Abstract
Antecedent physical exercise has emerged as a potentially promising treatment for reducing challenging behavior and increasing academic behavior in individuals with autism spectrum disorder (ASD). The purpose of this study was to evaluate the effects of physical exercise conducted prior to instructional sessions (antecedent physical exercise) on academic engagement and stereotypy during instructional sessions for two children diagnosed with ASD. Functional analysis results suggested stereotypy was maintained by automatic reinforcement for both participants. A multielement design was employed to evaluate academic engagement and stereotypy during instructional sessions following randomly sequenced conditions involving either (a) no antecedent exercise, (b) brief durations of antecedent exercise, or (c) antecedent exercise that continued until the participant engaged in a systematically determined behavioral indicator of satiation. Both participants demonstrated higher levels of academic engagement and reduced levels of stereotypy during the instructional sessions which followed antecedent physical exercise that continued until behavioral indicators of satiation occurred. This study replicates previous research suggesting that individuals with ASD may benefit from physical exercise prior to academic instruction and further suggests that the duration of antecedent exercise may be optimally individualized based on behavioral indicators of satiation.
Restrictive and repetitive behavior and interests (RRBIs) have been considered a characteristic of autism spectrum disorders (ASDs) since the disorder was first defined (Kanner, 1943). RRBIs are now identified as one of the two core characteristics of an ASD diagnosis (Diagnostic and Statistical Manual of Mental Disorders [5th ed.; DSM-5]; American Psychiatric Association [APA], 2013). RRBIs include repetitive motor movements or vocalizations (stereotypy), restricted interests, or insistence on sameness regarding routines and rules (APA, 2013; Hattier, Matson, MacMillan & Williams, 2013). Stereotypy may interfere with educational activities and hinder a student’s academic engagement (Gorman-Smith & Matson, 1985; Koegel, Koegel, Frea, & Smith, 1995; Lovaas, 1981).
Academic engagement is an essential component of academic achievement (Nicholson, Kehle, Bray, & Van Heest, 2011). Students who engage in academic tasks have more opportunities to respond, practice skills, and interact with educational materials and this increase in opportunity may lead to increased rates of learning and educational advancement (DiPerna, Volpe, & Elliott, 2002). Previous research has suggested that interventions designed to reduce stereotypy may also occasion improvements in academic responding (Kern, Koegel, Dyer, Blew, & Fenton, 1982; Powers, Thibadeau, & Rose, 1992; Rosenthal-Malek & Mitchell, 1997).
Antecedent physical exercise has been identified as one approach to reducing motor stereotypy (e.g., Bahrami, Movahedi, Marandi, & Abedi, 2012; Celebreti, Bobo, Kelly, Harris, & Handleman, 1997; Kern et al., 1982; Morrison, Roscoe, & Atwell, 2011). A review by Lang et al. (2010) identified 18 studies that reported decreases in challenging behavior for individuals with ASDs following various types of exercise (e.g., jogging, swimming, and weight training). Of those 18 studies, 4 studies reported subsequent increases in academic behavior including on-task behavior (Powers et al., 1992), increased correct responding (Rosenthal-Malek & Mitchell, 1997), and increased task completion (Kern et al., 1982; Rosenthal-Malek & Mitchell, 1997). In addition to these previous findings, a study by Nicholson and colleagues (2011) found beneficial effects of exercise on academic engagement independent of stereotypy. Specifically, Nicholson and colleagues evaluated the effects of a jogging program on the academic engagement of four higher functioning individuals diagnosed with Asperger syndrome who did not engage in stereotypy. Study results reported increased academic engagement following physical exercise, suggesting that exercise may have direct effects on academic engagement. Therefore, previous research supports the potential benefits of antecedent physical exercise as an intervention for reducing challenging behavior and increasing appropriate academic behavior.
Although antecedent exercise has demonstrated positive effects on the behavior of individuals with ASD, the mechanism of action responsible for changes in challenging behavior is unclear (Kasner, Reid, & MacDonald, 2012; Lang et al., 2010; Sowa & Meulenbroek, 2012). For example, it is possible that decreases in challenging behavior following physical exercise may be due to physical fatigue. Previous research has investigated this theory by manipulating the intensity of antecedent exercise, with differential effects on challenging behavior for high-intensity exercise versus no effects for low-intensity exercise (e.g., Kern, Koegel, & Dunlap). Although fatigue has not been completely discounted, researchers have indicated that corresponding increases in appropriate behavior suggest other mechanisms may be contributing to the behavioral change (e.g., Lang et al., 2010; Morrison et al., 2011). To investigate the potential that physical fatigue may result in decreased challenging behavior, researchers have identified a need to experimentally manipulate the duration and intensity of the physical exercise (Lang et al., 2010; Morrison et al., 2011).
