals in which Rosie engaged in the target behavior over four phases. Phase 1 was the baseline condition during which Rosie engaged in the target behavior on an average of 23.9% of intervals (range, 8%–54%). Phase 2 was the intervention condition consisting of 3 data points, these being the last 3 days of school prior to summer recess. She engaged in the target behavior during an average of 15.3% of intervals (range, 9%–23%). After the summer recess, Rosie began program in a new classroom with new staff members. All staff members were trained as described previously. Phase 3 depicts data collected during circle time with the interven- tion in place in the new classroom. Rosie engaged in stereo- typy between 8% and 14% of intervals (M = 13.3%) during this condition. We returned to baseline in Phase 4. There

0

20

40

60

80

100

1 5 10

% o

f i nt

er va

ls

Sessions

Baseline Swing Intervention

Figure 1. Percentage of intervals of stereotypy for Tristan.

at UNIV OF NORTHERN COLORADO on March 19, 2015foa.sagepub.comDownloaded from

18 Focus on Autism and Other Developmental Disabilities 30(1)

was a further reduction in the percentage of intervals with which Rosie engaged in the target behavior (M = 3.7%, range 0%–8%).

Discussion

The purpose of this investigation was to use well-designed experimental procedures to investigate several questions pertaining to the effectiveness of different sensory proce- dures on reducing various self-stimulatory behaviors of children diagnosed with autism. We wanted to explore whether students engaged in less stereotypy after being exposed to slow vestibular movements via the swing. In addition, would students engage in less stereotypy after being provided with “deep pressure,” and would the use of a “sensory diet” decrease rates of stereotypic behavior? Based on recommendations by the participants’ occupa- tional therapists, sensory procedures were implemented according to the protocols established by the OTs. Dependent measure and procedural fidelity were assessed and found to be at acceptable levels. The results showed that with no participant was there a clear functional rela- tionship between positive changes in dependent measures and the implementation of sensory procedures. In the cases where there was reduced behavior during the SIT phase, the behavior continued to decrease when SIT was removed, suggesting that the initial reduction was not causally related to the SIT procedures.

These results correspond with much of the literature about the effectiveness of sensory integration procedures on problem behaviors of persons with autism spectrum

disorders (e.g., Parham et al., 2007; Smith et al., 2005). Reviewers of the literature report inconsistent results at best. Often, when results seem promising, there are obvious research design flaws, such as no reliability of measurement or uncontrolled threats to internal validity. However, Van Rie and Heflin (2009) conducted a study that met rigorous research standards, and found some positive results, as did Schaaf et al. (2013). More research, with the same level of care when designing methodological controls approximat- ing Van Rie and Heflin, is sorely needed. Strengths of the current study include the use of a tightly designed investiga- tion of sensory integration procedures and utilization of common research methodologies to enhance both the inter- nal and external validity of the outcomes.

With Tristan, the swing had no decelerative effect; indeed, the target behavior actually increased slightly in rate during intervention. Prior to the study, Tristan’s OT noted that based on observation, Tristan seemed to be calm during swinging. Post results, the OT could not explain why the intervention had an excitatory effect; she noted that lin- ear movement is typically calming and organizing; a rotary movement would be considered to produce an excitatory reaction, and could not speculate why swinging had such an opposite effect on Tristan. In this case, there are at least two possible explanations. First, perhaps the OT’s initial obser- vations of Tristan—that he seemed to calm during swing- ing—were simply in error or biased. Without objective assessment based on data, the true nature of what transpires may be lost. Second, it is highly likely that there are signifi- cant differences across individuals with regard to respon- siveness to sensory stimulation. That is, individuals are

0

20

40

60

80

100

1 5 10 15

% o

f i nt

er va

ls

Sessions

Baseline Deep Pressure Baseline

Figure 2. Percentage of intervals of stereotypy for CJ.

at UNIV OF NORTHERN COLORADO on March 19, 2015foa.sagepub.comDownloaded from

Sniezyk and Zane 19

likely to respond to sensory stimuli as differently as they do to other environmental stimuli. A sensory theory predicting similar outcomes across different individuals seems unlikely.

