Preference Assessments With Individuals With Severe Disabilities: The Utility of Moderate- and Low- Preference Stimuli.

Thesis

Presented in Partial Fulfillment of the Requirements for the Degree Master of Arts in the Graduate School at The Ohio State University By Amanda R. Yeager, B. A. Graduate Program in Special Education

The Ohio State University 2010 Thesis Committee: Dr. Diane M. Sainato, Advisor Dr. Helen I. Malone

Copyright by Amanda R. Yeager 2010

Abstract Three children with mental retardation and/or autism participated in a study evaluating the reinforcing effectiveness of moderate- and low-preference stimuli. Nine full array Multiple-Stimulus Without Replacement (MSWO) preference assessments were conducted assessing tangible and edible stimuli. The items categorized as moderateand low-preference using the Ciccone, Graff, and Ahearn (2005) point weighting scoring method were reassessed. The items ranked high-, moderate-, and low-preference from the assessments were then evaluated during the reinforcer assessments. An alternating treatments design was implemented and data were compared to baseline using an A-B design. Two participants‘ responding increased as the low-, moderate-, and highpreference reinforcement contingencies were implemented. For one participant, the moderate- and high-preference stimuli produced responding at 100%. One participant had minimal responding relative to baseline across all three stimuli. This study extends previous research by demonstrating the MSWO is an effective and efficient technique to identify reinforcers.

ii

Acknowledgements I gratefully thank and acknowledge my advisor, Dr. Diane Sainato, for not only guiding me through my Master‘s studies and research, but for her positive and encouraging demeanor. I thank Dr. Helen Malone for providing me with my graduate research assistantship and writing opportunities in the BBAS lab. I am grateful for her feedback and I thank her for serving on my thesis defense committee. I also want to thank Senny Schnell for her feedback, advice, and encouraging words throughout my research studies. I thank my fellow graduate students who helped me with data collection for this project: Daniel Payne, Pei-Fang Wu, and Courtney Fleming. The time and care they contributed to collecting interobserver agreement and procedural integrity data is much appreciated. I thank the classroom teacher, Debbie, and the staff where the study was conducted for allowing me to work with their students, their effort to identify stimuli for the preference assessments, and their generosity with classroom space and materials. I also appreciate the participants‘ parents for providing consent to participant in research that can benefit more students with developmental disabilities. Finally, I want to thank Hartland for all of his encouragement and my family— Bruce, Kim, and Ashley—for all their support on my graduate school journey.

iii

Vita December, 31, 1986…………………………………………………....Born, Canton, Ohio 2007 to Present…………………………………...…Applied Behavior Analysis Therapist Columbus, OH 2009………………………………………………………………………B.A. Psychology The Ohio State University Columbus, OH 2009 to Present…………………………………….Applied Behavior Analysis Consultant Columbus, OH 2009-2010…………………………………………………… Graduate Research Assistant Buckeye Behavior Analysis Service (PI: Helen Malone) The Ohio State University Columbus, OH

Fields of Study Major Field:

Education

Specializations:

Special Education Applied Behavior Analysis

iv

Table of Contents Abstract……………………………………………………………………………………ii Acknowledgements………………………………………………………………………iii Vita…………………………………………………………………………………...…...iv List of Tables…………………………………………………………………………….vii List of Figures……...…………………………………………………...……………….viii Chapter 1: Review of the Literature………….……………………………………………1 Chapter 2: Method………………………………………………...……………………..15 Chapter 3: Results………………………………………………………………………..29 Chapter 4: Discussion……………………………………………………………………49 References……………………………………….……………………………………….67 Appendix A: Student Information Letter………………………………………………...72 Appendix B: Consent Form…………...…………………………………………………74 Appendix C: Caregiver Questionnaire………….………………………………………..80 Appendix D: Data Sheet for Experimenter and IOA for Preference Assessments………87 Appendix E: Ranking Stimuli Data Sheet…….…………………………………………89 Appendix F: Procedural Integrity Checklist for Preference Assessments….……………93 Appendix G: Procedural Integrity Checklist for Reinforcer Assessments/Baseline…… 95 Appendix H: Procedural Integrity Checklist for Reinforcer Assessments………..….….97 Appendix I: Social Validity Questionnaire………………………………………………99 v

List of Tables Table 2.1

Participants‘ ages and test scores……………………………..………….16

Table 3.1

Mean procedural integrity during the preference assessments…..………30

Table 3.2

Mean IOA for percent correct during preference assessments……..……31

Table 3.3

Mean procedural integrity during the reinforcer assessments………...…38

Table 3.4

Mean IOA of percent correct during reinforcer assessments……….……40

Table 3.5

Mean responding of each participant in each condition…………………45

Table 3.6

Social validity assessment……………………………………………..…47

Table 3.7

General social validity comments…………………………………..……48

vi

List of Figures

Figure 3.1

Stimuli rank during the preference assessments for each participant……33

Figure 3.2

Levi‘s reinforcer assessment data……………………………………..…41

Figure 3.3

Alvin‘s reinforcer assessment data………………………………………42

Figure 3.4

Jake‘s reinforcer assessment data…………………………………..……40

vii

Chapter 1: Review of the Literature

This chapter reviews the research on choice interventions, efficacy of preference assessment formats and components, and examines the use of high-, moderate-, and lowpreference stimuli as reinforcers. Effects of choice interventions Allowing individuals to make choices, particularly individuals with severe developmental disabilities, can be a parsimonious, yet effective way to manage behavior. Choice interventions are increasingly the focus of experimental studies and effectively used to reduce challenging behavior and increase appropriate behavior (e.g., Cannella, O‘Reilly, & Lancioni, 2005; Cannella-Malone, DeBar, & Sigafoos, 2009; Carlson, Luiselli, Slyman, & Markouski, 2008; Cole & Levinson, 2002; LeBlanc, Cherup, Feliciano, & Sidener, 2006; O‘Reilly, Lancioni, & Sigafoos, 2004; Piazza, Fisher, Hagopian Bowman, & Toole, 1996; Tiger, Hanley, & Hernandez, 2006). According to the criterion described by Horner et al. (2005), choice interventions are considered to be an evidence-based practice for individuals with severe to profound disabilities (Tullis, Cannella-Malone, Basbigill, Yeager, Fleming, Payne, & Wu, submitted). For example, Cole and Levinson (2002) compared the effects of embedded choice questions within instructional routines to verbal directives with two students with severe developmental disabilities whose behaviors were typically noncompliant and aggressive. Data were 1

examined using an ABAB design and demonstrated during the choice question condition that there was an overall decrease in challenging behavior. The results from this study signify and extend previous findings suggesting sometimes simple, nonaversive choice interventions can be effective at reducing challenging behavior and increasing appropriate behaviors (Cannella et al., 2005; Hanley et al., 2006; Lancioni, O‘Reilly, & Emerson, 1996). The opportunity to choose has also been examined as an independent variable. Tiger et al. (2006) examined choice as a reinforcer with six preschoolers, both with and without disabilities. The authors appear to have demonstrated a high degree of control over choice as an independent variable. The opportunity to choose was compared to a nochoice condition, the number of items from which to choose was evaluated, and a progressively increasing response requirement to access the opportunity to choose was used to identify the specific values of choice. This study found five participants preferred the opportunity to choose and two of the participants‘ data indicated choice alone could serve as a reinforcer. Another example of an effective choice intervention includes the study by O‘Reilly et al. (2004). A paired-choice assessment protocol was implemented with a fiveyear-old girl with severe intellectual disabilities to verify the reinforcing value of stimuli and activities maintaining the child‘s poor sleeping habits. After discovering the activity (i.e., being with her mother) maintaining the participant‘s behavior of leaving her room and staying awake, a fixed-time delivery of attention was used as an intervention. Data reported in a BCBC reversal design indicated a significant reduction in sleep disturbance.

2

Additional follow-up observations reported the reduction in the participants‘ sleep disturbances maintained for at least twelve months. Choice interventions can be a parsimonious intervention for relatively disruptive behaviors in the classroom. Carlson et al. (2008) used a choice-making intervention to eliminate public disrobing and urinating in clothing for two children with developmental disabilities in a school setting. The participants were given the choice to change into new, high preferred clothing during a scheduled period during the day. Data were examined and displayed in a multiple baseline across participants design demonstrating the choicemaking intervention was effective at decreasing and eliminating public disrobing and urinating in clothing with both of the participants. The authors concluded the choicemaking intervention functioned as an abolishing operation, decreasing the participants‘ motivation to disrobe and urinate. The effectiveness of various preference assessment formats and components The use of reinforcement is vital when acquiring, maintaining, and eliminating operant behaviors (Pace et al., 1985). Identifying preferences of individuals with severe intellectual disabilities can be challenging. Researchers have recently examined the methodological rigor of preference assessment in the literature. A plethora of recent research examines specific components of preference assessment to increase the efficacy and efficiency of identifying reinforcers (Ciccone et al., 2005; Carr, Nicolson, & Higbee, 2000; Ciccone, Graff, & Ahearn, 2006; Daly, Well, Swanger-Gagne, Carr, Kunz, & Taylor, 2009; DeLeon & Iwata, 1996; DeLeon, Iwata, & Roscoe, 1997; Fisher et al., 1992; Glover, Roane, Kadey, & Grow, 2008; Graff & Gibson, 2003; Hanley, Iwata, &

3

Roscoe, 2006; Horrock & Morgan, 2009; Pace et al., 1985; Piazza et al., 1996; Windsor et al., 1994). The need to discover effective, systematic ways to identify preferences of individuals with severe to profound developmental disabilities was identified by an important publication by Pace et al. (1985). A significant advancement in the field of developmental disabilities and behavior analysis was the first formal preference assessment (i.e., the single-stimulus approach method) developed by Pace et al. (1985). In this study, sixteen stimuli, both edibles and tangibles, were individually presented to each participant ten times. Approach behaviors were measured in an attempt to differentiate preferred and nonpreferred stimuli. One significant limitation to this format was that some of the participants consistently approached most or all of the stimuli on each trial (or presentation). This made it difficult to conclude which stimuli were preferred versus nonpreferred. Addressing the limitation posed by Pace et al. (1985), Fisher et al. (1992) developed the paired-stimulus (PS) preference assessment format. Before this, the singlestimulus assessment (SS) was the most common method used to identify reinforcers for individuals with severe to profound disabilities (Fisher et al., 1992). Fisher and colleagues used a concurrent operants paradigm as an extension of the SS to determine whether the forced-choice method (also known as paired-stimulus) of stimulus presentation can better identify preferred and nonpreferred stimuli. Three participants were presented sixteen stimuli individually during phase one and stimuli were presented in pairs during phase two. Overall data suggested the PS presentation had good