A second theory focuses on antecedent exercise as an intervention, which might affect an individual’s motivation to engage in specific behavior following the exercise. The concept that an antecedent event can alter an individual’s motivation to engage in a behavior is called a motivating operation (MO; Laraway, Snycerski, Michael, & Poling, 2003). There are two subcategories of MOs, abolishing operations (AOs) and establishing operations (EOs). An EO is an antecedent event that increases an individual’s motivation to engage in a specific behavior, and an AO is an antecedent event that decreases the individual’s motivation to engage in the specific behavior. The second theory posits that antecedent physical exercise may function as an AO for automatically maintained stereotypy by serving as matched stimulation for the behavior (Davis, Dacus, Strickland, Machalicek & Coviello, 2013; Lang et al., 2010; Morrison et al., 2011). Antecedent satiation on matched reinforcement for automatically maintained behavior has shown to be a promising intervention (e.g., Lang et al., 2009; Rispoli et al., 2013). Considering the theory that physical exercise may serve as matched stimulation for automatically maintained stereotypy, antecedent satiation on physical exercise may serve as an AO by decreasing the value of reinforcement obtained via stereotypy.
One method of recognizing a state of satiation in terms of a particular reinforcer is to identify a behavioral indicator of satiation. For example, a child who rejects (e.g., refuses to accept or pushes away) a previously identified highly preferred item after playing with the item for an extended amount of time may no longer find that item to be preferred and can be considered to be in a relative state of satiation in terms of that item’s reinforcing properties (O’Reilly et al., 2009). Previous research has utilized a behavioral indicator of satiation criterion of three rejections occurring during a time the child had free access to the item or activity (e.g. Lang et al., 2009; Rispoli, O’Reilly, Lang, et al., 2011; Rispoli, O’Reilly, Sigafoos, et al., 2011).
O’Reilly et al. (2009) provided participants presession access to the tangible maintaining their challenging behavior, in three different presession access conditions, to examine how variations in the duration of presession access influenced target behaviors during subsequent intervention sessions. These variations of presession access conditions included (a) no presession access condition, (b) brief presession access condition, and (c) a condition in which presession access continued until the individual produced the behavioral indicator of satiation. Results indicated high levels of challenging behavior following the brief duration antecedent exercise and no antecedent exercise conditions, with low levels of challenging behavior following the exercise until behavioral indicator of satiation condition.
This current study aims to expand the AO literature base by extending the use of behavioral indicators of satiation to antecedent physical exercise. To date, the application of behavioral indicators of satiation to antecedent physical exercise and subsequent effects on academic behavior and stereotypy have not been evaluated. The purpose of this study was to replicate previous research by evaluating the effects of varying conditions of antecedent physical exercise on stereotypy and academic engagement in instructional sessions for two children with ASD.
Method
Participants
Two children were recruited as participants from a university-supported autism clinic. These children were recruited because they had a diagnosis of ASD and engaged in motor stereotypy maintained by automatic reinforcement. In addition, the parents of both children reported that their child was able to engage in physical activity safely and provided informed consent for their child’s participation. The first two children considered were included and completed all study procedures.
Ally was an 8-year-old girl who was diagnosed by an outside agency with ASD. Ally scored 111 on the Autism Index of the Gilliam Autism Rating Scale–2 (GARS-2; Gilliam, 2006) and had a T-score of 79 on the Autism Spectrum Rating Scale (ASRS; Goldstein & Naglieri, 2009) further supporting her outside diagnosis of ASD. Ally’s stereotypy consisted of repetitive rocking (i.e., rhythmically bending her torso forward and backward or side to side) and repetitive bouncing (i.e., rapidly moving her torso up and down).
Chad was a 7-year-old boy who was diagnosed by an outside agency with ASD and severe intellectual disability (ID). Chad’s diagnostic information indicated that he scored a 97 on the GARS-2 (Gilliam, 2006) supporting his diagnosis of ASD. Chad’s stereotypy consisted of repetitive arm swinging (i.e., swinging his arms in back and forth) and repetitive head touching (i.e., raising his hands to head and touching or flicking his head with his fingers).
