USF Libraries
USF Digital Collections

The effects of reinforcement magnitude and session length on stereotypical behavior of an adolescent with autism


Material Information

The effects of reinforcement magnitude and session length on stereotypical behavior of an adolescent with autism
Physical Description:
Kotsios, Spiro
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:


Subjects / Keywords:
Functional analysis
Reinforcer duration
Dissertations, Academic -- Child and Family Studies -- Masters -- USF   ( lcsh )
non-fiction   ( marcgt )


ABSTRACT: Functional Analysis is an important component of creating a treatment plan for individuals with autism. A concern of functional analysis is that it might not match the natural environment in which the problem behavior occurs. Dimensions such as varying reinforcement duration and session length may affect response rates during functional analyses in natural settings. This study investigated if varying reinforcement duration and session length during functional analyses would affect the identification of the function of problem behavior across assessors, and the intervention designed based on the functional analyses would be effective in reducing stereotypic behavior and teaching communicative skills of an adolescent with autism at home. An alternating treatment design was used to conduct functional analyses and an ABA design with generalization probes was used to test the effectiveness the function-based intervention. Results indicated that the participant's stereotypic behavior was maintained by automotive reinforcement. The intervention developed based on the function of the stereotypic behavior was found to be effective in reducing the stereotypy and teaching communicative behavior.
Thesis (M.A.)--University of South Florida, 2009.
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Spiro Kotsios.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 59 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002068303
oclc - 606611207
usfldc doi - E14-SFE0003129
usfldc handle - e14.3129
System ID:

This item is only available as the following downloads:

Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200385Ka 4500
controlfield tag 001 002068303
005 20100412152315.0
007 cr mnu|||uuuuu
008 100412s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003129
HV697 (Online)
1 100
Kotsios, Spiro.
4 245
The effects of reinforcement magnitude and session length on stereotypical behavior of an adolescent with autism
h [electronic resource] /
by Spiro Kotsios.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 59 pages.
Thesis (M.A.)--University of South Florida, 2009.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
3 520
ABSTRACT: Functional Analysis is an important component of creating a treatment plan for individuals with autism. A concern of functional analysis is that it might not match the natural environment in which the problem behavior occurs. Dimensions such as varying reinforcement duration and session length may affect response rates during functional analyses in natural settings. This study investigated if varying reinforcement duration and session length during functional analyses would affect the identification of the function of problem behavior across assessors, and the intervention designed based on the functional analyses would be effective in reducing stereotypic behavior and teaching communicative skills of an adolescent with autism at home. An alternating treatment design was used to conduct functional analyses and an ABA design with generalization probes was used to test the effectiveness the function-based intervention. Results indicated that the participant's stereotypic behavior was maintained by automotive reinforcement. The intervention developed based on the function of the stereotypic behavior was found to be effective in reducing the stereotypy and teaching communicative behavior.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Advisor: Kwang-Sun Cho Blair, Ph.D.
Functional analysis
Reinforcer duration
0 690
Dissertations, Academic
x Child and Family Studies
t USF Electronic Theses and Dissertations.


The Effects of Reinforcement Magnitude a nd Session Length on Ster eotypical Behavior of an Adolescent with Autism by Spiro Kotsios, B.A. A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts Department of Child and Family Studies College of Behavioral and Community Sciences University of South Florida Major Professor: Kwang-Sun Cho Blair, Ph.D. Bobbie Vaughn, Ph.D. Frans van Haaren, Ph.D. Date of Approval: July 16, 2009 Keywords: Functional Analysis, Autism, Stereotypy, Reinforcer Duration, Sessions Copyright 2009, Spiro Kotsios


i Table of Contents List of Figures iii Abstract iv Section 1: Introduction 1 Functional Analysis 2 Functional Analysis in the Natural Settings 6 Stereotypy of Individuals with Autism 10 Preference Assessments 12 Schedules of Reinforcement 14 Magnitude of Reinforcement 16 Functional Analysis Based Interv entions to Reduce Stereotypy 21 Purpose and Research Questions 23 Section 2: Method 24 Participants 24 Setting 25 Response Measurement 25 Procedural Integrity 27 Data Collection Procedures a nd Interobserver Agreements 28 Preference Assessments 29 Design and Procedures 30 Descriptive Functional Assessment 30


ii Functional Analysis 31 Baseline 32 Intervention Procedures 33 Modified Intervention 37 Social Validity 37 Section 3: Results 38 Functional Analysis 38 Intervention Outcomes 38 Social Validity 40 Modified Intervention 40 Section 4: Discussion 43 References 50 Appendices 57 Appendix A: Treatment Acceptability Rating Form-Revised (TARF-R) 58


iii List of Figures Figure 1: Percentage of Stereotypy acr oss Functional Analysis Conditions 41 Figure 2: Percentage of Stereotypy and Communicative Behavior across Experimental Conditions 42


iv The Effects of Reinforcement Magnitude a nd Session Length on Ster eotypical Behavior of an Adolescent with Autism Spiro Kotsios ABSTRACT Functional Analysis is an important com ponent of creating a treatment plan for individuals with autism. A con cern of functional analysis is that it might not match the natural environment in which the problem beha vior occurs. Dimensions such as varying reinforcement duration and session length may affect response rates during functional analyses in natural settings. Th is study investigated if vary ing reinforcement duration and session length during functional an alyses would affect the iden tification of th e function of problem behavior across assessors, and the in tervention designed based on the functional analyses would be effective in reducing st ereotypic behavior and teaching communicative skills of an adolescent with au tism at home. An alternating treatment design was used to conduct functional analyses and an ABA design with generalization probes was used to test the effectiveness the function-based intervention. Results indicated that the participant’s stereotypic behavior was maintained by automotive reinforcement. The intervention developed based on the function of the stereotypic behavior was found to be effective in reducing the stereotypy and teaching communicative behavior.


1 Chapter 1: Introduction Autism can affect a variety of aspects of a person’s quality of life, including interacting with other people, communicating ideas and feeling, and understanding what others feel or think (Nati onal Research Council, 2001). Autis m begins in early childhood and persists throughout the lifetime of the aff licted individual (Crocket t et al., 2007; Prior, 2003), Individuals with autism require a hi gher level of support than the general population. This stems from impairments in so cial, communicative, and repetitive and stereotyped behaviors (Rap in, 1997; Ingersoll & Gergans, 2007), which are the cornerstone of progressing thr oughout various environments a nd situations that occur in everyday life. Problem behaviors are among th e most challenging and stressful issues faced by families in their efforts to provide a supportive environment. Recent surveys suggests that nearly 50-70% of individuals with autism present co-occurring behavioral and emotional problems (Gadow, Devicent, Pomeroy, & Azizian, 2004; Tonge & Einfeld, 2003). These disorders require a lifelong commitme nt of services for the individuals with autism and their families, costing upwards of $4 million per individual (Jacobson, Mulick, & Green, 1998). Researchers and professionals have recognized a powerful approach, called functional assessment, in the intervention fo r problem behavior of individuals with autism. Prior to the use of functional a ssessment approach, problem behavior of individuals with autism was t ypically treated by arbitrary co ntingencies of reinforcement or punishment over unknown sources of rein forcement for problem behavior (Mace,


2 1994). Functional assessment has made subs tantial contribution to the field of developmental disabilities including autism. Since the study by Iwata, Dorsey, Slifer, Bauman, and Richman (1982), an extensive body of research using functional analysis has been published in the behavior analytic literature demonstrati ng that environmental stimuli and operant function of problem behavior can be identified for certain classes of behavior. Functional assessment procedures are classified into three categories: informant assessment, descriptive assessment, and functional analysis (Lennox & Miltenberger, 1989). Informant assessments rely solely on the verbal report of careg ivers and clients for information regarding potential determinan ts of problem behavior. The descriptive assessment employs direct observations, collecti ng data in a linear fashion as events in natural settings (Lalli & Goh, 1993) In functional analysis, specific environmental events are experimentally manipulated to identif y the functional relationship between the environment and problem behavior. By contra st, functional analysis presents the most rigorous of the functional assessment met hods (Matson & Minshawi, 2007). The purpose of identifying the functional relations hip between the prob lem behavior and environmental stimuli is to develop an interven tion that reduces or eliminates the problem behaviors by creating conditions which make th ese controlling variab les absent from the client’s environment (Cihak et al., 2007). Nu merous studies have demonstrated that the individualized function-based intervention is effective in reducing or eliminating the problem behavior of individuals wi th disabilities. (Blair et al., 2006) Functional Analysis Functional analysis procedures ar e implemented to develop a powerful intervention by creating conditions which make these controlling variables absent from


3 the individual’s environment. The functional analysis methods developed by Iwata et al. (1982) and Carr and Durand (1985) marked th e beginning of a comprehensive approach to intervention for individuals with devel opmental disabilities, which has led to the development of more precise reinforcement-ba sed interventions and a decrease in the use of punishment. The primary conditions used in functional analysis include attention, alone, demand, and control (Iwata et al., 1982) A multi-element design can be used to test each condition, alternating between each condition (Healey et al., 2001; Sidman, 1960). One method of alternating between these conditions is to use a condition that might be responsible for the problem behavi or, followed by the control condition, then followed by another condition possibly res ponsible for the problem behavior, and followed by the control condition again, with th is pattern being repe ated till all possible conditions are tested (e.g., example, a ttention condition, control, demand condition, control, alone condition, control, demand condi tion, and finally control). This is known as the sequential, test-control methodology (Iwat a et al., 1994). This is an experimental design that involves alternati ng between the test and contro l conditions. This procedure differs along several dimensions from othe r functional assessment procedures. For example, a descriptive assessment employs observation of client and caregiver interactions and calls for instruction to the ca regivers to behave as they normally would (Thomposn & Iwata, 2001). An observer record s each incident of problem behavior and the consequence delivered from the caregiver. The purpose of the descriptive assessment is to reveal the naturally occurring cons equences for problem behavior (Thompson & Iwata, 2007). Rationale for using descriptiv e assessment is that the natural occurring