4

concurrent validity and the SS presentation may tend to have false negatives when identifying high-preference stimuli. To further investigate preference assessment formats, Windsor et al.,(1994) compared paired- and multiple-stimulus preference assessment formats. The two formats were also compared to staff reports. Six stimuli, all edibles, were available in pairs during the PS assessment and available for selection during all trials (50 trial total) for the multiple-stimulus condition. Results indicated the PS produced results more consistently across sessions and a more distinct ranking of the stimuli, but required more administration time. Staff reports were not significantly correlated with either format, but the authors reported the items identified as preferred were similar. DeLeon and Iwata (1996) examined two different formats of multiple-stimulus presentations and compared those methods to the PS presentation. Edible and tangible stimuli were used with seven adults with profound developmental disabilities. The first multiple-stimulus presentation was the multiple stimulus without replacement (MSWO). In this format, multiple stimuli (seven) were placed in an array. When an item was selected, it was not returned to the array of stimuli. The next stimulus presentation was the multiple stimulus with replacement (MSW). This format was similar to the MSWO except the selected item was returned to the array. Data revealed all three preference assessment formats produced similar results in identifying the high-preference stimuli. The rankings produced by the MSWO and PS procedures were more consistent across administrations when compared to the MSW. In addition, the MSWO took substantially less time to administer than the PS (i.e., the MSWO took 15.5 minutes whereas the PS

5

took 53.3 minutes). Effectively identifying reinforcing stimuli in a relatively short amount of time makes the MSWO a more practical choice for use in applied settings. Additional research has confirmed and extended these findings (Carr et al., 2000; Daly et al., 2009; Paramore & Higbee, 2005). Various components of preference assessments have been examined and researched. For example, preference assessments have been researched with a variety of populations, such as individuals with mild to severe developmental disabilities (e.g., Ahearn, Clark, Debar, & Florinino, 2005; Hanley, Iwata, & Roscoe, 2006), individuals with emotional and/or behavioral disorders (e.g., Daly et al., 2009; Taravella, Lerman, Contrucci, & Roane, 2000; Paramore & Higbee, 2005), physical and medical disabilities (e.g., LeBlanc et al., 2006), in educational settings (Hanley, Cammilleri, Tiger, & Ingvarsson, 2007) and with general education students (Resestar & Noell, 2008; Schanding, Tingstrom, & Sterling-Turner, 2008). Each study cited above reported positive findings except for LeBlanc et al. (2006) and Resestar and Noell (2008), who reported mixed results. Another broad component of preference assessments is choosing a selection- or duration-based procedure. Kodak, Fisher, Kelly, & Kisamore (2009) compared the effects of selection- versus duration-based preference assessment procedures with four participants who were two to ten-years-of age during the time of the study. The participants were diagnosed with developmental disabilities and/or autism. A multiplestimulus with replacement preference assessment (MSW; DeLeon & Iwata, 1996) was used as the selection-based procedure. A free-operant assessment (FO; Roane, Vollmer,

6

Ringdahl, & Marcus, 1998) was used as the duration-based preference assessment. A reinforcer assessment was conducted, and a reversal design was used to evaluate the data. The results indicated the MSW and the FO identified different items as preferred. For two of the participants, the MSW was able to identify at least one of the items most preferred. Data for the other two participants were undifferentiated. The authors recommended evaluating the circumstances under which the preferred items would be used. For example, because selection-based procedures typically allow access to stimuli for a brief amount of time, it may be useful to use a selection-based preference assessment in the classroom where reinforcement is typically brief. However, for circumstances like a long car ride where access to reinforcement may be relatively longer in duration, a freeoperant assessment may be more appropriate. Many researchers have evaluated what types of stimuli can be identified as reinforcing using preference assessment procedures. For example, Graff et al. (2006) examined the use of pictures versus tangibles during a PS preference assessment with four participants. Two of the participants were diagnosed with autism, one was diagnosed with a chromosomal disorder, and another was diagnosed with attention deficit hyperactivity disorder. The pictorial PS procedures were identical to the tangible PS procedures except the pictorial assessment consisted of pictures of the stimuli rather than the actual stimuli. Data were collected using an alternating treatments design embedded within a reversal design. The results of the preference assessments revealed both assessments identified similar hierarchies of the stimuli. The data from the reinforcer

7

assessments demonstrated high-preference stimuli were effective reinforcers as predicted by both assessments. Activities have also been evaluated as stimuli during preference assessments. Daly et al. (2009) implemented MSWO preference assessments to identify preferred activities that function as reinforcers in the classroom with individuals with behavioral disorders. The name of one activity was written in the center of a card and presented in the array with eight activities total. The effects of high-, moderate-, and low-preference activities and doing nothing were evaluated using an alternating treatments design. Regardless of the activities available, all participants demonstrated an increase in performance when reinforcing contingencies were introduced when compared to baseline. The authors suggest the MSWO preference assessment format may effectively identify reinforcing activities in a classroom setting. Wilder, Schadler, Higbee, Haymes, Bajagic, & Register (2008) evaluated the use of PS preference assessments to identify reinforcing olfactory stimuli with three individuals with mental retardation and autism. Examples of stimuli included citrus, apple, lavender, and vanilla. A reinforccer assessment was conducted and data were examined using a reversal with a multi-element design. The authors reported positive findings and responding appeared to increase for all of the participants. The olfactory stimulus identified as high-preference by the preference assessment demonstrated to be the most effective when compared to the moderate- and low-preference olfactory stimuli. Results from the PS corresponded with the data from the reinforcer assessment,

8

suggesting it may be effective at identifying reinforcing olfactory stimuli for children with autism. Additional stimuli evaluated using preference assessments included work tasks. Reid, Parsons, Towery, Lattimore, Green, & Brackett (2007) compared the efficiency and accuracy of PS procedures, MSWO procedures, and staff reports to identify work preferences for twelve participants ranging in age from 29-76 years-old with severe to profound cognitive disabilities. The results of this study are mixed: the staff reports were the most efficient but the least accurate. The MSWO was more efficient when compared to the PS but the PS was more accurate. Overall, results suggested preference-assessment protocol is a useful means to identify work-task preferences with adults with severe disabilities in a supported work environment. The way stimuli are scored and ultimately ranked has also been investigated. Ciccone et al. (2005) evaluated the accuracy of an alternative scoring method for the MSWO preference assessment. The traditional way of scoring stimuli from the MSWO was to measure the percentage a stimulus was approached or selected (Fisher et al., 1992). The authors state this scoring method does not consider the unequal number of opportunities to select each stimulus within and across sessions during the MSWO. The alternative scoring method, a point weighting method system, was developed to enhance reinforcer identification so data for each stimulus was analyzed on a per session basis to identify on which trial (e.g., first, second, third, etc.) selection occurred. The authors suggested the traditional method of scoring stimuli for the MSWO may reject stimuli serving as functional reinforcers and the point weighting method was able to identify

9

more reinforcing stimuli. Thus, the authors conclude the point weighting method is a superior method for the MSWO and allows for a preference hierarchy to which preferred and nonpreferred items are identified more accurately. Research examining high-, moderate-, and low-preference stimuli as reinforcers In applied settings, providing frequent reinforcement is central to learning and maintaining appropriate behaviors (Pace et al., 1985). Providing frequent reinforcement, however, can lead to satiation and therefore act as an abolishing operation in reducing the effectiveness of the reinforcing stimuli (Ciccone et al., 2005). Hence, having a number of items identified as reinforcing can prevent satiation from occurring. Research has examined the effects of low-, moderate-, and high-preference stimuli as identified by various preference assessment formats (Carr et al., 2000; Ciccone et al., 2006; Daly et al., 2009; Paramore & Higbee, 2005; Piazza et al., 1996; Roscoe, Iwata, Kahng, 1999; Taravella et al., 2000; Wilder et al., 2008). For example, Carr et al. (2000) conducted brief MSWO preference assessments with three participants, ages 2, 6, and 7, diagnosed with autism in a therapy room. Edible and tangible stimuli were assessed. The authors examined the stability of preference over time and low-, moderate-, and high-preference stimuli were evaluated using a low-frequency target behavior selected for each participant. For two of the participants, the low-preference stimulus had no significant reinforcing effect on responding when compared to baseline levels. The moderatepreference stimulus produced moderate effects on responding for one participant, only modest for another, and no reinforcing effect for one of three participants. The high-

10

preference stimulus produced the highest rates of responding when compared to the moderate- and low-preference stimuli and baseline levels for all three participants. A study replicating and extending the results from Carr et al. (2000) with adolescents was conducted by Paramore & Higbee (2005). The authors compared the reinforcing effects of low-, moderate-, and high-preference stimuli identified using a MSWO preference assessment with participants diagnosed with emotional-behavioral disorders. Only edible stimuli were used in this study and an alternating treatments design was implemented. Reinforcement was provided contingent upon on-task behavior in the classroom. Data were initially somewhat undifferentiated for all three participants and the authors suggested this was due to the reinforcement phase producing a general increase in on-task behavior as the contingency was simply added. For all participants, though, the high-preference stimulus later produced the highest percentage of on-task behavior. The reinforcing effect of the high-preference stimulus persisted over time. An interesting consideration when evaluating low-, moderate-, and highpreference stimuli as reinforcers is to examine the stimuli being presented (e.g., edibles versus tangibles). Taravella et al. (2000) conducted two preference assessments and a reinforcer assessment with two participants with mental retardation to evaluate the efficacy of low-preference stimuli. A SS preference assessment was conducted using ten edible and tangible stimuli. Items approached on at least 80% of trials were included in a PS preference assessment. The five lowest ranked stimuli from the PS were reassessed using similar procedures. The highest ranked item from the latter preference assessment was provided during the reinforcer assessment contingent upon correct responding. A