The data collected for CJ do not reveal a functional rela- tionship between “deep pressure” and stereotypy. While the target behavior did decrease in occurrence, it is important to note that the intervention data were within the same range as baseline; however, levels were more stable. The contin- ued decrease in the occurrence of the target behavior during the return to baseline condition does not allow a confident assumption of a functional relationship between the inter- vention and the target behavior. Some potential explana- tions for this post-intervention reduction include maturation or a cumulative effect of the intervention. A cumulative effect is unlikely based on the large lapse in time in between intervention and return to baseline. Any residual effects of the treatment would most likely have worn off during this time. It should be noted that the intervention phase was ter- minated on the completion of the extended school year pro- gram in August. CJ changed classrooms when school commenced in September. Although the change in class- rooms may be a confounding variable, it was considered to be a natural opportunity to terminate the intervention phase. Moreover, it is unlikely that the change in environment pro- duced a decrease in the target behavior because CJ’s alleged internal sensory issues would still be present in the new classroom. CJ did experience some difficulty with the tran- sition to the new classroom. He frequently engaged in tan- trum behavior that resulted in staff prompting. Data were

not collected on days that CJ engaged in tantrum behavior because his tantrum behavior was incompatible and inter- fered with the observation of the target behavior.

Rosie is another child for whom a functional relationship between improved behavior and the implementation of sen- sory procedures cannot be established. Possible explana- tions for the observed decreased occurrence of the target included are the same as for CJ. In terms of maturation, it should be noted that Rosie began imitating words and inde- pendently initiating verbalizations after the break in November. In addition, one would not expect the level of behavior to continue to decrease after the intervention was removed. The argument is the same for whether or not a cumulative effect was observed. One would not expect con- tinued improvements over intervention levels. It would seem that this would be an adequate amount of time for any effects of the treatment to “wear off” and therefore, one would observe an increase in the occurrence of the target behavior. In addition, there could be some unidentified confounding variable, since Rosie developed a new behavior involving manipulation of the seams in her clothes. This behavior was not noted during baseline and the beginning stages of inter- vention but became more prevalent as time went on. This behavior is topographically incompatible with the target behavior and may have interfered with its occurrence.

Several potential limitations should be addressed. There was no placebo control condition, which could have served as a test of the potential reactivity to having more attention by an adult. We chose not to perform such a placebo condi- tion because we were primarily interested in noting whether

0

20

40

60

80

100

% o

f i nt

er va

ls

Sessions

Baseline Sensory Diet -New Classroom BaselineSensory Diet

1 5 10 15 20 25 30 35 40

Figure 3. Percentage of intervals of stereotypy for Rosie.

at UNIV OF NORTHERN COLORADO on March 19, 2015foa.sagepub.comDownloaded from

20 Focus on Autism and Other Developmental Disabilities 30(1)

or not there was any positive effect of sensory integration procedures. A second potential limitation is that all aspects of a reversal design were not implemented with Tristan. With him, because the behavior did not change in a positive way from the baseline to intervention phase, it was assumed that going back to baseline would have again shown no sig- nificant change and thus the intervention ended. If a return to baseline had been implemented, there would have been one of three outcomes—a decrease in rate, an increase in rate, or no change in trend. None of these possibilities would have changed the conclusion that the sensory proce- dures had no positive impact. However, failure to return to baseline is a potential limitation to the study’s design and thus could affect the interpretation of the results.