4 consequences, from caregivers, will expose the function of the problem behavior. A disadvantage of this method is that it does not isolate each function while keeping others removed. The delivery of the consequence is determined by the caregiver, and this may not match the function of the problem behavior (Hall, 2005). In th e functional analysis procedure, only one consequence is delivered for each (e.g., during tangible condition, only a tangible is given contingent on the occurr ence of problem behavior). It is expected that problem behavior will only occur at high rates during the condition which corresponds to the function of the problem behavior, and will be at a low rate during conditions which do not correspond. Through the sequential method, this is further replicated across and the data can demons trate distinctions between each condition, therefore leading to the identification of the function of the problem behavior. This method is useful because it is important to identify the controlling variable responsible for the problem behavior (Ca rr & Owen-DeSchryver, 2007). During functional analyses, reinforcer s associated with each condition are presented contingent upon the target behavi or. For example in the attention condition with a child whose target beha vior is self-injurious behavi or, whenever the child hits himself, he is given attention for a brief pe riod. If the child exhi bits self-injurious behavior (SIB) at a significantly higher ra te than in the other conditions, then a hypothesis is formed that the function of the ta rget behavior is to ga in attention. In the control condition, the child is given access to reinforcers (e.g., attention and tangibles) and is free of any tasks (e.g., demand conditi on). Since the child has access to these reinforcers, the child is expected not to pe rform the target behavi or as a function of attention, escape, and access to tangibles. The control conditi on is compared to the other


5 conditions to determine the function of the problem behavior. Signi ficantly higher levels of the target behavior in one or more conditions indicate the controlling variables responsible for the target behavior. The importance of identifying these controlling variables is that they can be used in inte rventions focused on reducing problem behavior (Kahng, Abt, & Schonbachler, 2001). The effectiveness of functional analys is has been demonstrated through the replication of countless studies that have proven the appr oach to be valuable in identifying the variables responsible for problem be haviors. This is an efficient tool to be used before developing and implementing a detaile d treatment plan, as it can save a lot of “guesswork” as to why a child behaves the wa y he does (Alter et al ., 2008). Although the majority of functional analys is studies examine self-injur ious behavior the functional analysis methodology has also been widely used to identify functions of different problem behaviors, regardless of their topograp hy. It has been used to examine a large variety of problem behaviors su ch as aggression, yelling, dest ruction of objects, tantrums, pica, elopement, self-injurious behavior, a nd non-compliance (Hanle y et al., 2003; Moore, Fisher, & Pennington, 2004). Not only has func tional analysis been used to assess problem behaviors exhibited by i ndividuals with severe disab ilities, but also to assess topographies of problem behavi or exhibited by typically de veloping children (Hanley et al., 2003; Ward & Higbee, 2008). Despite the utility of the functional analysis procedures, literatu re indicates that it may be difficult to determine the function of problem behavior of individuals with developmental disabilities (Tiger, Hanley, & Bessete, 2006). In th e Piazza et al. study (1998), functional analyses dete rmined that the problem beha vior of a child with autism


6 was a function of escaping from tasks. Results also indicated that attention maintained destructive behavior, suggesti ng that escape from tasks al one was not enough to decrease the problem behavior. Fisher et al. (1996) alternated betw een demand, attention, tangible, and control conditions to demonstrate the function of self-injurious behavior and aggressive behavior in a child with ment al retardation. The duration of access to reinforcers was varied for each condition. Lowe r rates of SIB and aggressive behavior were associated with greater duration of acce ss to reinforcers. Kennedy et al. (2000) used a functional analyses including attention, demand, no attentio n, and recreation conditions for children with autism. There were mixed results, with two of the five children engaging in stereotypical be havior during all experiment al conditions, including the recreation condition. This study demonstrated the complexity of stereotypy, and the importance of careful functional analysis procedures. It is a priority to ensure that the i ndividual is kept safe when implementing functional analyses. For example, if severe a ggressive behavior is the target behavior, it can be useful to examine the precursors res ponsible for the target behavior, to prevent serious injury (Smith & Church ill, 2002). When examining aggressive behaviors leading to physical injury (Fisher et al., 1998), the objects used during th e functional analyses were of safe materials and could not ha rm the child if they were broken. Functional Analysis in the Natural Settings Functional analysis can be conducted in a wide variety of settings. Often, researchers prefer to have tight control over the settings. This environment is known as a controlled or analogue setti ng (Philips & Mudford, 2008). Th e controlled setting has its advantages mainly because it limits the num ber of confounding variables. Experiments


7 with the controlled setting are done by conduc ting the functional anal ysis in a “closed” environment, meaning being clos ed off from the natural environment which the client is usually exposed to (Hanley, Iwata, & McCo rd, 2003). The typical controlled setting might be in a clinic with limited distractions (or confoundi ng variables). An example of this can be a small room with a table and tw o chairs. Materials in the room would only be those which correspond to the condition (O’R eilly et al., 2006). For example, in the tangible condition, preferred items would be pr esent in the room. Before the attention condition, the preferred items are removed from the room, and then the attention condition begins. Other condi tions, variables are removed if they do not correspond to the condition being tested. This promotes a gr eat deal of control ov er both the client and the environment (O’Reilly et al., 2006). An issue concerning functional analysis is that the problem of concern may not occur in the controlled analogue settings (Call et al., 2005). For example, Roantree & Kennedy ( 2006) conducted a functional analysis on a child with severe mental retardation in an analogue setting which was a small room with tables and chairs. Results indicated that the function of stereotypy was attention. However, the function was determined only af ter noncontingent pre-se ssion attention was introduced. Had the functional analysis take n place in a natural setting, manipulation of pre-session attention might have not been requ ired since that is the setting where the stereotypy usually occurred. English & Ande rson (2003) compared the effects of using an unfamiliar therapist versus familiar caregivers during functional analysis with children with developmental disabilitie s. Prior to the study, the caregivers were trained to implement functional analysis. Re sults indicated that for 3 of the 4 participants, rates of responding were different depending on who implemented the functional analysis.


8 Sheridan et al. (1996) also showed sim ilar results. The researchers conducted a social skills intervention for school childre n with autism, in which intervention was implemented both in analogue and natura listic conditions. An analogue condition involved performing 48 social scenarios with a confederate child of the same age as interventionist. The naturalistic observations were conducted during morning and recess periods at the child’s school The target behaviors were social entry, maintaining interactions, and solving problems. From ba seline to treatment phases, results of the analogue condition indicated improvement in a ll target behaviors. Intervention during the natural conditions did not produce the same results, with no improvement for social entry and solving problems skills across baseline to treatment phases. Conducting functional analysis sessions dur ing natural routines at home has been emphasized in the literature (Jay et al., 2005; Richman & Lindauer, 2005). Ellingson et al. (2000) conducted a functional analysis wi th a high functioning child who engaged in excessive thumb sucking. It was revealed th at the child engaged in thumb sucking while alone, usually while watching television in th e living room. Theref ore, the functional analysis was conducted at her home in the living room. Results indicated significantly higher rates of thumb sucking during the al one condition, which suggested the function was maintained by automatic consequences. This was further confirmed when bandages were put on the finger and thumb sucking dr opped down again, and only occurred at high rates when the fingers were exposed. An Awareness Enhancement Device (AED) was placed on gloves, and thumb sucking dropped to n ear zero levels in the treatment phase. An important factor when using the natura l environment in functional analyses is deciding who will conduct the analyses (Ringda hl & Sellers, 2000). As shown by several


9 researchers (Brousard & Northup, 1997; Lewi s & Sugai, 1996; Northup, 1997), problem behavior of children may be se nsitive to attention provided by peers, but not by adults. The results of these studies indicate that unless the source of atte ntion is accurately identified by involving peers in the functi onal analysis, the inte rvention would not be effective. Likewise, stimuli are specific to certain homes and caregivers are specific to certain individuals. Since caregivers are th e people who must deal with the problem behavior. Involving families or caregivers in the functional analysis will not only contribute to accurate identification of the be havioral functions, but also increase the social validity of the analysis and interv ention (Martin et al., 1999). Several studies included parents in the functional analysis. Vollmer et al. (1996) conducted a functional analysis on children with severe tantrums a nd limited speech. In an outclinic setting they used the parents to perform the functional an alysis and determined that for one of the children, the function of the tantrums was to obtain attention from the mother and preferred items such as toys. Treatment involved using al ternative forms of communication as replacement behaviors for the tantrums. The maternal attention observed between mother and son may have not been observed if a therapist, rather than the mother, performed the functional analysis. Future studies can use several therapists (e.g. mother, father, teacher, therapist) and ex amine if the same functions of behavior are observed across therapists. It is relevant for people with whom problem behaviors occur the most to conduct the functional analysis, sinc e this most closely resembles interactions in the naturalistic context. Functional Analysis of problem behaviors ca n also occur in other settings such as a school classroom. Mueller et al. (2001) trea ted a child with the problem behavior of


10 hand flapping in the classroom. In this study, it was hypothesized that the function of hand flapping was associated with class dema nds. The teacher was instructed to place a task demand on the child throughout the cour se of a normal class period. Results indicated that high levels of hand flapping occurred when task demands were low, and low levels of hand flapping occurred when task demands were high. His stereotypical behavior occurred even in the absence of social consequenc es (e.g., attention), suggesting that the child enjoyed performing the tasks and was more likely to engage in hand flapping when task demands were low. It is also possible that ha nd flapping resulted in lower level rate of demands. Another po ssible explanation for the high rates of stereotypical behavior during high task demands is that it is possible that his hand flapping was automatically reinforcing, but performing the tasks was found to be more reinforcing which competed with his st ereotypical behavior. Stereotypy of Individuals with Autism Stereotypy can be defined as responses with properties that ar e relatively invariant over successive occurrences (C atania, 1998). According to Lewis & Baumeister (1982), stereotypy can be defined as repetitive or i nvariant behavior that serves no apparent social function. Stereotypic behavior is co mmon in individuals with autism as well as other developmental disabilities. Func tional analysis is a common method of determining the function of ster eotypy. Stereotypic behavior is viewed as core symptoms of autism. These behaviors can prevent indi viduals with autism from acquiring new and useful behaviors, as well as isolating them from inclusive settings (Jones Wint, & Ellis, 1990; Wolery et al., 1985). If unaddressed as early, the stereotypic behaviors can be a


11 precursor to more serious probl em behavior such as self-inj ury (Morrison & Rosales-Ruis, 1997). Topographies of these behaviors can in clude but are not limited to lining up objects, body rocking, repetitiv e verbalizations, hand mouthing, hand flapping, and hair twirling (Mueller, Sterling-Tu rner, & Scattone, 2001). Stereo typical behaviors can serve the function of escaping from tasks, because of their aversive nature towards caretakers, therapists, and school teacher s, as this has been dem onstrated through the use of functional analysis (Durand & Carr, 1987). Caretakers will often leave a child alone when they engage in stereotypical behavior and this can produce an escape function. There are studies reporting that stereot ypical behaviors can be used for sensory stimulation (Lovaas, Newsom, & Hickman, 1987). However, studies have also demonstrated the social reinforcement eff ects on stereotypical behavior (Roantree & Kennedy, 2006). Tang et al. (2002) examined stereotypical behaviors of a child with autism who engaged in stereotypical ear covering. Through descriptiv e analysis it was hypothesized that this behavior might be in response to another child’s screaming. Functional analysis confirmed that stereotypical ear covering only occurred in the presence of another child screaming. Durand & Carr (1987) conducted functional analyses using 3 different experimental conditions. These conditions were the baseline condition, decreased attention condition, and increased task diffi culty. Matching to sample and receptive identification tasks with picture cards were used and were separated into easy and difficult, and the easy tasks were used duri ng baseline, while the difficult tasks were placed into the increased task difficulty condi tion. Attention in the form of commands