11

reversal design was implemented and data suggested access to reinforcement substantially increased responding. The authors noted when less preferred tangible items appeared with highly preferred tangibles, the less preferred items did not appear to be potent reinforcers. When the less preferred (or low ranked items) were reassessed with other less preferred items, at least one of the items was approached 80% or more of trials. Results suggest low-preference stimuli may function as reinforcers, which is similar to findings by Roscoe et al. (1999). The study conducted by Daly et al. (2009) was previously discussed for its use of cards during preference assessments. This study also evaluated the efficacy of high-, moderate-, and low-preference stimuli as reinforcers and examined if the MSWO format could identify reinforcing activities and privileges to increase math problem completion. High-, moderate-, and low-preference activities were presented in an alternating treatments design. The results suggested the frequency of math problem completion increased relative to baseline when reinforcement contingencies (i.e., high-, medium-, and low-preference stimuli) were available. The authors noted low-preference stimuli can be potent reinforcers in the absence of more reinforcing stimuli. This finding suggests low-preference stimuli may function as effective reinforcers under specific circumstances and extends findings from Taravella et al. (2000). Ciccone et al. (2006) sought to extend the findings of Taravella et al. (2000) and Piazza et al. (1997) and evaluated if moderate-preference items were more likely to function as reinforcers when compared to low-preference items. In this study, seven participants ranging in age from 15 to 21 who were diagnosed with autism and/or mental

12

retardation were exposed to three MSWO preference assessments using only edible stimuli. The stimuli were ranked using the point weighting method developed by Ciccone et al. (2005). The items identified as moderate-preference were reassessed using similar procedures but adding additional stimuli as necessary to complete the array of seven items. The items identified as low-preference were reassessed and additional stimuli were added to this array as well. Results indicated items initially categorized as moderatepreference were scored as high-preference during the moderate item assessment. Only one of the nineteen stimuli initially categorized as low-preference was later categorized as high-preference. The authors suggested high-preference stimuli can mask other highpreference stimuli during MSWO preference assessments. No reinforcer assessment was conducted to verify the reinforcing effectiveness of the stimuli. Effectively and efficiently identifying multiple reinforcers for students, especially those with severe developmental disabilities can be a central component for managing classroom behavior. Having multiple reinforcers can help guard against satiation (Ciccone et al., 2006). The current study sought to extend the findings of Ciccone et al. (2006), Daly et al. (2009) and Taravella et al. (2000). Both edible and tangible stimuli were assessed as requested by the classroom teacher. Using edible and tangible stimuli allowed the opportunity to extend Taravella et al. (2000) findings noting tangible stimuli can be displaced when edible stimuli are available during MSWO preference assessments. The point weighting method developed by Ciccone et al. (2005) was used to categorize the stimuli. Moderate- and low-preference stimuli were reassessed with some procedures differing from those used by Ciccone et al. (2006). Rather than adding

13

additional stimuli (i.e., from other participants‘ array), nine instead of seven stimuli were used in the initial MSWO preference assessments. This was to address the limitation of adding confounding stimuli without a known preference. Nine initial preference assessments rather than one were conducted before developing a hierarchy to allow for more accurate scoring. Finally, a reinforcer assessment was conducted to verify the results of the preference assessments.

14

Chapter 2: Method This chapter presents a description of the participants, the settings and materials, dependent variables and data collection, interobserver agreements, procedural integrity, experimental and general procedures, experimental design, and social validity. Participants Three students from a class of seven diagnosed with mental retardation and/or autism participated in this study. Participants were between eight and eleven years-of-age at the time of the study. The experimenter probed to see if the participants understood the instruction ‘pick one’ by making the request prior to the beginning of the assessment. Students able to follow this instruction at least two times prior to participation in the preference assessments without training qualified for inclusion in this study. All of the participants chosen for this study were able to follow single-step instructions, had communication deficits, and attended a school for individuals with developmental disabilities. The students were using a community-based instructional curriculum, which taught daily self-help skills and communication skills (e.g., signing, pressing a switch). Two of the participants frequently engaged in stereotypic behavior (e.g., repetitive face touching, swinging of arms). All of participants were ambulatory and two were completely nonverbal. All seven students were potential participants, but one participant, who was cortically blind, was excluded from this study because he was unable to see the stimuli during the preference assessments. 15

Levi was a white male diagnosed with mental retardation and multiple disabilities (other than deaf and blind) and was 11.5-years-old at the start of the study. According to his school records, he scored in the moderate deficit range on the Vineland Adaptive Behavior Scale (VABS; Sparrow, Balla, Cicchetti, 1984). Levi‘s adaptive behavior composite score was a 47 (moderate deficit). For the communication domain, Levi‘s mean score was a 49 (moderate deficit). For the daily living skills domain, Levi‘s mean score was a 44 (moderate deficit).

Table 2.1. Participants‘ ages and test scores Levi

Alvin

Jake

Age at start of study

11.5 years

10 years

8 years

Adaptive Behavior Composite

47 (moderate deficit)

32 (severe deficit)

33 (severe deficit)

Communication Domain

49 (moderate-deficit)

36 (moderatesevere deficit)

33 (severe deficit)

Daily Living Skills Domain

44 (moderate deficit)

33 (severe deficit)

36 (moderatesevere deficit) Note. Adaptive Behavior Composite, Communication Domain, and Daily Living Skills Domain are scores taken from the Vineland Adaptive Behavior Scales II pulled from school records.

Alvin was a white male diagnosed with mental retardation and autism and was 10-years-old at the start of the study. According to his school records, Alvin scored in the 16

severe deficit range on the VABS. Alvin‘s adaptive behavior composite score was a 32 (severe deficit). For the communication domain, Alvin‘s mean score was a 36 (moderatesevere deficit). For the daily living skills domain, his mean score was a 33 (severe deficit). Jake was a white male diagnosed with mental retardation and autism and was 8years-old at the start of the study. According to his school record, Jake scored in the severe deficit range on the VABS. His adaptive behavior composite score was a 33 (severe deficit). For the communication domain, Jake‘s mean score was a 33 (severe deficit). For the daily living skills domain, his mean score was a 36 (moderate-severe deficit). During the time of the study, Jake had a behavior plan in place for aggressive behavior maintained by access to tangible items. Settings and Materials Preference assessments were conducted in an empty room in the participants‘ school. The room contained a table, two chairs, and the stimuli needed for the study. Reinforcer assessments took place in the participants‘ classroom. The classroom was 4.57 by 6.10 meters, with two tables lined in a row, ten chairs, a desk in the corner secluded from the class, the teacher‘s desk, and a variety of materials used in the classroom (e.g., art bins, book shelves). Materials needed for the study included the stimuli specific to each participant, such as a puzzle, Craisins, mirror, ice, bubbles, M&Ms, Playdoh, cardboard chest, rubbery fish, windmill, string cheese, string, juice, spinner, tambourine, wand, chips, toy truck, picture collage, plastic microphone, chocolate chip morsels, and a spin top.

17

Materials needed for the reinforcer assessment included 17.78 by 12.7 cm index cards, scissors, placemat with alphabet letters, green tape, and a wet cloth. Other materials needed in the study included a white laminated placemat, a red corduroy placemat, a stone-textured placemat, a cafeteria tray, and data observation sheets. Dependent Variables and Data Collection Preference Assessments. Several different variables were measured, including Approach, Selection, and Not Selected. For all preference assessments, Approach was defined as a student reaching for an item, but not making any physical contact with that item. A nonoccurrence of Approach included the participant making contact with an item or the participant reaching for something other than the stimuli presented on the table. Selection was defined as a student making physical contact with a tangible item or consuming an edible. Although Approaches were recorded, only Selected items were scored. A nonoccurrence of selection included the participant making any physical contact with an edible but not consuming it and/or the participant touching or grabbing two or more stimuli simultaneously. If the nonoccurrences took place or the participant touched or grabbed an item before the experimenter instructed him/her to do so, the item(s) were not counted and the experimenter waited 5 s and gave the instruction ‘pick one’ again. If the participant did not make a selection within 10 s from the start of the trial, the remaining items were recorded as Not Selected. The scoring method similar to the one described by Ciccone et al. (2005) was used, but of seven preference assessment sessions, nine were conducted so that additional

18

stimuli did not need to be added to the MWSO array when reassessing moderate- and low-preference stimuli. The point weighting system was used to analyze data on each stimulus per session and to identify during which trial the selection occurred. The item chosen during Trial 1 (i.e., the item chosen first in a session) resulted in a point value score of 9. Selection on Trial 2 = 8 points; Trial 3 = 7 points, Trial 4 = 6 points; Trial 5 = 5 points; Trial 6 = 4 points; Trial 7 = 3 points; Trial 8 = 2 points; Trial 9 = 1 point. If any item was not selected, a point value of 0 was assigned to that stimulus. Next, a trial position score was assigned. This was done by tallying the number of times an item was selected during a specific trial (e.g., first, second, etc.) and multiplying that number by the point value assigned to that trial (e.g., first trial‘s assigned point value was nine, second trial‘s point value was eight). The trial position scores for each stimulus from all nine preference assessment sessions were then added to calculate the total score. If a stimulus was not selected in one or more sessions, the sessions during which the stimulus was selected were added. For example, when a stimulus was chosen two times during the first trial (selected twice on the first trial (point value of 9), for a trial position score of 18), and five times on the third trial (selected five times on the third trial (point value of 7), for a trial position score of 35), the total score was 53 (18 plus 35). High-preference items were those that scored 81-64 points, moderate-preference 63-40 points, and lowpreference 39-0 points (Ciccone et al., 2005). If there were not at least three items in the moderate or low category, the item that scored nearest the criterion was moved to that category.