Setting and Materials
All instructional sessions for Ally and Chad were held in a 3 × 4 m classroom at the clinic. The room contained a small rectangular table and two small chairs. Only the researcher, participant, and a data collector(s) were present in the room during sessions. All sessions for both participants were videorecorded with an iPod® Touch. Sessions were 10 min in duration and were conducted once a day, 2 to 3 days per week over the course of 10 to 12 weeks.
Ally’s sessions were implemented by the first author and consisted of discrete trial teaching (DTT) of goals selected from Ally’s individual therapy plan. The academic skills targeted during instruction included identifying numerals up to 30, describing a picture using pronouns, counting up to 10 objects, rote counting up to 20, and following two-step gross motor imitations. Ally earned tokens for correct responses on a FR5 schedule. Tokens were traded for backup reinforcers at the end of each session.
Chad’s sessions were implemented by a doctoral student in Special Education. Sessions consisted of tasks selected from Chad’s individual therapy plan and consisted of following one-step directions (“put in bucket”), imitating one-step gross motor (“shake rattle” and “roll car”), matching puzzle pieces, and stacking blocks. Chad earned tokens for correct responses on a FR4 schedule and also exchanged tokens at the end of sessions. The DTT sessions were consistent with DTT intervention sessions that both participants were already receiving as part of their therapy at the autism clinic. The tasks, schedule of reinforcement, and therapist were held constant for both participants across all conditions.
Physical exercise for both participants included jumping on an indoor trampoline. The trampoline was located in a therapy room that was approximately 3 × 8 m and contained other supplies and materials unrelated to the study but common in clinical settings (e.g., shelves of instructional materials). The trampoline was 2 m in diameter and contained a safety net with a zipper designed to prevent the child from falling.
Dependent Variables and Data Collection
Data were collected on participants’ stereotypy and academic engagement. Data were collected using a 10 s partial interval recording procedure for the participants’ stereotypy and a 10 s whole interval recording procedure for academic engagement. Stereotypy was operationally defined individually for each participant using the descriptions already provided. Ally’s academic engagement was operationally defined as one or a combination of (a) looking at the task or instructor, (b) verbally responding to questions, or (c) appropriately responding to directives (i.e., imitating gross motor or counting items) without a full physical prompt. Chad’s academic engagement was defined as one of a combination of (a) looking at the task or instructor or (b) appropriately responding to directives (i.e., roll car) without a full physical prompt.
Functional Analysis
An analog functional analysis of each participant’s stereotypy was conducted. The functional analysis utilized a multielement design with five conditions: attention, escape, tangible, play (control), and an alone condition. Sessions were 5 min, and five sessions of each condition were conducted. The attention, escape, play, and alone conditions were designed following the procedures outlined by Iwata, Dorsey, Slifer, Bauman, and Richman (1982/1994). The tangible condition was implemented using the procedures described by Vollmer, Marcus, Ringdahl, and Roane (1995).
Prior to conducting the tangible condition, participants’ teachers were interviewed to identify potentially preferred stimuli. The participants’ preferences for these stimuli were assessed using a paired stimulus preference assessment (Fisher et al., 1992). The stimulus identified as most preferred was used as the tangible in the tangible condition. Prior to a tangible session, the researcher allowed the participant 10 s of free access to his or her preferred item (e.g., Legos). Following the 10 s access, the researcher removed the preferred item from the participant and placed it out of reach but within sight of the participant. Any attempts by the participant to obtain the preferred item were blocked. Contingent on target challenging behavior, the 10 s of access to the preferred item was provided. All other behavior was ignored.
Interobserver Agreement (IOA)
A second independent observer collected reliability data for 50% of baseline sessions and 25% of intervention sessions for Ally, and 80% of baseline sessions and 29% of intervention sessions for Chad. Reliability data were collected either in vivo (with both data collectors inside the therapy room) or by watching videorecorded sessions. IOA was calculated using block-by-block agreement. An agreement was scored if both raters identified that the target behavior occurred within a specified interval. A disagreement was scored if the raters differed on the occurrence of the behavior within a specified interval. IOA was calculated by dividing the number of agreements by the number of agreements plus disagreements. To obtain a percentage, the result was then multiplied by 100. The mean IOA for Ally’s stereotypy was 90% (range = 85%-100%) and the mean IOA of Chad’s stereotypy was 95% (range = 85%-100%). The mean IOA for Ally’s academic engagement was 90% (range = 83%-100%), and for Chad’s academic engagement was 91% (range = 80%-100%).