12 and praise was delivered con tingent upon correct responses on a variable ratio schedule of VR 3, and stereotypic beha vior was ignored. In the de creased attention condition, praise and commands were si gnificantly reduced during task s. In the increased task difficulty condition, praise and commands were delivered just as they were in baseline, except the tasks used were identified as di fficult. The increased task difficult condition had the highest frequency of stereotypical beha vior. This suggested that the stereotypic behavior served as a function of avoiding difficu lt tasks. In the second experiment of this study, the same baseline condition was use d. The other condition was the time-out condition. In the time-out condition, the ta sk was briefly paused (time-out) and the experimenter ignored the part icipant. The results showed an increase in stereotypic behavior in the time-out condi tion. These results suggest that stereotypic behavior functioned as escape from the task. Mace & Belfiore (1990) also found similar result with their study. They used functional analys es on a 38 year old lady with stereotypic behavior. The functional analyses demons trated that stereotypic behavior was maintained by termination of demands. When stereotypy occurs at high rates in an alone condition, it is often hypothesized that the function is self-s timulatory (Repp, Felce, & Barton, 1988; Sidener, Ca rr, & Firth (2005). Preference Assessments Preference assessments are a vital component of putting together a treatment plan when determining effective reinforcers fo r the individual (Paramore & Higbee, 2005). While reduction of problem behavior might be beneficial to the cl ient, this must be replaced by positive and/or replacement behaviors. Replacement behaviors are often rewarded by various reinforcers (e.g. praise, edib les, etc.), and each child is an individual


13 with different preferences for reinforcers. Thus, it is important to determine which items a child is interested in obta ining, more importantly, which items they want in exchange for positive behaviors. Typical preference assessment procedures systematically present reinforcers to determine the participant’s preferred item. This item is then used as a reinforcer during functional analysis and treatment. Ther e are several commonly used preference assessment procedures. For the Pace Proce dure (also known as single item procedure) one item is presented at a time (Pace et al., 1985). Choosing the item signals they prefer it. For example, there may be a total of 16 items, and each item is presented 10 times in random order. The number of times each item is presented is divided by the number of times it is approached. A limitation of this preference assessment is that it may overestimate preference, since it is possible fo r all items to be chosen when presented (Fisher et al., 1992). The Forced Choice Procedure, also know n as paired choice, (Carr, Nicolson, & Higbee, 2000) involves showing 2 items at a time and waiting, and then allowing the participant to choose one of the two items. A group of items is selected and then each item is paired with the other in the group. For example item A is paired with item B, item A paired with item C, item A paired with item D, etc. An advantage of this procedure, unlike the Pace Procedure, is that it is not po ssible for every item to be chosen 100% of the time. The participant must choose one of two items, which can create a hierarchy of preference, and items can be ranked relative to each other. In a study comparing the Pace Procedure and Forced Choice Procedure, Fish er et al. (1992) conduc ted both of these preference assessments. After the preference assessments were conducted, two chairs


14 were available to sit in. Depending on which chair they sat in, th ey received access to a certain item used in the preference assessment. Using this concurrent operant approach, they determined the Forced Choice Proce dure was significantly more accurate at predicting which item the participant would choose, while the Pace Procedure overestimated the preference of the participants. Harding et al. (2002) used reinforcer assessments to determine which reinforcers were responsible for specific behaviors. High lights how multiple reinforcers, rather than just one reinforcer may be responsible fo r problem behaviors, and how multi-component interventions are important. Zarcone et al (1999) assessed the effects of negative reinforcement during tasks. The study provides good information on reinforcer assessment and how they can be used in inte rventions aimed at de creasing destructive behavior. Poling (1987) used reinforcer asse ssments containing forced exposure as well as choice trials. The study was unique in the fact that it used more than one type of preference assessment, rather than just one which is common practice. Considering the fact that an item might have a different va lue in an analogue condition as opposed to an individual’s natural environm ent, it is important that the item determined in the preference assessment, which will be reinforcing the problem or competing replacement behavior in the natural environment shoul d be used during functional analysis and intervention. Items that are reinforcing in one environment might not be as reinforcing in another environment. Schedules of Reinforcement Schedules of reinforcement can be creat ed and manipulated in several ways during functional analysis. The vast majority of studies administered consequences on a


15 continuous reinforcement (CRF) schedule (Hanle y et al., 2003). With a CRF schedule, a programmed reinforcer is delivered after th e occurrence of each problem behavior. Studies that used intermittent schedules of reinforcement, delivered the reinforcer sporadically. This might be observed in the naturalistic environment as well, since often the caregiver does not deliver reinforcement af ter each incident of problem behavior (Lali & Casey, 1996). Issues with the intermittent schedule of reinforcement are that since the reinforcer is delivered intermittently, it is possible that the reinforcement might not be delivered sufficiently after the problem beha vior, which might extend the length of the functional analyses (Han ley et al., 1996). Tiger, Hanley, & Heal (2006) conducte d a study using multiple schedules of reinforcement to determine school children’s preferences. They used different color cards presented to the experimenter, and each card was associated with various schedules of reinforcement. The conditions included: two stimuli (e.g. different colored cards and leis around the experimenter’s neck) in which one was associated with reinforcement and another with extinction; only one stimuli wa s presented which was associated with reinforcement (continuous sc hedule of reinforcement; and the mixed schedule in which stimuli associated with reinforcement and extinction were not presented. The children who were able to discriminate better when the stimulus associated with extinction was presented preferred the mixed schedule of reinforcement. Children who were not as accurate at discriminating the stimulus a ssociated with extinction preferred the reinforcement only condition (continuous schedu le of reinforcement). Paisey, Whitney, and Hislop (1991) embedded trials of their func tional analyses within the natural routine (e.g. lunch time) of adults with intellectua l disabilities who engaged in aggressive


16 behavior. Due to the high frequency and se verity of a participant, they used an intermittent schedule of reinforcement (FR3). Results indicated aggressive behavior under all conditions, which suggested that he was under the control of social contingencies (e.g. disrupted during lunchtime) regardless of the condition. Therefore, future research in functiona l analyses in natural settings should conduct the analyses using either intermittent or continuous sc hedules of reinforcement depending upon the severity of the individual’s problem behavi or and the characteristics of caregiver responses on problem behavior. Magnitude of Reinforcement The term magnitude can refer to several dimensions including the size, quantity, or number of an object or time (Hoch et al., 2002). It can also make reference to the importance, or quality of an object. Rein forcement is defined as a consequence that follows an operant response that increase (or attempts to increase) the likelihood of that response occurring in the futu re (Skinner, 1953). Magnitude of reinforcement can be defined as “The rate, quality, intensity, or dur ation of reinforcement” (Hoch et al., 2002). Reinforcement itself has been manipulated, a nd examined in countless studies. It can be thought of as one of the most fundamental as pects of behaviorism. Using reinforcement is of extreme importance, if not entirel y required, when developing a behavioral intervention plan. This is obvious to most researchers and reinforcement has been examined in a variety of ways. However, few studies have directly evaluated reinforcement magnitude. The few studies th at have made magnitude of reinforcement the main focus of their studies have yielde d inconsistent results. Some studies have found that an increase in magnitude of reinfo rcement results in an increase in response


17 rates (Jenkins & Clayton, 1949; Reed, 1991). On the other hand, numerous studies have found that an increase in magnitude produces a decrease in response rates, which is an inverse relation (Belke, 1997). For example, Volkert et al. (2005) co nducted functional analyses on 6 children with developmental disabilitie s. Engagement in problem behaviors was examined during each condition. The researchers tested the effects of varying the magnitude of reinforcement duration during functional anal ysis using 3 sec., 20 sec., and 120 sec. reinforcement durations. Access to reinforcer s varied with 3 different lengths of time contingent on problem behavior. With the gr eatest length of time the children had access to reinforcers, rates of problem behavior were lowest. The results indicated no difference in the identification of problem beha vior across the 3 sec., 20 sec., and 120 sec. duration. On the other hand, other studies ha ve found an inverse relation in reinforcer duration and responding (Reed, 1991; Staddon, 1970). There are several explanations as to why the results are inconsistent with each other. A simple answer might be that not enough studies have directly examined magnitude of reinforcement. When examining any aspect of behavior results will vary due to many environmental variables (Volke rt, Lerman, & Vorndran, 2005). With such few studies being done on magnitu de of reinforcement during f unctional analysis with the target behavior of stereotypy, we can not reas onably come to a solid conclusion yet, and it is possible that certain proce dures were not valid or reliable The more research is done on the magnitude of reinforcement, the grea ter reliability we ha ve on the conclusions drawn from all studies. Another ex planation of the inconsistent results is that most of the studies used reinforcers different than what is usually evaluated in the functional analysis


18 for problem behavior. For example, access to candy rather than escape from demands or attention might be used for a child who is tr ying to escape from demands or attempting to receive attention. An inaccura te use of reinforcement will not produce valid results. The function of a problem behavior must be dete rmined, and the proper form of reinforcement addressing that behavior needs to be us ed in the functional analysis. Recently, there have been several studie s that directly and systematically examined the magnitude of reinforcement. Volkert et al. (2005) directly examined reinforcement magnitude, the dimension of duration of reinforcement. This was done during the functional analysis of children with mental disa bilities to examine if the duration of reinforcement effects the identifi cation of problem beha vior. Volkert and her colleagues used three different durations of reinforcement: small duration (3 seconds), medium duration (20 seconds), and large dura tion (120 s). Different durations can be related to durations of reinforcement in the natural environment. For example, a small duration of 3 seconds might be equivalent to verbally praisi ng a job, such as saying “good job” and smiling, or a pat on the shoulder. The medium duration of 20 seconds can be similar to having a short convers ation with the child. Asking the child what they made in art class or tickling a child may last approxi mately 20 seconds. The large duration of 120 seconds can be closer to the length of break time given for in betw een tasks, or sitting together and watching television. Although 120 seconds can be considered a long period of reinforcement, it may not even come clos e to the amount of reinforcement that might be seen in the natural environment (Lindberg et al., 2003). For example, a parent might play sports with their child in the park for 30 minutes at a time, and maybe even longer. A child might find a gym class at school wh ich lasts 50 minutes to be a stimulating