Another potential limitation was the lack of blinded data collection. We established the validity of the data through frequent and carefully arranged reliability observations by observers naïve to the purpose of the study. An additional design issue concerns the break of the sessions between the end of the school year and the beginning of the summer ses- sion for Rosie. There was approximately a 2-week vacation between the end of phase 2 and the beginning of phase 3. It is doubtful that this recess influenced the effect of the sen- sory procedures, since the behavior seemed to be minimally affected prior to the vacation period. A final limitation con- cerns the fact that different sensory procedures were used across the three participants. The variability and implemen- tation duration of these procedures could have had a con- founding effect on the results. Although the sensory procedures differed across the three participants, we believe that this is not a weakness in our design. The procedures varied across the participants because the occupational ther- apist planned the procedures based on the unique needs and characteristics of each child. We believe this treatment validity overrides any methodological or experimental con- cerns. It was important that each child received the therapy best suited for that child, in the professional opinion of the Occupational Therapist. To artificially implement treat- ments for a child that were not identified to be best suited for his or her unique needs, for the sake of experimental rigor, may have detracted from the clinical significance of the study. In addition, we used within-subject designs, using each participant as his or her own control. Regardless of the independent variable, the behavior measured during the independent variable phase was compared with the same behavior of the same participant during the baseline phases.

Thus, we did not need to worry about similarity of indepen- dent variable across the three participants. We believe that for sensory integration, we need to first determine if the procedures—whatever they are—have any positive effect. In our study, we provided three tests of three different procedures.

This study contributes to the investigation of sensory integration in several ways. First, the OTs—the profession- als most often using sensory procedures—were given free rein to determine what they thought were the behaviors to be influenced by sensory procedures, and exactly what those procedures would entail, including the exact steps of the procedure, the duration of each application, the fre- quency of applications daily, and how long-term the proce- dures were used. Second, we implemented well-crafted experimental designs that would allow an investigation of a causal relationship, should the behavior seem to be sensi- tive to the independent variables. Third, we rigorously applied IOA procedures for all dependent variables, and similarly checked procedural fidelity.

The clinical implication of this study is that service providers must use sensory procedures with caution. The current study adds more data from three participants to the emerging conclusion that we cannot assume sensory pro- cedures are evidenced-based. That is, the research that exists on testing the effectiveness of SIT is at best incon- clusive, at worst showing thus far that it is not an effective treatment strategy. There does exist proven treatments that can affect behaviors typically addressed by sensory proce- dures, such as noncontingent reinforcement (e.g., Cox, Gast, Luscre, & Ayres, 2009) and response interruption and redirection (e.g., Ahrens, Lerman, Kodak, Worsdell, & Keegan, 2011).

Future researchers should focus on developing clear pro- tocols for testing sensory procedures. The sensory integration professionals should drive these protocols; but the research- ers need to develop research design methodologies to control for threats to internal and external validity, and both depen- dent measure and procedural reliability. The structure of the experimental design is important to enhance, since if positive results were ever discovered, there would be some confi- dence that positive changes were due to the independent variable(s) used. Only through a tightening of the experimen- tal protocols to test sensory integration procedures could we build a research foundation that provided any empirical sup- port for this unsubstantiated treatment.

at UNIV OF NORTHERN COLORADO on March 19, 2015foa.sagepub.comDownloaded from

Sniezyk and Zane 21

Appendix

Treatment Fidelity Checklist

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.

“Deep pressure” via heavy work activities Date Date Date

Holds cart with both hands

Pushes cart to end of hall

Pushes cart back to class

Note. X = step implemented correctly; O = step implemented incorrectly.

“Meatball squeeze” Date Date Date

Staff and Rosie sit on floor facing the same direction

Rosie sits on staff’s lap Bends Rosie’s knees to her chest Wraps her arms around her

knees

Staff wraps arms around Rosie Staff squeezes for 10 s and

releases

Repeats 10 s squeeze a total of 10 times

Note. X = step implemented correctly; O = step implemented incorrectly.

Joint compressions Date Date Date

Staff grasps one hand above Rosie’s joint and one hand below the same joint

Squeezes joint and release Ten squeezes per joint Left elbow Left wrist Left five fingers Right elbow Right wrist Right five fingers Left hip Left knee Left ankle Right hip Right knee Right ankle

Note. X = step implemented correctly; O = step implemented incorrectly.