Treatment Integrity
Treatment integrity data were collected for at least 55% of all sessions, for each participant, and during each phase of the study using a researcher-developed task analysis for each phase. For baseline and no antecedent exercise conditions, there were four steps: (a) Researcher provided continuous instruction using a least-to-most prompting hierarchy (verbal, verbal + model, then verbal + physical prompt), (b) researcher delivered praise on successful completion of a trial/task, (c) researcher did not deliver any interactions/praise outside of the task conditions, and (d) researcher ignored stereotypic behavior. In addition to these four steps, the brief duration antecedent exercise and exercise until behavioral indicator of satiation conditions also included the following: (a) Researcher did not provide any interactions/attention during physical exercise and (b) on either the third behavioral indicator of satiation in the exercise until behavioral indicator of satiation condition or the expiration of the time limit for the brief duration antecedent exercise condition, researcher prompted the student to enter the room with academic demand using a least-to-most prompt hierarchy (gestural, verbal, verbal + model, verbal + physical). The data collector evaluated adherence to the intervention using this task analysis as a checklist. Treatment integrity was then calculated by dividing the total number of steps completed correctly by the total number of expected steps and then multiplying by 100. The mean treatment integrity during Ally’s sessions was 99% (range = 86%-100%). The mean treatment integrity during Chad’s sessions was 95% (range = 75%-100%).
IOA data were also collected for treatment integrity data. A second observer collected data during at least 30% of study sessions (43% of all the treatment fidelity sessions), for each participant, during each phase of the study. An agreement was scored if the first and second observer agreed on the occurrence or nonoccurrence on a step-by-step basis for each step in the task analysis. A disagreement was scored for any discrepancy. IOA for the treatment integrity data was calculated by dividing the number agreements by the number of agreements plus disagreements, and then multiplying by 100. The IOA for the treatment integrity data was 100% for both participants across all phases of the study.
Research Design
Prior to instructional sessions one of three different presession conditions were implemented within the context of a multielement research design: (a) no antecedent exercise, (b) exercise until behavioral indicator of satiation occurred, or (c) brief duration antecedent exercise. Prior to the multielement phase, a baseline involving consecutive sessions without any antecedent exercise was in place. This research design was chosen to compare stereotypy and engagement across conditions, and to compare both conditions with baseline condition prior to the introduction of exercise. Postbaseline conditions were randomly alternated to minimize potential sequence effects and establish experimental control (Kennedy, 2005).
Instructional Session
Instructional sessions were held constant across conditions and participants. Each instructional session was 10 min in duration and one session was conducted per day with no more than three sessions occurring per week. DTT involved a least-to-most prompting hierarchy and verbal praise contingent on completion of tasks. The researcher ignored all inappropriate behavior stereotypy and access to stereotypy was not blocked. The time of day that the sessions occurred was held constant with the only difference between sessions being the presession access to exercise on the trampoline.
Baseline and No Antecedent Exercise Condition
During the course of the study, both participants’ access to a trampoline was limited to programmed access provided in the study. Ally’s mother reported that she did not have any access to the trampoline prior to arriving at the clinic and the first author coordinated with Chad’s teacher to ensure he did not have access to the trampoline on days he participated in this study. For the baseline and no antecedent exercise condition, participants were not allowed access to the trampoline prior to the instructional sessions.
Antecedent Exercise Until Behavioral Indicator of Satiation
For the antecedent exercise until satiation condition, participants jumped on the trampoline until their individual behavioral indicator of satiation had occurred three times (e.g., O’Reilly et al., 2009). The behavioral indicator of satiation was individually determined for each participant using a two-step procedure. Researchers first discussed with his or her therapist how the participant communicated that he or she was done engaging in an activity. Researchers asked the therapists questions such as “what do they typically do when they are done with an activity?” and “how do you know when they are not done?” From these interviews, the researchers then formed a hypothesis on the behavioral indicator of satiation and observed the participants engaging in exercise on the trampoline for two sessions across 2 days. During these sessions, the researcher told the participant that they could jump on the trampoline. If the participant engaged in their hypothesized behavioral indicator of satiation, the researcher verbally prompted the participant by saying “You’re almost done. You can keep jumping.” On the participant’s third hypothesized behavioral indicator of satiation, the researcher told the participant “You can keep jumping if you want.” If the participant kept jumping, the hypothesized behavioral indicator was determined to be invalid and was revised and retested. Once the participant consistently ceased jumping across two sessions, the hypothesized behavioral indicator of satiation was determined to be confirmed. For Ally, her behavioral indicator was confirmed within three sessions and defined as “ceasing jumping or bouncing on the trampoline for 1 min.” Chad’s behavioral indicator was confirmed within two sessions and defined as “physically getting off the trampoline (his feet hitting the floor outside the trampoline).”