19 source of reinforcement. If a child is given only 2 minutes of a 50 minute gym class, it may have an adverse effect on behavior. 120 seconds of reinforcement might only be a fraction of the durations of reinforcement we see in the natural environment. However, for the purposes of conducting a functional an alysis, it is understandable to use 120 seconds for the “long” duration. This also ma kes a functional analysis more practical in most cases. The results of the functional analysis concluded that there were no differences in the identification of problem beha vior. It is possible th at these results were due to reinforcement duration be ing arbitrarily manipulated. This significant finding however, came w ith a few limitations. There were only 14 to 16 sessions for each participant. More sessions may have produced satiation effects (Lindberg et al, 2003). Also each session onl y lasted for 10 minutes. The session length might have affected the levels of target behavior during the analyses. In a study with children with intellectual disabilities, Roscoe et al. (2003) observed lower response rates of behavior during 30 minute se ssions when compared to 10 minute sessions. The results indicated that larger magnitudes of reinfo rcement produced a lower rate of responding. Wallace and Iwata (1999) manipulated session lengths of 5 min., 10 min., and 15 min. for individuals with mental reta rdation. The 10 min. and 15 min. sessions produced the same results. However there were a few discre pancies between the 5 min. and 15 min. sessions, with an increased frequency of problem beha vior towards the later parts of the 15 min. sessions. In the natural environment, sessions may last much longer. For example, a therapist or caregiver conducting discrete trial training at an in home setting might work with the child for up to 4 hours. Although br eaks may be provided in between each work


20 period, the therapist might wish to work w ith the child for 30 minutes at a time to increase aspects of attention or vigilance. This is important when looking at the format of teaching in school classrooms. Class periods typically last approximately 50 minutes. A child might be able to perform well for th e first 10, or perhaps even 20 minutes. For example, if the function of a child’s behavi or is to gain attention, a 10 minute session might not be sufficient to determine this func tion. Children are different and some might be able to go longer without attention than others. For example, a child may not engage in problem behavior for the first 10 minutes of a class period. However, after 20 min. of deprivation of attention, from the teacher or peers, may lead to problem behavior. If a classroom period lasts 50 minutes, a 10 minut e functional analysis session is only onefifth of that class period. Therefore, future studies should conduct the functional analysis in conditions more similar to the naturalist ic setting, which can be done by manipulating several variables. It is possibl e that behaviors that are not exhibited early in the session might start to occur and increase towards the end of the session. Increasing the session to longer than 20 minutes however, might be impr actical since it can interfere with other routines that caregivers wish the participant to engage in such as household chores, homework, discrete trial thera py, etc. Reinforcer duration shou ld also be manipulated to examine any differences across the durations. With a session length of longer minutes, it may be more reasonable to see problem behaviors occur with long durations of reinforcement (e.g. 120), which might not be ob served with shorter session lengths (e.g. 5-10 min.). Varying durations of reinforcement can match the different durations seen in the naturalistic environment. For example, a 3 sec. duration is approximately the amount of time a caregiver might provide attention in the form of a short statement, such as


21 “Don’t do that”. A duration of 20 second mi ght be equal to atte ntion being provided from a caregiver with by a coupl e of sentences, or the amount of time it takes to finish a small candy. The 120 second duration is rare ly used in functional analysis, although many instances of this duration, and even longer, occur every day in a child’s environment. Hugging a child and watching tele vision together, or be ing sent out of a classroom and avoiding a lesson can last wa y beyond 120 seconds. Varying both session length and duration of reinforcement during f unctional analysis can more closely match the environments, and increase ecological validity. Functional Analysis-Based Inte rvention to Reduce Stereotypy Intervention designed based on the functiona l analysis results closely link to the functions of stereotypy. For example, Ahearn et al. (2007) used functional analysis to determine the function of vocal stereot ypy in children with autism. After the identification of the function of stereotypy, the participant’s teachers were trained to use a response and interruption and redirection program, which was used to replace the stereotypy with appropriate voc alizations. Contingent upon vocal stereotypy, the teacher presented the children with simple vocal demands such as “What color is your shirt?” These vocal demands were presented until the child answered 3 consecutive times without an incident of vocal stereotypy. Results indicated a significant decrease in vocal stereotypy and an increase in appropriate vocalizations. Sidener, Carr, & Firth (2005) performe d a functional analysis on children who engaged in stereotypy. Stereotypy was highest in the alone condition and it was determined the function was automatic rein forcement. The researchers found that an environmental enrichment procedure signifi cantly reduced stereotypy. This procedure


22 involved access to three highly preferred it ems, suggesting that a reinforcer rich environment competed with stereotypical behavior. Through function based interventions, alte rnative behaviors can replace problem behaviors. Kennedy (2000) determined that stereotypical behavior was observed in the attention, demand, and no attention conditions It was hypothesized that the problem stereotypical behaviors occurred to communi cate for attention and escape demands. In the treatment phase, the alternative beha vior of signing was introduced and was reinforced, while the stereotypical behavi or was ignored. Significant decreases in stereotypical behavior were obs erved in the treatment phase. Family involvement has become incr easingly common in the function-based intervention literature. Parents have actively i nvolved not only as informants or assessors, but also as intervention agents in the process of intervention development and implementation (Arndorfer, Miltenberger, Wo ster, Rortredt, & Gaffa ney, 1994; Derby et al., 1997; Vaughn, Clarke, & Dunlap, 1997; Wa cker, Cooper, Peck, Derby, Berg, 1999). Arndorfer et al. (1994) involved pa rents in the brief experiment al analysis in the context of family routines in the home. Derby et al. (1997) involved moth ers in the functional analysis and intervention procedures. Lu cyshyn, Albin, & Nixon (1997) targeted four family routines in the home and community i nvolving parents of a 14year-old-child with multiple disabilities in comprehensive assessment including the experimental functional analysis and intervention procedures. In tensive training and maintenance support was provided to the parents during the implem entation of the intervention. Although the family involvement has been emphasized and valued in the literature, only a small


23 number of research studies have reported family involvement both in the functional analysis and intervention procedures (Dunlap, Newton, Fox, Benito, & Vaughn, 2001). Purpose and Research Questions The purpose of this research is to exam ine the effects of varying duration of reinforcement and session le ngth during functional analyses on stereotype of an adolescent with autism and the effects of intervention designed ba sed on the functional analysis results on his stereotype and re placement behavior. This study extended the literature by (a) examining the interaction e ffects of reinforcement duration and session length on stereotypy and (b) involving the family in the functional analysis and intervention procedures during a natural fa mily routine. Questions addressed were whether (a) the identification of function of stereotype would be di fferent across varying session lengths and varying dura tions of reinforcement; (b) the combinations of session lengths and durations of reinforcement w ould produce different results; and (c) the function-based intervention based on the func tional analysis results would lead to a decrease in stereotypica l behavior and increase in communicative behavior.


24 Chapter 2: Method Participants The participant in the study was Mike, a 19 year old adolescent diagnosed with autism by an independent psychologist according to the criteria of Diagnostic and Statistical Manual of Mental Disorders ( DSM-IV-TR ; American Psychiatric Association, 2000). Mike’s intelligence was reported to be 2.0 below the mean on the Standord-BinetIV (Thorndike, Hagan, & Sattler, 1986). The Vineland Adaptive Behavior Scale (Sparrow, Balla, & Cicchetti, 1984) indicated that he functioned at 2-ye ar-old level. On the Childhood Autism Rating Scale (Schopler, Reishler, & Renner, 1988), he scored in the severe range. Mike has also taken a co mmunication evaluation. For the ReceptiveExpressive Emergent Language Scale (REEL) Mike scored 7-9 months on Receptive Language and 20 months on Expressive La nguage. On the Sequenced Inventory of Communication Development (SICD), he scored 20 months for Expressive Communication, and 8 to 16 months for Recepti ve Communication. Mi ke lived with his parents in upper middle class neighborhood, a nd attended a special school for children with autism and other related disabilitie s. Mike had been on and off of several medications within the past year which in clude Depakote, Risperdal, Clonidine, and Melatonin. He appeared to be in good physic al health, and received medical care on a regular basis from his primary care physician. He had limited vocal skills, and only spoke the first phoneme, or at most two phonemes of a word. He performed well at tacting objects, and when asked to vocally identify an object, he could say the first phoneme of


25 the object, but had trouble pronouncing the entire word. His fine motor skills were rather poor and often needed physical guidance to pe rform such tasks. Mike’s community safety skills were very limited which put him at serious danger in th e community without appropriate supervision. Mike frequently engaged in stereo typical behavior at home and school, and had a history of doing so most of his life. He had worked with speech therapists, neuropsychologists, and most r ecently ABA therapists. Most of his recent therapy had focused on skill acquisition such as speech, motor skills, and social skills. The discrete trial method had been used for the majority of his skill acquisition. He did not have a history of ther apy to reduce his stereotypy. Setting The functional analysis and treatment were conducted in the participant’s home. The routine that was targeted for analysis a nd intervention was transition from meal after school to house chores with his mother. The transition period occurred after Mike was done with his first meal afte r school, which was usually eat en soon after being dropped off from his school bus. Genera lization sessions were conducte d during transition activity time that occurred after the discrete trial activ ities with his therapist. Functional analyses were conducted in the living room area, wh ere Mike spent most of his time during transition. During transition time, Mike did not engage in any structured activities, spending his time mostly walking around the house, laying down on the couch, and watching television, while he fre quently engaged in stereotypy. Response Measurement Stereotypical behavior fo r functional analysis and in tervention, and replacement communicative behavior for intervention was measured using a 10-s partial interval


26 recording procedure. Data on the target be haviors was converted to a percentage of intervals for each session. Stereotypical be havior included occurrence of hand flapping, repetitive vocalizations, jumping up and dow n, stomping feet, saying “wa-hee”, twirling an object, and shaking an object. Hand flapping was defined as rapidly moving an open hand back and forth at leas t twice within 5 seconds. Repetitive vocalizations were defined as repeating the same sound at leas t two times within 5 seconds. Jumping up and down was defined as the feet leaving the gr ound and the body moving ve rtically, unless the current activity requires jumping (e.g. jumping jacks, trampoline, etc.). Stomping feet was defined as lifting one or two feet off the ground and hitting them against the ground at least two times within 5 seconds. Saying “wa-hee” is defined as the client verbally saying the phrase “wa-hee”. Twirling an object was defined as holding an object in hands or fingers and moving the object around in a circular motion at least twice in 5 seconds. Shaking an object was defined as holding an obj ect in hands or fingers and moving it back and forth at least twice within 5 seconds. Communicative behavior was defined as using appropriate vo calizations (e.g., saying the name of a preferred object or activity, or a close approximation) or pointing to an item or a specific picture of item or activity to request items or activities on verbal prompts (e .g., What do you want?). To be considered a correct communicative response, the participant would have to clearly indicate a specific preferred activity or item upon being prompted or independently, and the parent would have to clearly understand his response. A response must indicate which item or activity Mike wants on the first attempt. For example, if Mike’s mother asked him “What do you want?” and the participant vo cally said the item or pointed to the item and the mother understood what he said or pointed, that would qualify as a