Swing Date Date Date

Student sits on pillow in net swing Staff moves slowly move swing from

side to side

Student is in moving swing for 3 to 5 min

Staff sings to student

Note. X = step implemented correctly; O = step implemented incorrectly.

Joint compressions Date Date Date

Student sits in a chair Staff sits in chair either in front or

behind student

Staff places left and right hand on student’s left and right shoulder

Staff applies pressure by squeezing for 2 s and releasing

Staff moves hands down student’s arms to the wrist, applying 2 s of pressure every 2 to 3 in.

Staff place left and right hand on student’s left and right thighs

Staff applies pressure by squeezing for 2 s and releasing

Staff moves hands down student’s legs to the ankle, applying 2 s of pressure every 2 to 3 in.

Staff places left and right hand on student’s trunk under his or her left and right arm

Staff applies pressure by squeezing for 2 s and releasing

Staff moves hands down student’s trunk to his or her waist applying 2 s of pressure every 2 to 3 in.

Note. X = step implemented correctly; O = step implemented incorrectly.

CJ “Deep Pressure.”

Rosie Sensory Diet.

“Deep compression” via therapy ball Date Date Date

Rosie on stomach Staff places ball on her back Rolls to her ankles and back Repeats 15 times

Note. X = step implemented correctly; O = step implemented incorrectly.

Tristan Swing.

at UNIV OF NORTHERN COLORADO on March 19, 2015foa.sagepub.comDownloaded from

22 Focus on Autism and Other Developmental Disabilities 30(1)

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Ahrens, E. N., Lerman, D. C., Kodak, T., Worsdell, A. S., & Keegan, C. (2011). Further evaluation of response interrup- tion and redirection as treatment for stereotypy. Journal of Applied Behavior Analysis, 44, 95–108.

American Psychiatric Association. (2013). Diagnostic and statis- tical manual of mental disorders (5th ed.). Washington, DC: Author.

Arendt, R., MacLean, W. E. Jr., & Baumeister, A. (1988). Critique of sensory integration therapy and its application in mental retardation. American Journal of Mental Deficiency, 92, 401–411.

Ayres, A. J. (1979). Sensory integration and the child. Los Angeles, CA: Western Psychological Services.

Baranek, G. T. (2002). Efficacy of sensory and motor inter- ventions for children with autism. Journal of Autism and Developmental Disorders, 32, 397–422.

Cooper, J. O., Heron, T. E., & Heward, W. L. (2007). Applied behavior analysis (2nd ed.). Englewood Cliffs, NJ: Prentice- Hall.

Cox, A. L., Gast, D. L., Luscre, D., & Ayres, K. M. (2009). The effects of weighted vests on appropriate in-seat behaviors of elementary-aged students with autism and severe to profound intellectual disabilities. Focus on Autism and Developmental Disabilities, 24, 17–26.

Daems, J. (Ed.). (1994). Reviews of research in sensory integra- tion. Torrance, CA: Sensory Integration International.

Dubouloz, C. J., Egan, M., Vallerand, J., & Von Zweck, C. (1999). Occupational therapists’ perceptions of evidence-based practice. American Journal of Occupational Therapy, 53, 445–453.

Fertel-Daly, D., Bedell, G., & Hinojosa, J. (2001). Effects of a weighted vest on attention to task and self-stimulatory behav- iors in preschoolers with pervasive developmental disorders. American Journal of Occupational Therapy, 55, 629–640.

Goldstein, H. (2000). Commentary: Interventions to facili- tate auditory, visual, and motor integration: “Show me the data.” Journal of Autism and Developmental Disorders, 30, 423–425.

Grandin, T. (1992). Calming effects of deep pressure in patients with autistic disorder, college students and animals. Journal of Child & Adolescent Pharmacology, 2, 63–72.

Hardy, P. M. (1990, July). Anxiety and related disorders. Paper presented at the Annual Conference of the Autism Society of America, Los Angeles, CA.