During exercise until behavioral indicator of satiation, the researcher minimized attention and interactions to keep this potentially confounding variable consistent throughout study conditions. For Ally, the researcher stood outside of the trampoline room and observed through a one-way mirror, and for Chad, a researcher was on the trampoline with him but did not provide any attention. If the participant engaged in his or her behavioral indicator of satiation, the researcher verbally prompted the participant by saying “You’re almost done. You can keep jumping.” On the participant’s third behavioral indicator of satiation, the researcher immediately instructed the participant to the classroom using a least-to-most prompting hierarchy. The instructional session then began on entering the classroom.
Brief Antecedent Exercise
For the brief antecedent exercise condition, participants were allowed access to the trampoline for a predetermined amount of time. Access time was individually calculated as 20% of the time to satiation as measured during the first exercise until behavioral indicator of satiation session (2 min 36 s for Ally and 1 min 38 s for Chad). In determining this duration, it was hypothesized that the duration of satiation sessions would likely vary across participants. Therefore, a percentage of the time to satiation was selected to allow for procedural consistency across participants. Twenty percent of the time to satiation was selected to be sufficiently short to ensure participants would not satiate during the brief antecedent exercise conditions.
During the brief exercise on the trampoline, the researcher minimized attention and interactions to keep this potentially confounding variable consistent throughout study conditions. The researcher did not prompt the participant to engage in exercise. On the expiration of the time limit, the researcher prompted the participant to the classroom. The instructional session then began on entering the classroom.
Results
Functional Analysis
Figure 1 displays the percentage of 10 s intervals with stereotypy across each condition of the functional analysis. Both participants engaged in stereotypy during all of the conditions of the functional analysis, including the alone condition, suggesting stereotypy was maintained, at least in part, by automatic reinforcement.
Results from analog functional analysis.
Antecedent Exercise Analysis
Figure 2 displays the percentage of 10 s intervals with stereotypy across the exercise until behavioral indicator of satiation, brief duration of antecedent exercise, and no antecedent exercise conditions for Ally and Chad in the upper and lower panels, respectively. Ally’s mean stereotypy was 34% of intervals (range = 15%-50%) following the exercise until behavioral indicator of satiation condition, 59% of intervals following the brief duration of antecedent exercise (range = 47%-72%), and 66% of intervals (range = 58%-72%) following the no antecedent exercise conditions. Chad engaged in lower levels of stereotypy following the exercise until behavioral indicator of satiation condition (M = 12%; range 0%-22%) than the brief duration of antecedent exercise (M = 28%; range 22%-43%) and no antecedent exercise conditions (M = 28%; range = 22%-32%). Only the exercise until behavioral indicator of satiation condition produced responding lower than baseline levels for both Ally and Chad.
Percentage of intervals with stereotypy across antecedent exercise conditions.
Figure 3 displays the percentage of 10 s intervals with academic engagement following the exercise until behavioral indicator of satiation, brief duration of antecedent exercise, and no antecedent exercise conditions for Ally and Chad. Academic engagement was higher following the exercise until behavioral indicator of satiation condition than in the brief duration of antecedent exercise and no antecedent exercise conditions. In addition, academic engagement was slightly higher for both participants following the brief duration of antecedent exercise than the no antecedent exercise condition. Ally’s academic engagement was highest following the exercise until behavioral indicator of satiation condition (M = 89%; range = 82%-93%) and higher following the brief duration of antecedent exercise (M = 63%; range = 57%-73%) than the no antecedent exercise condition (M = 53%; range = 42%-60%). Ally engaged in equivalent levels of academic engagement in the no antecedent exercise condition as compared with the baseline condition, with baseline responding for Ally averaging 50% of the intervals (range 42%-57%).
Percentage of intervals with academic engagement across antecedent exercise conditions.
Chad’s academic engagement was highest following the exercise until behavioral indicator of satiation condition (M = 78%; range = 68%-93%) and higher following the brief duration of antecedent exercise condition (M = 52%; range = 28%-67%) than the no antecedent exercise condition (M = 31%; range = 20%-43%). Chad engaged in equivalent levels of academic engagement in the no antecedent exercise condition as compared with the baseline condition, with baseline responding for Chad averaging 34% (range = 27%-45%).
Time to Satiation
The mean time to satiation on physical exercise on the trampoline for Ally was 9 min 37 s (range 5 min 40 s-13 min 52 s). The mean time to satiation for Chad was 6 min 18 s (range 4 min 4 s-8 min 18 s).