27 communicative response. However, if the mo ther did not understa nd, and then said “Do you want to go bike riding”? and Mike responded “yes”, this would not qualify as a correct communicative response. A correct co mmunicative response must be clear to the person providing the activity or item, and not require guessing or a process of elimination by asking a series of potential item s or activities that the child might be interested in. Procedural Integrity Therapist and parent procedural integrity was measured to determine the accurate implementation of functional analysis and intervention procedures. It measured the therapist’s and parents correct responses contingent upon stereotypical behavior. For example, during the attenti on condition of functional analysis, only the delivery of attention contingent on problem behavior was marked as a correct response. In addition, the delivery of the correct corresponding rein forcer within 3 seconds and the delivery of the reinforcement duration matching the durat ion designated for that condition within 2 seconds were recorded. For example, if th e condition was for the 20 sec. reinforcement duration, then reinforcement must have b een provided between the ranges of 18-22 seconds. Also, session length must have lasted within no more than 5 seconds of its designated condition. An example of this is for the 8 min. session to last no less than 7 min. and 55 sec., and no more than 8 min. and 5 sec. During the attention condition of the functional analysis, attention delivery was defined as providi ng social attention in the form of statements such as “Stop doing that ”, or “Hey what are you doing?” contingent upon stereotypical behavior, and would last for 3 sec., 20 se c., or 120 sec. depending on the designated duration of reinforcement fo r that session. During tangible condition, tangible delivery was defined as access to a pr eferred item contingent upon stereotypical


28 behavior for 3 sec., 20 sec., or 120 sec. depending on the designated duration of reinforcement for that session. During the intervention phase, parent deli very of sensory extinction occurred by removing stimulation directly produced by the stereotypy such as a pen, string, popsicle stick, and sock. Creating an enriched envir onment by providing activities with parents, and delivery of correct prompts, time delay, and reinforcement contingent upon communicative behavior were measured. Percen tage of correct use of procedures was measured to determine the procedural integrity by dividing the number of correct use of procedures by the total numb er of opportunities to impl ement the procedures and multiplying by 100. The results of fidelity asse ssment indicated that both implementers (researcher and mother) did adhere to all trea tment procedures scoring a mean of 98% for the researcher and 94% for the mother, ove rall mean of 96% and overall range of 92%100%. IOA for treatment fidelity was 94%. Data Collection Procedures and Interobserver Agreements 46% of the functional analysis and intervention sessions were video recorded, and the video recorded sessions were scored by the researcher and a data collector (a graduate student in the Applied Behavior Analysis program). The data collectors scored the sessions using paper and pencil while they liste ned to an auditory cue emitted from an audiotape. They practiced observations using the 10-s interval recording procedure while they observed prerecorded video sessions. Upon attaining a minimum criterion of 90% across behaviors, they concu rrently but independently, scor ed the video-recorded data. Interobserver agreements were obtained across conditions and behaviors. Mean interobserver agreement for FA was 97% (ra nge, 90% to 100%). Mean interobserver


29 agreement for stereotypy during baselin e and intervention were 96% and 94%, respectively. Mean interobserver agreement for communicative behavior during baseline and intervention were 98 % and 100%, respectively. Preference Assessment The Forced Choice Procedure (Fisher et al., 1992) was used prior to each session to determine highly preferred items to be us ed in the functional analysis phase. Three items (i.e., jelly beans, crackers, and potato ch ips) and 3 activities (i.e., ball tossing game, reading comic books, and doing picture card task s) determined to be highly preferred through an interview with the mother and obs ervations by the researcher were used. Each item was paired with every other item in the group until all possible combinations were used. The order in which the items we re paired and presented was in randomized order to control for sequence effects. For ex ample item A was paired with item B, item A was paired with item C, Item B was paired wi th item C, etc. Each item was presented for approximately 5 seconds. The item that had th e highest percentage of being chosen was used as a reinforcer for the tangible condition of the functional analysis and intervention phase. During each session, the second data collector (a graduate student) recorded whether or not the participant approached a presented item. Each stimulus was presented once with other stimuli in the group for a total of 2 presenta tions. Approaching a stimulus resulted in approximately 5 seconds of access. If the stimulus was not approached within 5 seconds, the participant was prompted to samp le one of the stimuli. If the stimulus was not approached, it was briefly removed and pl aced in front of the participant and was prompted again to make a choice. A total of 3 presentations occu rred prior to each session. The percentage of times each stimulus was approached was divided by the


30 number of times it was presented. The item w ith the highest percentage was considered the most highly preferred item, and was used for that session. Design and Procedures The study used a multi-element treatment design for functional analysis procedures and a reversal design (i.e., AB-A-B’-A design) (Dewein & Miller, 2008) for intervention testing. The revers al design involved baseline, intervention, and intervention with generalization promotion. Descriptive functional assessment: Prior to the functional analysis, Mike’s parents were given the Motivation Assessment S cale (MAS: Durand & Crimmins, 1992). The MAS uses a Likert-type scale and asks res pondents to rate the lik elihood of the problem behavior occurring in various situations (e.g., receiving social at tention, escape from difficult situations, receiving tangibles, and sensory stimul ation). The respondents rate on a scale of 0-7 (e.g. 0 = never, 7 = always). The scores are totaled and the category with the highest score is assumed to be the main taining function of the problem behavior. For example, if the attention category receives the highest score, it is assumed that attention is most likely the function of the problem behavi or. In addition, an in terview with Mike’s parents using Functional Assessment Intervie w Form (O’Neill et al., 1997) and two days of ABC observations (Bijou, Peterson, & Alt, 1968) were conducte d during the target transition time to corroborate the MAS results Results of the MAS indicated that the highest score of 19 was for the sensory function. This was followed by tangible at 17, attention at 15, and escape at 9. The inte rview with the parents and direct ABC observations revealed he engage d in high rates of stereotypy when he was not engaged in any structured activities and was in transiti on periods between activ ities. Therefore, it


31 was determined that stereotypy was least likel y to occur during situa tions that involved performing activities such as house chores with his mother structured leisure activities, and discrete trials with his mother or therap ist, and most likely to occur during situations when Mike had to be left alone and not provided any activ ities. For this reason, the demand condition which was the least likely to maintain the function of Mike’s stereotypy was excluded from th e functional analysis in this study in order to make the functional analysis more efficient, reduc ing the amount of time the participant and implementer had to spend duri ng the functional analysis. Functional analysis. A functional analysis of ster eotypy based on the procedures used by Iwata et al. (1982; 1994) was conducte d, by the researcher and Mike’s mother. Four conditions were tested during in the functional analysis phase: Control (play condition), attention, tangible, and alone. Th e functional analysis manipulated session lengths (Roscoe, Iwata, & Rand, 2003) using 8 min., and 15 min. session lengths. There were 3 durations of reinforcement which were varied: 3 sec., 20 sec., and 120 sec. (Volkert et al., 2005). All possible combinati ons of the functional analysis conditions (control, attention, tangible, and alone) along with the session lengths (8 min. and 15 min) and durations of reinforcement (3 sec., 20 sec., and 120 sec.) were combined during the analysis phase. A total of 24 conditions (4 functional analysis conditions x 2 session lengths x 3 reinforcement durations) were re peated 2 times for each possible combination, once by the researcher and once by Mike’s moth er. Some of the conditions were repeated once more either by the researcher or the mother. During attention conditions, Mike was asked to play with a moderately preferred item, while the implementer was seated appr oximately 4 feet away reading paperwork.


32 Contingent upon stereotypical be havior, Mike was given atten tion in the form of mild reprimands and conversation that varied in length based on the duration of reinforcement designated for that condition (3 s., 20 s., or 120 s.). Prior to the tangible conditions, Mike was given access to a preferred item for 1 to 2 minutes. At the beginning of the session, the experimenter placed the item approximate ly 3 feet from Mike and out of reach. Contingent on the occurrence of stereotypi cal behavior, Mike was given access to a preferred item that varied in length based on the duration of reinfor cement designated for that condition. At the end of the pre-specified duration of reinforcement, the reinforcer was removed, and access was only given contin gent upon the occurrence of stereotypical behavior. During alone conditions, Mike was left alone in the liv ing room. No leisure activities were provided during this condition. The experiment er observed the participant from outside the living room, and no consequences were provided contingent upon stereotypical behavior. During control conditions, the participan t had access to preferred items or activities, and attention from th e researcher or mother. The implementer provided frequent attention in the form of brief verbal comments. There were no consequences for stereo typical behavior. Mike’s mother was involved in all conditions. She was trained to deliver positive (i.e., attention and tangible) reinforcem ent correctly contingent upon stereotype. Modeling, role play with the child, and feedba ck procedures were used during a 2-hour training session. This was done to ensure th at the mother could properly implement the functional analysis procedures. Baseline. Before the implementation of interv ention, baseline data was collected during transition time. Sessions with the therapist we re conducted in the participant’s


33 home in the living room. The room remained the same as it typically was during the baseline phase, with no alterations. Transiti on lasted about 15-20 minutes. For Mike, the transition time involved resti ng in the living room, engaging in stereotypy using a string, pen, or beads. Mike’s stereotypy was ignore d, but sometimes Mike’s mother came to Mike to ask what he wanted and provided activ ities or items that were available. During baseline, Mike’s mother spent her time in the same fashion as she typically would. She spent her time preparing dinner, cleani ng up, answering phone calls, doing paperwork, etc. Baseline observations were conducted dur ing 15 min sessions. Du ring each session, a trained observer ( a graduate student) recorded each instance of stereotypical behavior and communicative behavior using a 10-s partial in terval recording proc edure. 46% of the sessions were video recorded to obtain interobserver agreements. Intervention. Before the intervention was implemented, Mike’s parents worked together with the researcher to develop inte rvention procedures that would reduce Mike’s stereotypy and to increase his communicat ive behavior. The intervention focused on Mike’s behavior when he was most likely to be alone during transitional periods. The intervention package consisted of providing an enriched environment by providing access to preferred activitie s that competed with Mike’s stereotypy (Sidener et al., 2005), removing of stimuli related to Mike’s stereotypy (Rincover, 1978), teaching communicative behavior of re questing activities (Jones, Dr ew, & Weber, 2000; Mace et al., 1992) using prompts, and providing reinforcement contingent upon Mike’s engagement in the activities and communicative behavior. After careful planning between the research er and parents, it was determined that physical activities and tasks would be benefici al to Mike’s well bei ng, as well as reducing