19

Reinforcer Assessments. The single-step tasks that the students were asked to complete were scored as Correct or Incorrect, in all conditions. Each session in all conditions consisted of five trials. Correct was scored if the participant completed the task accurately. Incorrect was scored if the participants did not complete the task or did so inaccurately. If approximations of the task occurred, they were scored as Incorrect. All incorrect responses resulted in hand-over-hand prompting to aid the participant in completing the task correctly. Procedures to Ensure Accuracy and/or Believability of Data Observer Training. Observers were graduate students from The Ohio State University Special Education program. They were given the definition of the dependent variables, including examples of occurrences and nonoccurrences. Observers were trained to identify the stimuli for each participant. Additionally, observers were trained on the criteria used to score trials during the reinforcer assessments. Experimenter The experimenter was a Master of Arts student in the Special Education and Applied Behavior Analysis program at The Ohio State University. The student holds a Bachelor of Arts degree in psychology from The Ohio State University. She has spent the last three years working with children with autism at home, school, and in the community. Additionally, she has served as a consultant for parents and teachers, trained

20

others in applied behavior analysis therapy, and worked on a behavior analysis team in a school for students with developmental disabilities. Interobserver Agreement A second independent observer recorded Approach, Selection, Non Selection, Correct, and Incorrect individualized to each participant. The second observers were graduate students from The Ohio State University Special Education and Applied Behavior Analysis. Interobserver agreement (IOA) was calculated at least twenty percent of sessions, across all conditions (i.e., preference assessments and reinforcer assessments) and participants. IOA was calculated by dividing the smaller number of occurrences by the larger number of occurrences. This number was then multiplied by 100% to yield the percentage of IOA (Cooper, Heron, and Heward, 2007). Procedural Integrity The degree to which the experimenter followed the procedures outlined was assessed by a second independent observer across all conditions and at least twenty percent of sessions in each condition. The second observer had a checklist of each procedural step the experimenter was to follow in each condition and simply circled ‗yes,‘ ‗no,‘ or ‗not applicable.‘ The percentage of procedural integrity was determined by dividing the number of steps followed by the experimenter by the total number of steps, then multiplying by 100%.

21

Experimental Procedures Selection of Stimuli. The stimuli selected for the first phase of preference assessments were based on a structured interview with the participants‘ caregiver (i.e., the person identified as assuming care and supervision of the participant throughout the day). The Reinforcer Assessment for Individuals with Severe Disabilities (RAISD) developed and described by Fisher et al. (1996); see Appendix C. was used to determine preliminary potentially preferred items. The experimenter also discussed at length with the teacher the participants‘ preferences. Structured observations of the participants during leisure time were also used in selecting stimuli. The stimuli chosen were items the students had been previously exposed to. Preference Assessments. A maximum of two preference assessments were conducted during one day. Participants were given a break (i.e., three to five minutes) before beginning another assessment. Prior to the start of the preference assessment, each participant was exposed to all the stimuli that were used in the array for 10 s or until the edible item was consumed. Participants were exposed to three separate, but procedurally identical sets of Multiple Stimulus Without Replacement (MSWO) preference assessments. Procedures for the MWSO preference assessments were similar to those described by DeLeon and Iwata (1996), using both edible and tangible stimuli. Each session began with nine stimuli (e.g., M&M, windmill, wand, chip) placed in a linear array, approximately 5 cm

22

apart and 30 m in front of the participant. The experimenter sat across the table from the participant and instructed the participant to select an item by saying, ‘pick one.‘ Phase I. During phase one of the preference assessments, when a participant did not respond within 10 s of the instruction ‘pick one,’ the instruction was repeated. The instruction was given up to two times before the session was terminated. After a selection was made, the participant was allowed to consume the edible item or given 10 s access to the tangible item before it was removed from the array. When nonoccurrences of the dependent variable took place, the participant was asked again to make a selection by the experimenter stating ‘pick one’ 5 s after the nonoccurrences took place. After selection, the experimenter repositioned the stimuli by taking the item on the far left of the array and moving it to the right end and repositioning the other stimuli so they are equally spaced in front of the participant. This process was continued until all nine items were selected or the session was terminated due to the participant not making a choice. The original nine stimuli were assessed in an additional eight sessions to ensure reliability of selections and to allow for the application of the point weighting method described by Ciccone et al. (2005). Phase II- Moderate-Preference The stimuli categorized as moderate-preference according to the Ciccone et al. (2005) point weighting scoring system from phase one were reassessed using similar procedures. No additional stimuli were added to the array. The stimuli used (a minimum

23

of three were required) were positioned, spaced, and rotated using the same procedures as the first preference assessment. Phase III- Low-Preference Preference assessments included the stimuli that were categorized as lowpreference in the first preference assessments, using the same procedures used in phase two. Reinforcer Assessments. After completing all three sets of preference assessments, the highest ranked item from the first preference assessment (high-preference stimuli), second preference assessment (moderate-preference stimuli), and third preference assessment (lowpreference stimuli) were used as reinforcement for the reinforcer assessment. Levi‘s task was cutting across a 15.24 cm by 10.16 cm index card. Alvin‘s task was matching each letter of his name to a placemat containing the letters of his name. Jake‘s task was wiping a 24.5 cm by 25.4 cm space on a table. Levi and Alvin‘s tasks were recommended for use by the classroom teacher and Jake‘s task was taken from his individualized education programs (IEP). Experimental Design The effects of the students following directions contingent on high-, moderate- and low-preference stimuli were evaluated using an alternating treatments design (Cooper et al., 2007). This design entails the rapid alternation of two or more independent variables while the effects on the dependent variables are measured. The three independent variables, or treatments, in this study were the high, moderate, and low preference

24

stimuli. The dependent variable, or the target behavior, was the completion of the singlestep task that was predetermined through a teacher interview and/or a goal in the students‘ IEP. When data in baseline were stable, the intervention was implemented by randomly counterbalancing the treatments from session to session. General Procedures With the exception of the baseline condition, after the completion of the preference assessments the participants were shown the high-, moderate-, or lowpreference item and told the item they would receive if they completed the task. In all conditions, the participants were asked to complete the predetermined tasks by receiving a predetermined verbal discriminative stimulus (Sᴰ), such as ―cut the paper‖ (Sᴰs remained constant throughout all conditions). To indicate the experimental condition in which the participants were engaged in, different textured and colored placemats were used. A white laminated placemat was used during baseline, a brown cafeteria lunch tray was used when a high-preference item was used as reinforcement, a stone textured placemat was used when a moderate-preference item was used as reinforcement, and a red corduroy placemat was used when a low-preference item was used as reinforcement. Each reinforcer session contained five discrete trials that were scored as correct or incorrect. The experimenter waited 10 s for the participant to complete the task. If the participant initiated the task within 10 s but did not complete it, the participant was stopped and the trial was scored at incorrect. If the participant did not attempt to initiate or complete the task within 10 s, the trial was scored at incorrect. Accurate completion of the task within 10 s was scored as correct.

25

Baseline Throughout this condition, a white laminated placemat was used. Baseline data were taken on the participants‘ ability to complete the single-step task with no reinforcement contingencies in place. The experimenter stated: ‘Hi (participant’s name). Thanks for working with me today,’ and then gave the Sᴰ for the task. For example, the experimenter may state ‘Hi Alex. Thanks for working with me today. Cut the paper.‘ When baseline data were stable, the intervention was implemented immediately. High-Preference Item Throughout this condition, a brown cafeteria lunch tray was used. The experimenter began by stating: ‘Hi (student’s name). Thanks for working with me today. You will be earning (the experimenter stated the high-preference item and show the item to the participant) if you do your work,’ and then gave the Sᴰ for the task. For example, ‘Hi Alex. Thanks for working with me today. You will be earning a cracker (experimenter showed item to the participant) if you do your work. Cut the paper.’ If the participant completed the task, he was immediately given 10 s access or allowed to consume the high preference item. Access to reinforcement was contingent upon correct completion of the single-step task. If the participant did not comply within 10 s, a hand-over-hand prompt was provided to aid him in completing the task. Moderate-Preference Item A stone textured placemat was used in this condition. The experimenter began by stating: ‘Hi Alex. Thanks for working with me today. You will be earning (the experimenter stated the moderate-preference item and showed the item to the participant)

26

if you do your work,’ and then gave the Sᴰ for the task. For example, ‘Hi Alex. Thanks for working with me today. You will be earning a cracker (experimenter shows item to the participant) if you do your work. Cut the paper.’ If the participant completed the task, they were immediately given 10 s access or allowed to consume the moderate-preference item. Access to reinforcement was contingent upon correct completion of the single-step task. If they did not comply within 10 s, a hand-over- hand prompt was provided to aid the participant in completing the task. Low-Preference Item A red corduroy placemat was used to in this condition. The experimenter began by stating: ‘Hi Alex. Thanks for working with me today. You will be earning (the experimenter stated the low-preference item and showed the item to the participant) if you do your work,’ and then gave the Sᴰ for the task. For example, ‘Hi Alex. Thanks for working with me today. You will be earning a book (experimenter shows item to the participant) if you do your work. Cut the paper.’ If the participant completed the task, they were immediately given 10 s access or allowed to consume the low-preference item. Access to reinforcers was contingent upon completion of the single-step task. If they did not comply within 10 s, a hand-over-hand prompt was provided to aid the participant in completing the task. The low-preference item was counterbalanced randomly from session to session with the high- and moderate-preference items. Social Validity The effectiveness of the intervention, satisfaction with procedures, and likelihood of using preference assessments was assessed with one of the consumers, the

27

participants‘ teacher. Specifically, the teacher filled out a survey (see Appendix I) after the conclusion of the study. Five-point Likert scales and an opportunity to voice general comments were employed.

28

Chapter 3: Results This chapter presents results from the preference assessments, experimental data, measures of interobserver agreement, measures of procedural integrity, and social validity assessments. Procedural Integrity Procedural Integrity of Preference Assessments Data on the procedural integrity of preference assessments are displayed in Table 3.1. During Phase I of the preference assessments for Levi, 22% of sessions were measured, during Phase II and Phase III, 100% of sessions were measured, and overall, 36% of all sessions were measured for Levi and procedural integrity was calculated to be 100%. For Alvin in Phase I, 22% of sessions were measured, during Phase II and Phase III, 100% of sessions were measured, and overall, 36% of all sessions were measured and procedural integrity was calculated to be 100%. During Phase I for Jake, 33% of sessions were measured and procedural integrity was 99% (range, 99-100%). During Phase II and Phase III, 100% of sessions were measured and procedural integrity was 100%. Overall, 45% of sessions were measured and procedural integrity was calculated to be 99.67% (range, 99-100%).