Discussion
The purpose of this study was to replicate previous research by evaluating the effects of varying conditions of antecedent physical exercise on stereotypy and academic engagement in instructional sessions for two children with ASD. The greatest reduction in stereotypy was observed following the exercise until behavioral indicator of satiation condition for both participants. Academic engagement was highest in the exercise until behavioral indicator of satiation condition, followed by the brief duration of antecedent exercise, and lowest in the no antecedent exercise condition.
This study supports previous findings regarding presession exercise as a potential treatment for stereotypy (e.g., Kern et al., 1982; Morrison et al., 2011). To our knowledge, this is the first study to apply behavioral indicators of satiation in determining the amount of physical exercise to provide an individual. The use of behavioral indictor of satiation on exercise is supported by the results, given that the exercise until behavioral indicator of satiation condition produced better outcomes for both stereotypy and academic engagement as compared with brief duration of antecedent exercise condition for both participants. However, these results are preliminary and should be replicated with additional participants.
Two theories have been posited to explain the effects of antecedent exercise on subsequent behavior changes. The first suggested mechanism is fatigue, or the idea that antecedent exercise leads to subsequent decreased stereotypy due the participants being tired from the exercise. However, previous research has identified that increases in academic engagement and other appropriate behaviors following antecedent physical exercise discount the fatigue theory (Kern et al., 1982; Lang et al., 2010; Powers et al., 1992). The fact that antecedent physical exercise led to increased academic engagement for both of the participants in this study replicates previous findings and suggests that a mechanism beyond physical fatigue may be responsible for the effects of physical exercise on challenging behavior.
A second potential explanation is that antecedent exercise may function as an AO for the interfering stereotypy. It is plausible that physical exercise on the trampoline may have functioned as a sort of matched stimulation for automatically maintained motor stereotypy. If this were the case, then presession satiation on matched stimulation may have functioned as an AO by reducing the value of automatic reinforcement and the frequency of behavior previously correlated with automatic reinforcement (motor stereotypy). In addition to the separation of responding following the exercise until behavioral indicator of satiation condition (as compared with brief or no antecedent exercise conditions), the increases in academic engagement following the exercise until behavioral indicator of satiation condition suggest participants allocated their responding toward academic tasks and away from stereotypy. This may be indicative of the decreased value of automatic reinforcement following satiation on matched reinforcement. Future research could further investigate this theory by examining the effects of matched versus unmatched physical activity on various forms of automatically maintained stereotypy, such as verbal stereotypy.
This study extends the research on academic engagement and individuals with ASD. Academic outcomes and performance for individuals with ASD is an underdeveloped area of research (May, Rinehart, Wilding, & Cornish, 2013). This study adds to the current literature base by investigating an antecedent intervention to reduce challenging behavior and increase academic engagement. Engagement in challenging behavior is a potential barrier to academic achievement, and, in particular, stereotypic behaviors can interfere with academic opportunities (Hart & Whalon, 2013; Zandt, Prior, & Kyrios, 2007). This study demonstrates the potential utility of antecedent physical exercise as an intervention to decrease stereotypic behavior and increase academic engagement.
Some important limitations must be noted for this study. First, the instructional sessions following the physical exercise were only 10 min in duration and no behavioral data were taken after the session. Therefore, it is not known how long the effects were maintained and future research should consider investigating the duration of this behavioral change. The duration of the effects could have important implications for practice as a short duration of effect could limit the validity and feasibility of this intervention. Second, the method for determining the brief duration antecedent exercise condition was based on the first antecedent exercise until behavioral indicator of satiation session. While the duration was set to be relatively brief, it is possible that the duration could have approached satiation during some sessions. However, both participants consistently engaged in higher levels of academic engagement and lower levels of stereotypy during the antecedent exercise until behavioral indicators of satiation sessions as compared with brief duration antecedent exercise conditions, suggesting that participants did not reach satiation levels during the brief duration exercise conditions.
Finally, data were not taken regarding the participants’ accuracy of responding during the session, and, therefore, it is not known whether the increased academic engagement led to meaningful academic progress. Future research should attempt to replicate these findings with additional participants and populations. In addition, as academic performance is correlated with lifetime achievement, employment, and overall quality of life, future research should focus on developing the research base for individuals with ASD to ensure access to high quality academic instruction (Fleury et al., 2014; Migliore, Timmons, Butterworth, & Lugas, 2012).
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.