34 his stereotypy. It was observed that Mike often enjoyed an ac tivity that was considered a “chore” (e.g. laundry) rather than an activity that may seem mo re stimulating or leisurely (e.g. bicycle riding). When identifying the c hores, higher preferred ch ores were identified rather than lesser preferred chores since preference in tasks could result in a lower amount of problem behavior (Vaughn & Horner, 1997). Preference for chores was determined through interview with the mother, and chores that were reported to result in the least amount of problem behavior were us ed more often than lesser preferred chores. However, this was not much of an issue, since Mike performed all chores with few instances of problem behaviors. Physical activities and house chores such as doing laundry, jogging on the treadmill, taking out the trash, bicycle riding and rollerblading were selected to be used during interventi on. Soon as Mike was done eating after being dropped off from his school bus, Mike’s mother directed him to engage in a household chore. If a household chore wa s not available at the moment, he was instructed to choose from one of the preferred phys ical activities. The focus wa s on keeping Mike occupied and avoiding giving him the opportunity to be alone and unengaged fr om activities. Mike’s stereotypy would ofte n begin while coming into contact with an object that could be swirled or shaked. These objec ts were small such as a pen, string, popsicle stick, sock, etc. Before the intervention, th ese objects were often in reach and Mike would use them to engage in stereotypy. The parents were instructed to keep these objects out of Mike’s reach as often as possi ble. The intervention also involved teaching communicative behavior and providing rein forcement contingent upon when Mike appropriately engaged in communicative behavior and engaged in activities. An issue of Mike being alone was that he would rarely initi ate communication to request preferred


35 activities. Although his stereotypy was not mainta ined by tangible reinforcers, the parents and the researcher agreed that teaching Mike how to request activities would be meaningful and increase his access to pr eferred activities, which would result in decreases in his stereotypy. Due to his poor verbal communication skills, and inability of signing due to his limited fine motor skills, th e parents and researcher determined to use pictures as a means of initiati ng activities. Pointing to a pict ure of a prefer red activity or verbally sounding the name of the activit y was designated as a proper communicative behavior. First, pictures of preferred physical activities were taken, such as pictures of Mike’s treadmill, trampoline, rollerblades, bicy cle, etc. Pictures of the laundry machine, garbage bin, and garden were also included as preferred activities. The pictures were then placed on a table in the same fashion as they would be during his tacting picture cards task. Before implementing the intervention, Mi ke’s mother taught Mike during six 15 minute sessions to discriminate each obj ect by using a manding procedure (e.g., “Where’s the treadmill?) and asking to point at the picture of the object or pronounce the name of the object while his mother pointed at the card by asking questions (e.g. “What is this?”). The correct response wa s prompted through least to most prompting, and praise was delivered immediately upon the correct re sponse. The researcher assisted Mike’s parents with the training. Once Mike was able to successfully discriminate the picture cards, he was then introduced the contingency of choosing a ca rd by his mother or father. Between 6 to 9 cards (4” x 6” inch cards) were arranged on a table in rows of 3 with each card approximately 1 inch apart from each other. If there were 7 or 8 cards used, then the first two rows closest to the participant had 3 cards in each row and the top row had 1 or


36 2 cards respectively. Mike stood next to the table where the cards were placed and the first row was approximately 6 inches from th e participant. Prompting consisted of being asked “What do you want to do?”, and as soon as he pointed to a card he was immediately guided to that activity. For exam ple, if he pointed to a picture of the trampoline, the parents would say “Alright, let’s go to the trampoline”, and he was guided outside and directed to get on the tr ampoline. Praise for engaging in an activity was given often by the parents, and no atte ntion was given to st ereotypical behaviors The parents received 2-hour training from the researcher before they implemented the intervention procedures to learn how to provide cues, verbal prompts, time-delay procedures, and how to respond to Mike’s communicative behavi ors (Lafasakis & Sturmey, 2007). Training strategies incl uded scenarios, role-play, modeling, and feedback. During training, parents were asked a series of questions providing scenarios of Mike engaging in stereotypical behavior, and what they woul d do in those situations, and they participated in role plays to practice implementation skills. The researcher asked the parents to review the scenarios and inte rvention procedures dur ing the first three intervention sessions and reminded them of what the procedures were. During the implementation phase, the researcher used 4 steps similar to those used by Dib & Sturmey (2007) to support parent implementatio n of intervention: a) caregiver behavior checklist, b) positive and corrective feedback, c) modeling and instructions, and d) continued feedback and modeling. The researcher and the parents reviewed the progress data daily by examining the level and trend of stereotypical and communicative behavior. Both parents were encouraged to provide any feedback and questions regarding the intervention.


37 Modified intervention. Following the intervention and the reversal baseline phases, a replication phase of intervention with generalization promotion was conducted to promote parents’ independent implementati on of intervention strategies and the generalization of Mike’s target behaviors. The intervention invol ved developing picture cards of new items of activities and providing Mike with choices on new activities. Four new picture cards of activities were added to the choice of activiti es he could choose. However, the parents did not re ceive any training or feedback from the researcher during implementation of the intervention. Per centage of stereot ypy and communicative behavior were measured. This data was collected by the researcher, and IOA was determined through the use of a trained secondary observer. Social Validity A Social Validity Scale (see Appendix), adapted from the Treatment Acceptability Rating Form-Revised (Reimers et al., 1992), was used to measure parent acceptance of the intervention procedures. It consists of 15 items that are rated on a 5point scale, where 1 = not at all acceptable, 3 = ne utral, and 5 = very acceptable. Social validity was measured by both parents after th e intervention phase. Both Mike’s mother and father filled out the social validity rating scale at the end of intervention.


38 Chapter 3: Results Functional Analysis The results of the functional analysis are depicted in Figure 1. As seen in Figure 1, the results indicated th at the function of stereotypy wa s the same under both 8min. and 15 min. session lengths regardless of who conducte d the functional analysis. The frequency of stereotypical behavior was significantly higher in the alone c ondition than any other conditions for both session lengths across the ex perimenters (i.e., researcher and mother). Although the tangible condition had the second highest level of stereotypy, the levels of stereotypy during tangible conditions were al so similar in both the 8 min. and 15 min. session lengths. Mike’s mean percentage fo r stereotypy during alone conditions in the 8 min sessions was 68% (83% with mother a nd 58% with research er). In the 15 min sessions, his stereotypy averaged 62% (78% w ith mother and 45% with researcher). His stereotypy occurred at higher rates during sess ions with his mother than during sessions with the researcher. It was f ound that determining the function of stereotypy was also not effected by all 3 reinforcement magnitudes of 3 sec., 20sec., and 120 sec. As seen Figure 1, stereotypy was significantly higher in the alone condition than in any other condition across reinforcement magnitudes. Combinati ons of different reinforcement magnitudes and session lengths did not hinde r the identification of the function of stereotypy. Intervention Outcomes Figure 2 displays the results of the in tervention. During the first baseline phase, when no environmental enrichment, removal of stimuli, communicative behavior


39 instruction, and reinforcement procedures were present, Mike’s stereotypy occurred at a mean percentage of 51% (range = 26% 89%) of intervals. Although there was a decrease in stereotypy in the second session, Mike’s stereotypy increased again in the third session. During intervention when Mike was presented with the environmental enrichment procedure by being allowed to e ngage in a variety of activities with his parents, when the stimuli associated with stereotypy was eliminated, and when the communicative behavior was taught with prompt s with the combinati on of reinforcement procedure, a significant decrea se in Mike’s stereotypical behavior was observed. His stereotypy occurred at a mean percentage of 7% (range = 0% 29%) As seen in Figure 2, Mike’s stereotypy stabilized as the session progressed occurri ng at very low rates. During the repeated baseline phase when the in tervention was withdrawn, his pattern of stereotypy reversed, occurring at a mean pe rcentage of 62% (range = 48% 93%). The results also indicated that Mike’s communicative skills through the use of pointing to pictures of pr eferred activities or vocaliz ation of word approximation increased significantly during the interventi on phase. During the first baseline phase, Mike did not engage in any communicative behavior across se ssions. He did not initiate requesting when verbal prompts were provide d. On the other hand, when the intervention was introduced, Mike demonstrated the communicative behavior 80% of the time throughout the intervention sessions when verbally being prompted to use the pictures of preferred activities. Mike indi cated his preference on activitie s by pointing at pictures or verbalizing word approximation at 80%. When the intervention was withdrawn during the second baseline condition, Mike did not init iate any requesting during verbal prompts.


40 Modified Intervention During the second intervention with gene ralization promotion, Mike engaged in low levels of stereotypy while maintaini ng high levels of communicative behavior. Mike’s mean percentage of intervals of stereotypy was 7%. His communicative behavior occurred 100%. However, when the interven tion was withdrawn, his stereotypy increased again. During the final baseline condition Mike’s mean percentages of intervals were 43% for stereotypy and 0% for communicative behavior. Social Validity Overall, the parents found the function based interventi on to be very acceptable (see Appendix A). According to their scores they were very willing to carry out the intervention, did not think th ere might be disadvantages following the intervention, felt that a moderate amount of time was needed to carry out the intervention, and were confident the plan would be effective and make permanent improvements in their child’s behavior. They found the procedures to be not all disruptive, liked th e procedures used in the intervention, and felt other staff was help ful in carrying out th e intervention. The parents felt their child experienced a low level of discomfort, they were willing to change their routines to carry out th e intervention, and the behavior plan fit in very well with their existing routine. They felt the behavior plan was effective in teaching their child appropriate behavior and the intervention fit with the team’s goals to improve their child’s behavior.