29

Table 3.1 Mean Procedural Integrity during the preference assessments Preference Assessments: Phase I

Preference Assessments: Phase II

Preference Assessments: Phase III

Levi

100 (2)ª (22)ᵇ

100 (1) (100)

Alvin

100 (2) (22)

100 (1) (100)

100 (1) (100)

Jake

99 (3) (33) (99-100)ᶜ

100 (1) (100)

100 (1) (100)

100 (1) (100)

Note. Phase I contained nine complete MSWO preference assessments. Phase II and Phase III refer to the reassessment of the stimuli chosen in Phase I.

ª Number of procedural integrity sessions ᵇ Percent of sessions with procedural integrity ᶜ Range

Interobserver Agreement of Preference Assessments For all participants, 39% of preference assessments were included for interobserver agreement (IOA). The exact number of sessions and percentage of sessions containing IOA for each participant in each condition of the preference assessments is displayed in Table 3.2. For Levi, IOA was measured in 22% of sessions during Phase I and calculated to be 100%. During Phase II and Phase III, IOA was included and measured in 100% of sessions and IOA was 100%. Overall mean IOA for Levi was 100% in 36% of sessions. For Phase I for Alvin, IOA was measured in 22% of sessions and calculated to be 100%. In Phase II and Phase III, IOA was measured during 100% of

30

sessions and calculated to be 100%. Overall, IOA in 36% of sessions was 100%. For Jake in Phase I, IOA was measured in 33% of sessions and calculated to be 100%.

Table 3.2 Mean IOA for percent correct during the preference assessments Preference Assessments: Phase I

Preference Assessments: Phase II

Preference Assessments: Phase III

Levi

100 (2)ª (22)ᵇ

100 (1) (100)

100 (1) (100)

Alvin

100 (2) (22)

100 (1) (100)

100 (1) (100)

Jake

99 (3) (33)

100 (1) (100)

100 (1) (100)

Note. Phase I contained nine complete MSWO preference assessments. Phase II and Phase III refer to the reassessment of stimuli chosen in Phase I.

ª Number of IOA sessions ᵇ Percent of sessions with IOA

ᶜ Range

During Phase II and Phase III, IOA was measured in 100% of sessions and calculated to be 100%. Overall, the mean IOA for Jake was 100% in 45% of sessions. Preference Assessments The data displaying the stimuli used during the preference assessments and the rank in which they were chosen for Levi, Alvin, and Jake are displayed in Figure 3.1.

31

Levi Stimuli included in Luke‘s preference assessments included a rubber toy with liquid and a plastic fish inside (fish); a puzzle that contained a variety of locks and doors; Craisins; plastic mirror; cup of ice; bubbles; yellow M&M candies; green Playdoh; and a cardboard chest that opened and closed(chest). During Phase I of the preference assessments, Levi most often chose edible items (e.g., Craisins, M&M, ice) first from the array. Out of the nine full array MSWO preference assessments, the fish and the chest were not selected during two of the sessions and the sessions were terminated prematurely (see method section for detail). The mirror and the Playdoh were not selected during three of the nine sessions. The bubbles were not selected during four of the nine sessions. The puzzle was not chosen during one of the nine sessions. Each edible item (i.e., the Craisin, ice, and M&M) was selected during all nine of the preference assessments in Phase I. Using the point weighting scoring method, the order from most preferred (i.e., received a higher point weighting score) to the least preferred was documented as: Craisin (75 points), M&M (74 points), ice (52 points), puzzle (45 points), chest (31 points), bubbles (26 points), fish (22 points), Playdoh (17 points), and mirror (16 points). The only documented approach was for the mirror, during the fifth assessment. For Phase II of the preference assessments, the stimuli categorized as moderate-preference were reassessed using the same procedures as in Phase I of the preference assessments. For Levi, only two items (ice and puzzle) were categorized as moderate- preference (i.e., point weighting score between 40 and 63). In order to conduct a MWSO

32

Rank

9 8 7 6 5 4 3 2 1 0

Rank

Levi 9 8 7 6 5 4 3 2 1 0

Rank

Alvin

9 8 7 6 5 4 3 2 1 0

Jake Figure 3.1. The stimuli presented during the first phase of the preference assessments are displayed across the x-axis for each participant. Items selected first are represented by the one on the y-axis and the items selected last are represented by the nine on the y-axis.

33

preference assessment, the array had to contain at least three stimuli. For this reason, the stimuli with the next highest point weighting score, the chest (31 points), was moved into the moderate-preference category and was reassessed in Phase II of the preference assessments. Upon reassessment, Levi selected the ice during the first trial of the only session in this phase. Consequently, the ice was used as the moderate-preference stimulus throughout the reinforcer assessment. For Phase III of the preference assessments, the stimuli categorized as lowpreference (i.e., point weighting score between 0 and 39) were reassessed using the same procedures as in Phase I and Phase II. Stimuli included the bubbles, Playdoh, the fish, and the mirror. When the stimuli were reassessed, Levi selected the bubbles during the first trial of the only session in this phase. The bubbles were used as the low-preference stimulus throughout the reinforcer assessment. Alvin Stimuli included in Alvin‘s preference assessments included a yellow M&M candy; bubbles; a spinner that lit up and spun when a button was pressed; a toy truck, a picture collage containing pictures of Alvin; his classmates and teachers; a plastic microphone; a chocolate chip morsel; and a small piece of tortilla chip. During Phase I of the preference assessments, similar to Jake, Alvin most often chose the edible items (e.g., M&M, chip, and chocolate chip) first from the array. As displayed in Figure 3.1, edible items were usually chosen during the first trials of the preference assessments. The picture collage and top were not selected during two of the nine sessions. The truck was not selected during three of the sessions. The spinner was not selected during four

34

sessions and the bubbles were not selected in eight of the nine sessions (i.e., the bubbles were chosen only once). Using the point weighting scoring method, the order from most preferred (i.e., received a higher point weighting score) to the least preferred was documented as: chip (67 points), chocolate chip (66 points), M&M (63 points), microphone (51 points), picture collage (38 points), top (37 points), truck (29 points), spinner (19 points), and the bubbles (6 points). There were no documented approaches. During Phase II of the preference assessments for Alvin, the stimuli categorized as moderate-preference (i.e., point weighting score between 40 and 63) were reassessed using the same procedures as in Phase I. The stimuli included the M&M, microphone, and the picture collage. Upon reassessment, Alvin selected the microphone during the first trial of the only session in this phase. The microphone was used as the moderatepreference stimulus throughout the reinforcer assessment. During Phase III of the preference assessments, the stimuli categorized as lowpreference (i.e., point weighting score between 0 and 39) were reassessed using the same procedures as in Phases I and Phase II. Stimuli included the top, truck, spinner, and bubbles. When the stimuli were reassessed, Alvin selected the truck during the first trial of the only session in this phase and it was used as the low-preference stimulus throughout the reinforcer assessment. Jake Stimuli included in Jake‘s preference assessments were: a windmill; a small piece of mozzarella cheese; a plastic tube containing green liquid and glitter (wand); a piece of 35

string; spinner that lit up and spun when a button was pressed; Capri Sun juice; a plastic tambourine; bubbles, and a small piece of tortilla chip . During Phase I of the preference assessments, unlike Levi and Alvin, Jake most often chose tangible items (e.g., spinner, tambourine, and bubbles) first from the array. As displayed in Figure 3.1, edible items were usually chosen during the last trials of the preference assessments. The bubbles, string, juice, tambourine, and chip were not selected in one of the nine trials. The cheese and chip were not selected during three of the nine sessions. The spinner and the windmill were the only two items selected in all nine preference assessments. Using the point weighting scoring method, the order from most preferred (i.e., received a higher point weighting score) to the least preferred were documented as: spinner (67 points), windmill (63 points), bubbles (51 points), tambourine (50 points), string (49 points), wand (40 points), juice (37 points), cheese (17 points), and chip (9 points). There were no documented approaches. During Phase II of the preference assessments, the stimuli categorized as moderate-preference are reassessed using the same procedures as in Phase I of the preference assessments. For Jake, the five items that were categorized as moderatepreference (i.e., point weighting score between 40 and 63) were the windmill, bubbles, string, wand, and tambourine. When the items were reassessed, Jake chose the wand during the first trial of the only session conducted in this phase. The wand was used as the moderate-preference stimulus during the reinforcer assessment. During Phase III of the preference assessments, the stimuli categorized as lowpreference (i.e., point weighting score between 0 and 39) were reassessed using the same

36

procedures as in Phases I and II. Stimuli included the chip, cheese, and juice. Upon reassessing the stimuli, Jake selected juice during the first trial of the only session in this phase and it was used as the low-preference stimulus throughout the reinforcer assessment. Experimental Data Procedural Integrity The extent to which the experimenter administered the reinforcer assessment procedures as described in the methods section was measured as procedural integrity. Procedural integrity was calculated by dividing the number of steps implemented correctly by the sum of the number of steps implemented correctly and incorrectly then multiplying by 100%. Procedural Integrity of Reinforcer Assessments Data on the procedural integrity of the reinforcer assessments are displayed in Table 3.3. During Baseline for Levi, 50% of sessions were measured, during the highpreference stimulus condition, 33% of sessions were measured, for the moderatepreference stimulus condition, 22% of sessions were measured , during the lowpreference stimulus condition, 44% of sessions were measured and procedural integrity was 100% for all conditions. Overall, 30% of all reinforcer assessment sessions were measured for Levi and mean procedural integrity was 100%. For Alvin in baseline, 33% of sessions were measured, during the high-preference stimulus condition, 33% of sessions were measured, during the moderate-preference stimulus condition, 22% of

37

sessions were measured, for the low-preference stimulus condition, 44% of sessions were measured and procedural integrity was 100% for all conditions. Overall for Alvin, 23% of sessions were measured and mean procedural integrity was 100%.