Hoehn, T. P., & Baumeister, A. A. (1994). A critique of the appli- cation of sensory integration therapy to children with learning disabilities. Journal of Learning Disabilities, 27, 338–350.

Lalli, J. S., Casey, S. D., & Kates, K. (1997). Noncontingent rein- forcement as treatment for problem behavior: Some proce- dural variations. Journal of Applied Behavior Analysis, 30, 127–137.

McKerchar, T. L., Kahng, S., Casioppo, E., & Wilson, D. (2001). Functional analysis of self-injury maintained by automatic reinforcement: Exposing masked social functions. Behavioral Interventions, 16, 59–63.

Miller, L. J. (2003). Empirical evidence related to therapies for sensory processing impairments. Communique, 31, 34–37.

Newborg, J. (2005). Battelle Developmental Inventory—Second edition. Itasca, IL: Riverside.

O’Brien, G., & Pearson, J. (2004). Autism and learning disability. Autism, 8, 125–140.

Olson, L., & Moulton, H. J. (2004). Occupational therapists’ reported experiences using weighted vets with children with specific developmental disorders. Occupational Therapy International, 11, 52–66.

Ottenbacher, K. J., Tickle-Degnen, L., & Hasselkus, B. R. (2002). From the desk of the editor—Therapists awake! The chal- lenge of evidence-based occupational therapy. American Journal of Occupational Therapy, 56, 247–249.

Parham, L. D., Cohn, E. S., Spitzer, S., Koomar, J. A., Miller, L. J., Burke, J. P., . . .Summers, C. A. (2007). Fidelity in sensory integration intervention research. The American Journal of Occupational Therapy, 61, 216–226.

Parham, L. D., & Mailloux, Z. (1996). Sensory integration. In J. Case-Smith, A. Allen, & P. Pratt (Eds.), Occupational therapy for children (3rd ed., pp. 307–356). New York, NY: Mosby.

Schaaf, R. C., Benevides, T., Mailloux, Z., Faller, P., Hunt, J., van Hooydonk, E., . . .Kelly, D. (2013). An intervention for sen- sory difficulties in children with autism: A randomized trial (Department of Occupational Therapy Faculty Papers, Paper 15). Retrieved from http://jdc.jefferson.edu/otfp/15

Shaw, S. R. (2002). A school psychologist investigates sensory integration therapies: Promise, possibility, and the art of pla- cebo. Communique, 31, 5–6.

Smith, T., Mruzek, D. W., & Mozingo, D. (2005). Sensory inte- grative therapy. In J. Jacobson, R. Foxx, & J. Mulik (Eds.), Controversial therapies for developmental disabilities: Fad, fashion, and science in professional practice (pp. 331-350). Mahwah, NJ: Pergamon Press.

Stagnitti, K., Raison, P., & Ryan, P. (1999). Sensory defensiveness syndrome: A pediatric perspective and case study. Australian Occupational Therapy Journal, 46, 175–187.

VandenBerg, N. L. (2001). The use of a weighted vest to increase on-task behavior in children with attention difficulties. American Journal of Occupational Therapy, 55, 621–628.

Van Rie, G. L., & Heflin, L. J. (2009). The effect of sensory activities on correct responding for children with autism spec- trum disorders. Research in Autism Spectrum Disorders, 3, 783–796.

Vollmer, T. R., Iwata, B. A., Zarcone, J. R., Smith, R. G., & Mazaleski, J. L. (1993). The role of attention in the treat- ment of attention-maintained self-injurious behavior: Noncontingent reinforcement and differential reinforcement of other behavior. Journal of Applied Behavior Analysis, 26, 9–21.

Zimmerman, I. L., Steiner, V. G., & Pond, R. E. (2002). Preschool Language Scales—Fourth edition. Bloomington, MN: Pearson.

at UNIV OF NORTHERN COLORADO on March 19, 2015foa.sagepub.comDownloaded from