41 Figure 1 Percentage of Stereotypy across Functional Analysis Conditions 123456Sessions 0 10 20 30 40 50 60 70 80 90 100 MotherResearcherMotherResearcherMotherResearcher 3 s Duration 20 s Duration 120 s Duration 15 Minute Session 7891012 11 Percentage of Intervals 0 10 20 30 40 50 60 70 80 90 100 MotherResearcherMotherResearcherMotherResearcher 3 s Duration 20 s Duration 120 s Duration 8Minute Session 7891012 11 Percentage of Intervals Tangible Control Alone Attention


42 Figure 2. Percentage of Stereotypy and Communicative Behavior across Experimental Conditions 1920 0 20 40 60 80 100 120 123456789101112131415161718Percentage of Intervals Sessions A B A Stereotypy Communicative Behavior B A


43 Chapter 4: Discussion This study evaluated the functional an alysis procedures and function-based intervention for a 19-year old adolescent w ith autism in the home setting. One purpose of the study was to examine the effects of vary ing the duration of reinforcement and session length during functional analyses on stereotypical behavior of the individual with autism in the natural home setting. The second purpos e was to examine the effects of an intervention designed by the re searcher and caregivers and was implemented by the natural caregiver based on the functional analysis results on the individual’s stereotypical behavior and communicative beha vior. Descriptive functional assessment revealed that Mike’s stereotypy was least like ly to occur when activities, regardless of the levels of demand, were provided. The results of the functional analysis involving attention, tangible, alone, and control conditions indi cated that Mike’s stereotypy occurred at higher rates during the alone condition acr oss reinforcement ma gnitudes and session durations than during any othe r conditions, indicating that hi s stereotypy was maintained by automatic reinforcement. Consistent with previous research (Roscoe et al., 2003; Volkert et al., 2005), this study suggests that th e function of stereot ypy can be determined regardless of the magnitude of reinforcement and session length. The levels of stereotypy during the alone condition were similar across reinforcement magnitude and session lengths. The results also i ndicated that the levels of responding were somewhat undifferentiated within the same conditions across different session lengths as well as reinforcement magnitudes However, it may be possible th at the level of stereotypical


44 behavior would increase in longer session durations as indicated by Wallace and Iwata (1999). Wallace and Iwata found that the frequenc y of problem behavior toward the later parts of the longer session increased. During the intervention implementation phase, the function-based intervention, using the environmental enrichment, remova l of sensory stimuli, and systematic communicative skill instruction with the contingent reinforcement procedure decreased the occurrence of stereotypical behavior a nd increased the communicative behavior. The intervention procedures were implemented in the context of a naturally occurring family routine. Before intervention, Mike was mostly left alone during transition time in which structured activities were not provided. Limited activities with family members were available during this time, which resulted in Mike’s stereotypy. Duri ng the intervention, a variety of activities chosen by Mike were provided. Pare nts actively engaged in the activities providing an enriched environment. Physical activities and chores that were preferred by Mike were av ailable during intervention. Study findings suggest that enriching the environment by providing preferred activities and social engagement is an effectiv e intervention for stereo typy if the behavior occurs most frequently when an individual is deprived of access to stimulation (Horner, 1980). Studies have shown that increases in activity engagement may produce the same effects. For example, physical exercises have been found to be effective in decreasing stereotypic behaviors (Ellis, Maclean, & G azdag, 1989; Levinson & Reid, 1993; Watters & Watters, 1980). It is suggested that the non -contingent nature of this environmental enrichment makes it simple for caregivers to implement the procedure relative to the rearrangement of behavioral contingencies (Vollmer, 1994).


45 In addition to stereotypy, Mike’s parent s’ concern was his limited communication skills. Mike rarely initiated communication to indicate what he wanted or what activities he wanted to engage in. Although a variety of possible activities were available for Mike, it was a burden for the parents to constantly fi gure out what he wanted to do. Due to his poor verbal and signing skills, teaching comm unicative skills using picture cards of preferred activities was included in the inte rvention. Mike was taught how to request by pointing to the activity he wanted or verbal ly saying the words when verbally prompted. His communicative behavior was heavily depend ent on verbal prompts, but he was able to request what he wanted. His parents were very excited about having the reciprocal interaction between Mike and them. As Mace et al.(1992) suggested, teaching Mike communicative skills of requesting preferred activities was an effective way of reducing stereotypy that is reinforced by automatic reinforcement. Although his stereotypy was not maintained by the communicative function of obtaining tangible items or activities, teaching him communicative skills contributed to increases in access to preferred activities which result ed in decreases in Mike’s stereotypy. Moreover, the second intervention phase data indicated that Mike successfully used communicative skills to request new activities. His parents were succe ssfully able to implement the intervention procedures with minimum researcher’s support, which promoted Mike’s use of communication skills and access to preferred activities and resulted in decreases in stereotypy and acquisition of communicative skills. Another strategy included in the intervention was the el imination of stimuli that were associated with Mike’s stereotypy. Mike ’s stereotypy with stim ulus objects such as pens, strings, and popsicle sticks often occurred for a large portion of stereotypy prior to


46 the intervention phase. Keeping these objects out of reach and sight in part contributed to a decrease in Mike’s stereotypy. Studies (R incover, 1978) indicate that elimination or extinction of stimuli that ar e directly produced by stereo typy is an effective way of reducing stereotypy. The use of parents as in tervention agents could have a significant impact on the support of individu als with autism within home settings. Family or parent implemented functional analysis and inte rvention would enhance positive social interaction between parents and individuals with autism and contribute to success at school and inclusion in the community. As suggested by previous studies, this study indicates that involving parents actively not only as informants or assessors, but also as intervention agents in the process of inte rvention development and implementation would enhance the intervention effectiveness and pr omote generalization (Lucyshyn et al., 1997; Miltenberger, et al., 1994; Vaughn et al., 1997) Despite the favorable results achieved in this study, there are several limitations that must be addressed in inte rpreting the resu lts of the study. One of the limitations is the difficulty of controlling extran eous variables due to the im plementation of the functional analysis and intervention procedures in the natural family setting. The functional analysis was conducted in the living room of the participant’s home, and the mother conducted a large portion of the functional analysis. Alt hough this can be of great value due to the similarity between the functional anal ysis setting and the real life setting of the participant’s environment, it does come al ong with a loss of control, which is an important part of conducting functional analys is. Certain environmental variables may vary across sessions which cannot be controlle d. For example, visiting family members or unexpected guests might have altered Mike’s behavior in the home. As shown in the


47 functional analysis data, the variability of Mike’s stereotypy in a few sessions during functional analysis and base line was relatively high. Another limitation of the study is the limited data collected during functional analysis and baseline. Although the brief functio nal analysis was found to be effective in identifying the function of Mike’s stereotypy as indicated in previous studies (Cihak et al., 2007), the data were not sufficient to compar e the levels of Mike’s stereotypy during conditions with his mother to those during conditions with the researcher. The third limitation is that the study only examined stereotypical behavior in an in home setting. The frequency of stereotypy while being in public places such as during school was not recorded. Perhaps a reduction in stereotypy in one setting and during one portion of the day may have an effect on the frequency of stereotypy in other settings and during other times of the day. The fourth limitation is that setting even ts were not examined during this study. Changes in the participant’s school day routin es that occurred prior to the routine that was targeted in the study may have had an effect on data during the functional analysis and baseline phases. Routines such as taking a different bus home, getting in trouble at school, and not getting enough sl eep the night before, may have affected the occurrence of Mike’s stereotypical behavior later on in the day. The fifth limita tion of the study is no examination or no instruction of teaching Mi ke communicative skills without prompting during the intervention phase. Mike’s communicative behavior taught and recorded in the study was prompt dependent behavior. Consid ering the fact that Mike was able to successfully use the communicative behavior w ith prompts, the intervention should have


48 included the phase of teac hing Mike spontaneous comm unicative behavior without prompts using time delay procedure. This study has extended the literature by demonstrating th at the interaction effects of reinforcement duration and session length may not change the id entification of the function of stereotypy. Overal l, the function remained the same through all conditions. This suggests that a session length of a pproximately 8 min. and short reinforcement duration (e.g. 3-20 s.) may be sufficient to determine the function of stereotypical behavior (Volkert et al., 2005) This study also demonstrates the importance of involving the family in the functional analysis and in tervention phases, while engaging in natural family routines. While a trained therapist may be able to implement effective intervention procedures, ultimately it is the caretaker’s long term responsibility to ensure these intervention procedures remain effectiv e over the long term. By including the participant’s parents from the beginning of the study, a higher sense of participation and input in the treatment plan may increase th e likelihood of the parents implementing a plan which they had a great deal of creating along with the ther apist (Dunlap et al., 2001). Future studies that conduct functional analysis in a naturalistic setting can examine setting events which occur prio r to the functional analysis sessions. Setting events occur in a naturalistic setting just as the functi onal analysis does. Examining setting events would further increase the social validity of the study, as well as crea te interventions that can be applied outside the home setting. Th e parents of this study expressed great satisfaction with the process and outcomes of the function-based intervention. Although the functional analysis process may have been less socially valid due to the time consumed, this study offers a relatively easily implemented intervention that may assist


49 families in managing stereotypy of individuals with autism. By providing families with effective procedures for their children with autism, the likelihood of successful familybased intervention will be dramatically increased.


50 References Alter, P. J., Conroy, M.A., Mancil, G.R ., & Haydon, T. (2008). A comparison of functional behavior assessment methodol ogies with young children: Descriptive methods and functional analysis. Journal of Behavioral Education, 17, 200219. Andorfer, R. E., Mintenberger, R. C., Wost er, S. H., Rortvedt, A. K., & Gaffaney, T. (1994). Home-based descriptive and experi mental analysis of problem behaviors in children. Topics in Early Childhood Special Education, 14 (1), 64-87. Belke, T. W. (1997). Running and responding reinforced by the opportunity to run: Effect of reinforcer duration. Journal of the Experimental Analysis of Behavior, 67, 337351. Blair, K. C., Liaupsin, C.J., Umbreit, J ., Kweon, G. (2006). Function-based intervention to support the inclusive placements of young children in Korea. Education and Training in Developmental Disabilities, 41, 48-57. Carr, E. G., Nicolson, A. C., & Higbee, T. S. (2000). Evaluation of a brief multiplestimulus preference assessment in a naturalistic context. Journal of Applied Behavior Analysis, 33, 353-357. Carr, E. G., & Owen-DeSchryver, J. S. ( 2007). Physical illne ss, pain, and problem behavior in minimally verbal people with developmental disabilities. Journal of Autism and Developmental Disorders, 37, 413-424.


51 Cihak, D., Alberto, P. A., & Fredrick, L. D. (2007). Use of brief functional analysis and intervention evaluation in public settings. Journal of Positive Behavioral Interventions, 9, 80-93. Crockett, J. L., Fleming, R. K., Doepke, K. J., & Stevens, J. S. (2007). Parent training: Acquisition and generalization of discrete trials teaching skills with parents of children with autism. Research in Developmental Disabilities, 28 23-36. Derby, K. M., Wacker, D. P., Berg, W., De Raad, A., Ulrich, S., Asmus, Harding, J. Prouty, A., Laffey, P., & Stoner, E. (1997) The long-term effects of functional communication training in home setting. Journal of Applied Behavior Analysis, 30 507-531. Dunlap, G., Newton, S., Fox, L., Benito, L., & Vaughn, B. (2001). Family involvement in functional assessment and positive behavior support. Focus on Autism and Other Developmental Disabilities, 16 (4), 215-221.English, C. L., & Anderson, C. M. (2004). Effects of familiar versus un familiar therapists on responding in the analog functional analysis. Research in Developmental Disabilities, 25, 39-55. Fisher, W. W., DeLeon, I. G ., Rodriguez-Catter, V., & Keeney, K. M. (2004). Enhancing Fisher, W. W., Piazza, C. C., & Chiang, C. L. (1996). Effects of equal and unequal reinforcer duration during functional analyses. Journal of Applied Behavior Analyses, 29, 117-120. Fisher, W. W., DeLeon, I. G., Rodriquez-Ca tter, V., & Keeney, K. M. (2004). Enhancing the effects of extinction on attention-maintained behavior through noncontingent delivery of attention or st imuli identified via a competing stimulus assessment. Journal of Applied Behavior Analysis, 37, 171-184.