Table 3.3 Mean Procedural Integrity during the reinforcer assessments

Baseline

High-Preference

Moderate-

Low-Preference

Stimulus

Preference Stimulus

Stimulus

Levi

100 (2)ª (50)ᵇ

100 (3) (33)

100 (2) (22)

100 (4) (44)

Alvin

100 (1) (33)

100 (2) (22)

100 (2) (22)

100 (2) (22)

Jake

100 (2) (67)

100 (1) (50)

--

100 (1) (50)

Note. Conditions were alternated using a predetermined random order.

ª Number of procedural integrity sessions

ᵇ Percent of sessions with procedural integrity

For Jake during baseline, 67% of sessions were measured, during the high-preference stimulus condition, 50% of sessions were measured and procedural integrity was 100% for each condition. There were no procedural integrity data taken during the moderatepreference stimulus condition due to early termination of the intervention. For the lowpreference stimulus condition, 50% of sessions were measured and overall, 57% of

38

sessions were measured and procedural integrity was 100% for Jake. Procedural integrity was 100% for each participant in each condition for the reinforcer assessments. Interobserver Agreement of Reinforcer Assessments Approximately 31% of sessions across participants and reinforcer assessment sessions were selected for IOA. The exact number of sessions and percentage of sessions during which IOA was assessed for each participant in each condition of the reinforcer assessments is IOA was 100%. During the high-preference stimulus condition, 33% of sessions were measured, during the moderate-preference stimulus condition, 22% of sessions were measured, during the low-preference stimulus condition, 44% of sessions were measured, and overall for Levi, 30% of all reinforcer assessment sessions were measured and IOA was 100%. During baseline for Alvin, 33% of sessions were measured, during the highpreference stimulus condition, 33% of sessions were measured, during the moderatepreference stimulus condition, 22% of sessions were measured, for the low-preference stimulus condition, 44% of sessions were measured, and overall, 23% of sessions were measured and IOA was 100% for Alvin. For Jake during baseline, 67% of sessions were measured, during the highpreference stimulus condition for Jake, 50% of sessions were measured and IOA was 100%. There were no IOA data taken during the moderate-preference stimulus condition due to early termination of the intervention. For the low-preference stimulus condition, 50% of sessions were measured, and overall for Jake, 57% of sessions were measured

39

and IOA was 100%. IOA was 100% for each participant in each condition for the reinforcer assessments.

Table 3.4 Mean IOA of percent correct in each condition for each participant Baseline

High-Preference

Moderate-

Low-Preference

Stimulus

Preference Stimulus

Stimulus

Levi

100 (2) ª(50)ᵇ

100 (3) (33)

100 (2) (29)

100 (4) (44)

Alvin

100 (1) (33)

100 (2) (22)

100 (2) (22)

100 (2) (22)

Jake

100 (2) (67)

100 (1) (50)

--

100 (1) (50)

Note. Conditions were alternated using a predetermined random order.

ª

Number of IOA sessions

ᵇ

Percent of sessions with IOA

Experimental Data Levi Levi‘s data are presented in Figure 3.2 and Table 3.5. During baseline, Levi‘s mean responding was 0.25 correct responses (out of five). Number of correct responses ranged from zero to one during four total baseline sessions. The intervention was

40

Baseline

High- vs. Moderate- vs. Low-Preference Items

5

Correct Responses

Moderate-P Item

High-P Item

4

3

2

Low-P Item

1

Levi 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

Sessions

Figure 3.2. Levi‘s reinforcer assessment data. Levi‘s task was cutting through a 15.24 cm by 10.16 cm index card. The depicts responding in baseline (no reinforcement). The depicts responding in the high-preference stimulus (Craisin) condition. The depicts responding in the moderate-preference stimulus (ice) condition. The depicts responding in the low-preference stimulus (bubbles) condition.

41

5

BL

High- vs. Moderate- vs. Low- Preference Items

Correct Responses

4

3

2

Alvin

1

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Sessions Figure 3.3. Alvin‘s reinforcer assessment data. Alvin‘s task was matching each letter of his name to a placemat containing the letters to his name. The depicts responding in baseline (no reinforcement). The depicts responding in the high-preference stimulus (chip) condition. The depicts responding in the moderate-preference stimulus (microphone) condition. The depicts responding in the low-preference stimulus (truck) condition.

42

Baseline

High- vs. Moderate- vs. Low-Preference Items

Phase I

5

Correct Responses

4

3

2

High-P Item Moderate-P Item

Low-P Item

1

Jake 0 1

2

3

4

5

6

Sessions

7

8

9

10

11

Figure 3.4. Jake‘s reinforcer assessment data. Jake‘s was wiping a 25.4 cm by 25.4 cm space on a table. The depicts responding in baseline (no reinforcement). The depicts responding in the high-preference stimulus (spinner) condition. The depicts responding in the moderate-preference stimulus (wand) condition. The depicts responding in the low-preference stimulus (juice) condition.

43

introduced when baseline levels were stable at zero. During the low-preference stimulus treatment condition, mean responding was 1.4 (range, 0-2). Data were stable at two correct responses and then fell to zero. By the end of data collection, data were stable at one correct response in the low-preference stimulus condition. During the moderatepreference stimulus condition, Levi‘s mean responding was 3.5 correct responses (range, 1-5). This was the highest mean responding of all the conditions. There were seven sessions in this condition. Data appeared with an immediate upward trend and throughout this intervention data were mostly stable. By the end of the intervention, two consecutive moderate-preference stimulus data points were at five correct responses (100%). During the high-preference stimulus condition, the mean number of correct responses was 3.25 (range, 1-5). There were eight total sessions and data in this condition were variable. The first data point in this condition was at three correct responses and then fell to two correct responses. Data then began showing an upward trend, reaching five correct responses (100%). Data indicated an increasing trend to one correct response and then back up to end the intervention with two consecutive sessions of five correct responses. Levi‘s moderate- and high-preference stimulus condition data paths demonstrated a separation from the low-preference stimulus condition toward the latter half of the intervention. It is important to note that two of Levi‘s data points (i.e., session 30 and session 31) were discarded due to Levi aggressing toward the experimenter, causing the intervention to not be carried out completely. (See discussion for more details)

44

Alvin Alvin‘s data are displayed in Figure 3.3 and Table 3.5 During baseline, there were zero correct responses. There were three sessions in this condition, and intervention was introduced when data were stable at zero. During the low-preference stimulus condition,

Table 3.5. Mean responding of each participant in each condition

Baseline Number of sessions

Levi .25 (0-1) 4

Alvin 0 3

Jake 0 3

Low-Preferred Item*

1.4 (0-2)

1.3 (0-2)

0

Number of sessions

7

9

2

Moderate-Preferred Item*

3.5 (1-5)

1.89 (1-3)

0

Number of sessions

8

9

2

High-Preferred Item*

3.25 (1-5)

1.13 (0-3)

.33 (0-1)

Number of sessions

8

8

3

Note. Each session in each condition included five discrete trials. * Average number of correct responses in each condition

the mean number of correct responses was 1.3 (range, 0-2) and data were variable. Responding started at two correct responses then fell to zero. Data exhibited an upward 45

trend to two correct responses that later leveled off. The last data point in this condition ended at zero. During the moderate-preference stimulus condition, the mean number of correct responses was 1.89 (range, 1-3) for Alvin. This was the highest of all the treatment conditions. There were nine total sessions in the moderate-preference stimulus condition. Data in this condition appeared highly variable. Responding began at one correct response and then increased to two correct responses. Responding varied from one to two correct responses. In the latter half of the intervention, data responding was stable at two correct responses. During the last session in this condition, Alvin achieved three correct responses. During the high-preference stimulus condition, the mean number of correct responses was 1.13 (range, 0-3) across eight total sessions. Jake Jake‘s data are displayed in Figure 3.3 and Table 3.5. During baseline, there were no correct responses. There were three total sessions in baseline and the intervention was introduced when data were stable at zero. For the low- and moderate-preference stimulus condition, there were zero correct responses. There were two sessions total in the lowpreference stimulus condition and two sessions total in the moderate-preference stimulus condition due to the termination of treatment (see Discussion section). In the highpreference stimulus condition, the mean number of correct responses was .33 (range, 0-1) across three sessions. Due to minimal responding during intervention, the experimenter decided to have Jake sit down at the table rather than stand in hope sitting down would serve as a discriminative stimulus to do work and may improve upon Jake‘s attending skills. This phase change is represented by Phase I in the data displayed in Figure 3.3. 46

Social Validity After the study, one consumer (the teacher in the classroom) rated the effectiveness of intervention, satisfaction with procedures, and likelihood of using preference assessments. The results are displayed in Table 3.6. The rating scale was as follows: 1 (not at all satisfied/effective/likely), 2 (somewhat not satisfied/effective/likely), 3 (neutral), 4 (somewhat satisfied/effective/likely), and 5

Table 3.6. Social Validity Assessment Survey Question

Response

1. How satisfied are you with the changes in the students‘ behavior regarding the assigned tasks (i.e., Luke: cutting paper, Andy: matching letters to name)?

4

2. How effective do you believe the preference assessments were in identifying reinforcers?

5

3. How likely are you to use preference assessments in your classroom?

4

4. How likely are you to recommend preference assessments?

5

Note. Results of the social validity assessment of the effectiveness and satisfaction of the preference assessments reported by the classroom teacher.

(very satisfied/effective/likely). When asked the level of satisfaction with the changes in the students‘ behavior regarding the assigned tasks, the teacher answered 4 ―somewhat satisfied.‖ The teacher rated the preference assessments as ―very effective‖ (which was a 47

5) at identifying reinforcers for the students. Regarding the likelihood of using the preference assessment in the classroom, the teacher responded she was ―somewhat likely‖ (which was a 4) to use them. The teacher responded she was ―very likely‖ (which was a 5) to recommend the use of preference assessments. The mean level of satisfaction was a 4.5. The teacher also gave some general comments, which are displayed in Table 3.7.