52 Harding, J. W., Wacker, D. P., Berg, We ndy, K., Barretto, A., & Rankin, B. (2002). Assessment and treatment of severe be havior problems using choice-making procedures. Education and Treatment of Children, 25, 26-46. Healey, J. J., Ahearn, W. H., Graff, R. B., & Libby, M. E., Extended analysis and treatment of self-injurious behavior. Behavioral Interventions, 16, 181-195. Hoch, H., McComas, J. J., Johnson, L., Fara nda, N., & Guenther, S. L. (2002). The effects of magnitude and quality of reinforcement on choice responding during play activities. Journal of Applied Behavior Analysis, 35, 171-181. Iwata, B. A., Duncan, B. A., Zarcone, J. R., Lerman D. C., & Shore, B. A. (1994). A Sequential, test-control methodology for conducting functional analyses of Selfinjurious behavior. Beha vior Modification, 18, 289-306. Jacobson, J. W., Mulick, J. A., & Green, G. (1998). Cost-benefit estimates for early intensive behavioral interventi ons for young children with autism. Behavioral Interventions, 13 201-226. Jenkins, W. O., & Clayton, F. L. (1949). Rate of responding and amount of reinforcement. Journal of Comparative and Physiological Psychology, 42, 171-181. Jones, R. S. P., Wint, D., & Ellis, N. C. (1990) The social effects of stereotyped behavior. Journal of Mental De ficiency Research, 34 261-268. Kahng, S., Abt, K. A., Schonbachler, H. E. ( 2001). Assessment and treatment of low-rate High intensity problem behavior. Journal of Applied Be havior Analysis, 34, 225228.


53 Kennedy, C. H., Meyer, K. A., Knowles, T., & Shukla, S. (2000). Analyzing the multiple functions of stereotypical behavior for students with autism: Implications for assessment and treatment. Journal of Applied Behavior Analysis, 33, 559-571. Lindberg, J. S., Iwata, B. A., Roscoe, E. M ., Worsdell, A. S., & Hanley, G. P. (2003). Treatment efficacy of noncontingent re inforcement during brief and extended application. Journal of Applied Behavior Analysis, 36, 1-19. Lucyshyn, J. M., Albin, R. W., & Nixon, C. D. (1997). Embedding comprehensive behavioral support in family ecology: An experimental, single-case analysis. Journal of Consulting and Clinical Psychology, 65, 241-251. Mace, F. C. (1987). Analysis of demand conditions associated with stereotypy. Journal of Behavior Therapy and Experi mental Psychiatry, 18, 25-31. Mace, F. C., & Belfiore, P. (1990). Behavior al momentum in the treatment of escapemotivated stereotypy. Journal of Applie d Behavior Analysis 23, 507-514. Matson, J. L., & Minshawi, N. F. (2007). F unctional assessment of challenging behavior: Toward a strategy of applied settings. Research in Developmental Disabilities, 28, 353-361. Meuller, M.M., Palkovic, C.M., Maynard, C.S. (2007). Errorless learning: Review and practical applications for teaching ch ildren with pervasive developmental disorders. Psychology in the Schools, 44, 691-700. Moore, J. W., Fisher, W. W., & Pennington, A. (2004). Systematic application and removal of protective equipment in the assessment of multiple topographies of self-injury. Journal of Applied Behavior Analysis, 37, 73-77.


54 Morrison, K., & Rosales-Ruis, J. (1997). Th e effect of object preferences on task performance and stereotype in a child with autism. Research in Developmental Disabilities, 18 127-137. National Research Council. (2001). Educating children with autism Committee on Educational Interventions for Children w ith Autism. Division of Behavioral and Social Sciences and Autism. Wash ington, DC: National Academy Press. O’Reilly, M. F., Edrishinha, C., Sigafoos, J., Lancioni, G., & Andrews, A., (2006). Isolating the evocative and abative e ffects of an establishing operation on challenging behavior. Behavioral Interventions, 21, 195-204. Paramore, N. W., & Higbee, T. S. (2005). An evaluation of a brief multiple-stimulus preference assessment with adolescents with em otional-behavioral disorders in an educational setting. Journal of Applied Behavior Analysis, 38, 399-403. Philips, K. J., & Mudford, O. C. (2008). Functi onal analysis skills training for residential caregivers. Behavioral Interventions, 23, 1-12. Piazza, C. C., Hanley, G. P., Fisher, W.W., Ruyter, J. M., & Gulotta, C. S. (1998). On establishing and reinforcing effects of termination of demands for destructive behavior maintained by positiv e and negative reinforcement. Research in Developmental Disabilities, 19, 395-407. Poling, A., Blakely, E., Pellettier e, V., & Picker, M. (1987). Choice between sequences of fixed-ratio schedules: Eff ects of ratio values and pr obability of food delivery. Journal of the Experimental Analysis of Behavior, 47, 225-232. Reed, G. K., Ringdahl, J. E., Wacker, D. P., Barretto, A., & Andelman, M. S. (2005). The effects of fixed time and contingent schedules of negative reinforcement on


55 compliance and aberrant behavior. Research in Developmental Disabilities, 26, 281-285. Repp, A.C., Felce, D., Barton, L.E. (1988). The treatment of st ereotypic and selfinjurious behaviors on hypotheses of their causes. Journal of Applied Behavior Analysis, 21, 281-289. Roantree, C. F., & Kennedy, C. H., (2006). A paradoxical effect of presession attention on stereotypy: Antecedent attention as an establishing, not an abolishing, operation. Journal of Applied Behavior Analysis 39, 381-384. Roscoe, E. M., Iwata, B. A., Rand, M. S. ( 2003). Effects of reinforcer consumption and magnitude on response rates duri ng noncontingent reinforcement. Journal of Applied Behavior Analysis, 36, 525-539. Sidman, M., (1960). Tactics of scientific research. Boston: Authors Cooperative. Skinner, B. F. (1953). Science & Human Be havior. New York: The MacMillan Company. Tang, J., Kennedy, C. H., Koppekin, A., & Caru so, M. (2002). Functi onal analysis of Stereotypical ear covering in a child with autism. Journal of Applied Behavior Analysis, 35, 95-98. Tiger, J. H., Hanley, G. P., & Bessete, K. K., (2006). Incorporating descriptive assessment results into the design of functional analysis: A case example involving a preschooler’s hand mouthing. Education and Treatment of Children, 29, 107-124. Vaughn, B. J., Clarke, S., & Dunlap, G. (1997). Assessment-based intervention for severe behavior problems in a natural family context. Journal of Applied Behavior Analysis, 30, 713-716.


56 Vaughn, B.J., & R.H., Horner (1997). Identi fying instructional tasks that occasion problem behaviors and assessing the eff ects of student versus teacher choice among these tasks. Journal of Applied Behavior Analysis, 30, 299-312. Volkert, V. M., Lerman, D.C., & Vorndran, C. M. (2005). The effects of reinforcement Magnitude on functional analyses outcomes. Journal of Applied Behavior Analyses, 38, 147-162. Wacker, D. P., Cooper, L. J., Peck, S. M., Derby, K. M., & Berg, W. K. (1999). Community-based functional assessment. In A. C. Repp & R. H. Horner (Eds.), Functional analysis of problem behavior. From effectiv e assessment to effective support (pp. 32-56). Belmont, CA: Wadsworth.Wallace, M. D., & Iwata, B. A. (1999). Effects of session duration on functional analysis outcomes. Journal of Applied Behavior Analysis, 32, 175-183. Ward, R. D., & Higbee, T. S., (2008). Noncontingent reinforcement as treatment for tub-standing in a toddler. Education and Treatment of Children, 31, 213-222. Zarcone, J.R., Crosland, K., Fi sher, W.W., Worsdell, A.S., Herman, K. (1999). A brief method for conducting a negative-reinforcement assessment. Research in Developmental Disabilities, 20, 107-124.


57 Appendices


58 Appendix A: Treatment Acceptabili ty Rating Form-Revised (TARF-R) Please score each item by ci rcling the number that best indicates how you feel about the function-based intervention. 1. Given this child’s behavior problems, how accepta ble did you find the function-based intervention? 1 2 3 4 5 Not at all Neutral Very acceptable acceptable 2. How willing were you to carry out this intervention? 1 2 3 4 5 Not at all Neutral Very willing willing 3. To what extent did you think there might be disadvantages in following this intervention? 1 2 3 4 5 None Neutral Many likely likely 4. How much time was needed each day for you to carry out this intervention? 1 2 3 4 5 Little time Neutral Much time will be needed will be needed 5. How confident were you that the behavior plan would be effective for your child? 1 2 3 4 5 Not at all Neutral Very confident confident 6. How likely do you think this intervention can make permanent improvements in your child’s behavior? 1 2 3 4 5 Unlikely Neutral Very likely 7. How disruptive was it to carry out this intervention? 1 2 3 4 5 Not at all Neutral Very disruptive disruptive


59 8. How much did you like the procedur es used in the intervention? 1 2 3 4 5 Do not like Neutral Like them them at all very much 9. How willing did other family member he lp to carry out this intervention? 1 2 3 4 5 Not at all Neutral Very willing willing 10. To what extent did undesirable side-effects result from this intervention? 1 2 3 4 5 No sideNeutral Many sideeffects likely effects likely 11. How much discomfort did your child experience during this intervention? 1 2 3 4 5 No discomfort Neutral Very much at all discomfort 12. How willing were you to change your routines to carry out this intervention? 1 2 3 4 5 Not at all Neutral Very willing 13. How well was carrying out this intervention to fit into the existing routine? 1 2 3 4 5 Not at all Neutral Very well well 14. How effective was the intervention in teaching your child appropriate behavior? 1 2 3 4 5 Not at all Neutral Very effective effective 15. How well did the goal of the intervention fit with the team’s goals to improve the child’s behavior? 1 2 3 4 5 Not at all Neutral Very much (Adapted from the TREATMENT ACCE PTABILITY RATING FORM—REVISED; TARF-R, Reimers & Wacker, 1988)