Table 3.7 General Social Validity Comments General Comments Strengths

―The assessments did a great job of determining the students‘ preferences, therefore, determining great reinforcers to be used with the students throughout the day.‖ ―This was a great project for me to observe and learn from.‖

Concerns

―It was difficult to find a task for a couple of the students.‖ ―Since we have a functional curriculum and the task needed to be something that could be done at the table, I struggled coming up with good table tasks.‖

48

Chapter 4: Discussion This study evaluated the efficacy of moderate- and low-preference stimuli as reinforcers using Multiple Stimulus Without Replacement (MSWO) preference assessments. There were nine stimuli for each participant, both edible and tangible items, and nine full array MSWO preference assessment implemented. The stimuli from the preference assessments were categorized using the point weighting scoring method developed by Ciccone et al. (2005). For two of the three participants, responding increased when reinforcement contingencies were implemented. This finding is consistent with previous literature demonstrating the MSWO can effectively identify reinforcers (Carr et al., 2000; Daly et al., 2009; DeLeon & Iwata, 1996; Reid et al., 2007) and implementing reinforcement contingencies can increase responding, regardless of the stimuli (Daly et al., 2009; Paramore & Higbee, 2005; Taravella et al., 2000). For one participant, the moderate- and high-preference item were effective at increasing responding at 100%, which extends Ciccone et al.‘s (2006) results noting moderatepreference stimuli may be likely to function as reinforcers. Data for one participant were undifferentiated, which may have been due to carry over effects. The point weighting scoring system (Ciccone et al, 2005) was used to categorize stimuli as low-, moderate-, and high-preference and a reinforcer assessment was conducted to verify results. According to findings, this is the first study to extend the point weighting scoring system

49

to rank stimuli during MSWO preference assessments and to also evaluate it using reinforcer assessments. The remainder of this chapter discusses the results relative to the research questions, addresses limitations of the study, suggests implications for practice, and offers recommendations for future research. Discussion of Results by Research Question Research Question 1: To what extent do high-, moderate- and low-preference stimuli, such as edibles and tangibles, function as reinforcers? Overall responding increased for two of the three participants during the reinforcer assessment and all of the stimuli (i.e., low-, moderate-, and high-preference), had a reinforcing effect. For Levi, data were initially undifferentiated during the reinforcer assessment, but the moderate- and high-preference data later separated. The literature suggests the high-preference stimulus typically has a greater effect on responding and this effect usually persists over time (Carr et al., 2000; Graff et al., 2006; Paramore & Higbee, 2005). Data for Alvin were undifferentiated throughout the entire reinforcer assessment. These results (i.e., undifferentiated data) suggest that simply implementing reinforcement contingencies can increase responding, which is consistent with previous findings (Daly et al., 2009; Paramore & Higbee, 2005; Taravella et al., 2000). Alvin‘s results may have also been due to carry over effects. For all participants, it seems the lack of demands in the classroom made the reinforcer tasks seemingly aversive. Participants were asked to join the experimenter at the table in the classroom to work while other students in the classroom continued engaging in leisure play. All participants occasionally displayed challenging behaviors during the reinforcer 50

assessments. Following are analyses of this research question with each participant, particularly in terms the effects of the stimuli on responding. Levi. During the preference assessments, Levi selected edible stimuli within the first trials, in accordance with previous studies noting tangible items may be displaced during MSWO preference assessments when edibles are available (DeLeon et al., 1997). Levi‘s responding increased during the reinforcer assessment immediately after reinforcement contingencies were implemented. There was little differentiation of data during the initial sessions of the reinforcer assessment. The low-, moderate-, and highpreference stimuli all had a reinforcing effect on Levi‘s responding. Mean responding during baseline was .25. Mean responding in low-preference stimulus condition (bubbles) was 1.4, demonstrating an increase. For Levi, the low-preference stimulus was the least effective. Mean responding in the moderate-preference stimulus condition (ice) was 3.5 and was the highest of all the conditions; these results contradict results reported in the literature (e.g., Carr et al., 2000; Paramore & Higbee, 2005). Data in the high-preference stimulus condition were variable and mean responding was 3.25, slightly below the mean during the moderate-preference condition. During sessions 21 and 22 where responding decreased, procedural integrity and IOA were measured by a second, independent observer. Levi appeared easily distracted by the second observer, which may explain the decrease in correct responses during these sessions. The high-preference stimulus was a Craisin, which was consumed immediately every time. The moderate-preference stimulus was a full cup of ice and Levi was given 10 s access contingent on correct responding. It is unclear why the moderate-preference

51

stimulus had a similar effective compared to the high-preference stimulus as the participant chose the Craisin almost immediately during the preference assessments but waited until the additional edibles were gone before selecting the ice. At no time did Levi appear satiated. One consideration is the duration of the reinforcer assessment. As mentioned, the moderate- and high-preference stimuli were effective reinforcers and allowed Levi‘s responding to reach 100%. It is unknown how long the reinforcing effects of the moderate- and high-preference stimulus would have been maintained if more sessions had been conducted. Previous literature suggests the reinforcing effects of the high-preference stimulus are more likely to persist over time (Carr et al., 2000). Levi achieved five correct responses (i.e., 100%) in the last four consecutive sessions of the reinforcer assessment. However, two sessions were completed after this, but are not displayed in the graph (refer to results). After consideration, the experimenter discarded the data due to speculation of serious confounding variables effecting the participant‘s responding, attending skills, and overall affect. During one of the sessions in which data were not included, Levi aggressed toward the experimenter in a way not observed before and eloped from the work table. Levi also screamed continually during these assessments. During the reinforcer assessment, the experimenter initially had to aid Levi to correctly place the scissors in his hand and to pick up the index card. This prompt was faded throughout the reinforcer assessment and during the last seven sessions, Levi was independently picking up the index card and scissors and cutting.

52

It is noteworthy to mention none of the stimuli used in the reinforcer assessment were listed in the RAISD. Stimuli mentioned in the RAISD that were included in the preference assessments for Levi included the puzzle, chest, and mirror. The puzzle was ranked fifth, chest was ranked sixth, and the mirror was ranked last (item chosen the least during the nine complete MSWO assessments). The stimuli used in the reinforcer assessment (e.g., craisin) were items Levi‘s teacher suggested, which contrasts findings by Resestar and Noeell (2008) finding teacher ranked items never ranked in the top for stimuli chosen during an MSWO. Alvin. Similar to Levi, Alvin selected the edible stimuli during the first trials of the preference assessments, contributing to findings noting a preference for edible items over non-edible items during preference assessments (e.g., DeLeon & Iwata, 1996; DeLeon et al., 1997; Taravella et al., 2000). Stimuli used in Alvin‘s preference assessments were selected by a structured interview with his teacher as Alvin‘s caregiver did not complete the RAISD. Many stimuli were not selected during the preference assessments. The experimenter noted that after the edible stimuli were selected, Alvin would engage in stereotypic behavior (e.g., spinning body, vocalizations) Also, when Alvin was told to pick one during the preference assessments, on several occasions he would whine and/or scream, which is the behavior Alvin displays when given demands from the classroom staff. Engaging in stereotypic behavior and perceiving the instruction to pick one as a demand rather than an opportunity to select preferred items may explain why some stimuli were not selected during preference assessments. Changing the Sᴰ to something like what do you want may have reduced inappropriate behaviors.

53

Overall, responding increased relative to baseline where responding was zero. Throughout the reinforcer assessment, Alvin‘s data paths were undifferentiated. By inspecting the graph, it is unclear which stimulus had a greater effect. Mean responding during the low-preference stimulus condition (truck) was 1.3, the second highest of all conditions. When Alvin was given access to the truck, he most often displayed stereotypic behavior rather than engaging with the item. Mean responding in the moderate-preference stimulus condition (microphone) was 1.89, the highest of all conditions. This finding was also demonstrated in Levi‘s data. Previous literature suggests moderate-preference stimuli can function as reinforcers (Ciccone et al., 2006; Daly et al., 2009; Piazza et al., 1996), but this literature does not suggest they function more effectively as reinforcers than high-preference stimuli. Mean responding in highpreference stimulus condition (chip) was 1.13, the lowest of the reinforcement conditions. This finding differs greatly from previous findings (Carr et al., 2000; Ciccone et al, 2006; Daly et al., 2009; Graff et al., 2006; DeLeon & Iwata, 1996; Graff et al., 2006; Paramore & Higbee, 2005; Piazza et al., 1996). Penrod, Wallace, and Dyer (2008) noted the lowpreference stimulus was equally effective as the high-preference stimulus for one of four participants under increasing schedule requirements However, data for the other three participants in this study suggested the high-preference stimuli maintained responding under high schedule requirements relative to low-preference stimuli. Although procedures for Penrod et al. (2008) and Roscoe et al. (1999) are dissimilar, Alvin‘s data extend the finding that it is possible for the low-preference stimulus to produce greater responding than the high-preference stimulus when looking at the mean. This finding

54

should be taken into consideration when administering reinforcement in applied settings to prevent satiation (Ciccone et al., 2006). Looking at trends in the data, the moderatepreference stimulus produced the most stable and effective results. Responding in the low-preference stimulus condition began to decrease during the latter half of the intervention. It is unclear why data were differentiated throughout the intervention. One explanation may be carry over effects. As discussed, Alvin selected the edible items during the first trials of the preference assessments and most of the time appeared disinterested in the remaining stimuli. At no point during the preference or reinforcer assessments did Alvin appear satiated. Alvin is the only participant whose stimuli came only from teacher interview and not from the RAISD. Another consideration is the utility of the Ciccone et al., (2005) scoring system in regards to effectively identifying and ranking reinforcing stimuli. Jake. During the preference assessments, Jake selected the tangible items during the first trials and edible stimuli during the last trials or not at all. This contrasts previous findings that edible stimuli typically displace tangible stimuli during MSWO preference assessments (DeLeon et al., 1997; Taravella et al., 2000). Two of the stimuli (i.e., spinner and bubbles) listed in the RAISD completed by his caregiver were selected for use in the preference assessments. The spinner was ranked as the high-preference stimulus by scoring the highest point weighting score. It was used during the reinforcer assessment as the high-preference stimulus. The bubbles had the third highest point weighting score but were not selected upon reassessment of the moderate-preference items. The moderate-

55

and low-preference stimuli used in the reinforcer assessment were items suggested by Jake‘s teacher and previous literature (e.g., Mintz, Wallace, Najdowski, Atcheson, & Bosch, 2007; Reid et al., 2007; Resestar & Noell, 2008) note teacher-selected stimuli may be effect reinforcers. Jake‘s responding during baseline was zero. Responding during the reinforcer assessment remained similar. Responding during the low- and moderate-preference stimulus condition was zero and mean responding during the high-preference stimulus condition was 0.33. Jake only responded correctly one time during the reinforcer assessment. Throughout the entire reinforcer assessment, Jake displayed several problem behaviors. For example, he displayed acts of aggression, such as hitting and kicking, toward the experimenter nearly every session. During the study, Jake had a behavior plan in place for acts of aggression maintained by access to tangible items as determined by a functional assessment. Jake would reach for the reinforcer during the conditions and had to be prompted to complete the task instead. This may have made Jake frustrated because he was not contacting reinforcement contingencies during the reinforcer assessment. The experimenter had to prompt and physically guide Jake to stay at the table to complete the task. Of relevance are Jake‘s minimal attending skills, which is a prerequisite skill for almost every task and this deficit is likely to have contributed to Jake‘s lack of responding. Intervention was eventually terminated due to Jake‘s problem behaviors and minimal attending skills, which is speculated to be the cause of his lack of responding. The experimenter probed other tasks, such as matching letters of his name to a placemat,

56

placing round shapes through holes in a bucket, and a picture exchange task to gain access to tangible and edibles. During these tasks, Jake still engaged in severe challenging behavior and attempted to elope from the tasks several times. Summary. Levi and Alvin preferred edible stimuli over tangible stimuli during preference assessment and the stimuli scoring the highest using the point weighting method (high-preference stimulus) was an edible item, contributing to previous findings (DeLeon et al., 1997; Taravella et al., 2000). For Levi, the moderate- and high-preference stimulus produced the highest responding at 100%, contributing to literature suggesting the moderate-preference stimulus can function as a reinforcer (Ciccone et al., 2006). Levi and Alvin‘s data contribute to findings (e.g., Penrod et al., 2008; Roscoe et al., 1999) indicating low-preference stimuli can function as reinforcers. After reinforcement contingencies were implemented, Levi and Alvin‘s responding immediately increased, suggesting reinforcement, no matter the preference, can increase responding. Jake‘s responding was stable at zero (with the exception of one session) due to challenging behaviors and lack of attending skills. Research Question 2: Do high-preference stimuli, such as edibles and tangibles, mask other high-preference stimuli during a multiple-stimulus without replacement preference assessment? For Levi and Alvin, the moderate-preference item had the greatest effect on responding, yet was ranked third (Levi) and fourth (Alvin) using the point weighting scoring method. If stimuli categorized as moderate-preference were not reassessed and available for reinforcement contingent on responding, the reinforcing effectiveness of

57

those stimuli would have been unknown and most likely underestimated. Previous literature (e.g., DeLeon et al., 1997; Taravella et al., 2000) note tangible items can be displaced when edible items are available during a multiple-stimulus preference assessment. It is possible during the preference assessments the edible stimuli masked tangible stimuli for Levi and Alvin. For Levi, though, it was an edible item (ice) that produced the highest mean responding and it was ultimately edible items (ice and Craisin) that produced responding at 100%. Taravella et al. (2000) found less preferred stimuli may not appear to be potent reinforcers when more preferred stimuli are available, but when the less preferred stimuli are reassessed, they appear to be reinforcing. Findings in this study are consistent with Taravella et al. (2000) and may suggest that highpreference stimuli do mask other high-preference stimuli during multiple-stimulus preference assessments. During MSWO preference assessments, three or more stimuli are available. It is common for researchers to select the highest ranked item from various preference assessment formats and use this item alone for comparison to baseline responding, rather than evaluating the utility of moderate- and low-preference stimuli (Ahearn et al, 2005; Kodak et al., 2009; LeBlanc et al., 2006; O‘Reilly et al., 2004; Resetar & Noell, 2008; ). The results from the current study demonstrated that regardless of the stimulus preference (i.e., low, moderate, or high), simply implementing reinforcement contingencies increased responding for Levi and Alvin. An interesting, and potentially important point from this finding suggests using stimuli categorized as moderate-preference (as identified by MSWO preference assessments and Ciccone et al. (2005) point weighting scoring

58

method) as reinforcers may be equally or more effective as the high-preference stimuli. In addition, because low-preference stimuli increased responding for Levi and Alvin, they also should be considered reinforcers until proven otherwise. Research Question 3: How beneficial did relevant consumers (e.g., teachers) of this study find using a multiple-stimulus without replacement preference assessment to identify preferences and reinforcers? The classroom teacher generally found the preference assessments to be effective and useful to identify reinforcers. The participants‘ teacher observed many preference assessments being conducted while the experimenter explained and demonstrated the procedures. She also completed a social validity survey rating the effectiveness of the intervention, satisfaction with procedures, and likelihood of using preference assessments. The teacher rated the preference assessments as very effective at identifying reinforcers for students. She also made general comments, such as ―the assessments did a great job of determining the students‘ preferences,‖ and ―this was a great project for me to observe and learn from.‖ She responded she was somewhat likely to use them in her classroom. The teacher seemed genuinely interested in the intervention and responded she was very likely to recommended preference assessments. The teacher was also asked how satisfied she was with the participants‘ behavior change, and she responded somewhat satisfied. Most of her general concerns were premised around the reinforcer assessments. She stated ―it was difficult to find a task for a couple of the students.‖ Despite the research supporting the use of preference assessments to identify reinforcers for students, the majority of the research has not been

59

conducted by direct care providers (Cannella et al., 2005). As Tullis et al. (submitted) note, there is a need for further research in training professionals to conduct preference assessments. Limitations There are several limitations of this study discussed. First, an alternating treatments design was used to evaluate the effects of the high-, moderate-, and lowpreference stimuli. Due to the relatively short duration of the reinforcer assessments, certain limitations to this design, such as multiple treatment interference, were not able to be teased out. It is unclear how implementing only one stimulus at a time, rather than rapidly alternating them, would have impacted responding. Also, the treatments were not significantly different from one another, making it unclear if the participants were able to discriminate between treatment conditions. Although different color and textured placemats were used, data for two the participants were undifferentiated, which may or may not be due to their inability to discriminate between treatment conditions. A reversal design would have been more suitable for Alvin. In addition, it is unclear which stimulus would have produced greater responding if treatment had continued. The relatively short duration of the reinforcer assessment is a limitation of this study. If the intervention had been in place for a longer period of time, it would have been possible to evaluate which stimulus produced higher responding over time and/or acted as an abolishing operation due to satiation effects. Previous literature indicates the high-preference stimulus maintains higher rates of responding for longer, so

60

it is possible the high-preference stimulus would have functioned more effectively as a reinforcer if the intervention continued. Next, reinforcing effects were only evaluated under a rich schedule of reinforcement (i.e., FR 1). It is unclear if undifferentiated data for two participants would have changed if the schedules of reinforcement for each stimulus increased. DeLeon, Iwata, Goh, and Worsdell (1997) demonstrated under increasing schedule requirements, responding increased for one item when items were similar (e.g., two food items) and responding was undifferentiated when items were dissimilar (e.g., a food and a leisure item). Another limitation to this study is the reinforcing value of each stimulus was not evaluated. It remains unclear if the additional stimuli categorized as high-, moderate-, and low-preference would have functioned as reinforcers. For all the participants, many point weighting scores only varied by one or two points. For example, for Levi, the point weighting score for the Craisin was 75 and the point weighting score for the M&M was 74. It is possible the M&M or another stimulus would have produced greater responding. It is also quite possible that all the stimuli would have functioned as reinforcers. Finally, an additional limitation to this study is the use of the point weighting method. Ciccone et al., (2005) states it is a superior method to the traditional way of categorizing stimuli using a MSWO preference assessment. However, it appears as though this is the first study to use the point weighting method and conduct a reinforcer assessment. The validity of this method is unknown. Ciccone et al. (2006) suggests it is a possibility this method overestimates preference for stimuli. Not comparing the results

61

from the point weighting method to a traditional scoring method is a limitation of this study. Implications for Practice The current study has implications for practice that may potentially benefit staff and students. Throughout the literature and discussion, there is great emphasis placed on the use of reinforcement. In special education, the need for evidence-based practices is tremendous. Preference assessments have undergone methodologically rigorous analysis and have been considered an effective means of identifying reinforcers (e.g., Daly et al., 2009; DeLeon et at., 1996; Fisher et al., 1992; O‘Reilly et al., 2004). Choice interventions have been described as an evidence-based practice for individuals with severe to profound disabilities (Tullis et al., submitted). The MSWO preference assessment is beneficial for teachers, therapists, or other staff working in applied settings needing to identify various reinforcers. When evaluating the use of the MSWO, it is important to take into consideration the procedures (i.e., multiple stimuli in an array). The procedures used in a MSWO preference assessment may not be conducive for identifying preferences with individuals with vision impairments. For teachers, the MSWO offers the ability to effectively identify reinforcers in a relatively short period of time. Compared to other preference assessment formats, the MSWO is more efficient and effective at identifying reinforcers. After identifying reinforcers, research demonstrates reinforcers can be used to manage classroom behavior by increasing appropriate behaviors and decreasing challenging behaviors (e.g., Cannella et al., 2005; Carlson et al., 2008; Cole & Levinson, 2002).

62

Recommendations for Future Research The data presented from this study offer many directions for future research. First, further research should be conducted evaluating the Ciccone et al., (2005) point weighting method of categorizing stimuli during MSWO preference assessments. Researchers should evaluate reinforcer effectiveness of stimuli identified using the point weighting method compared to stimuli identified using the percentage approach method (Fisher et al., 1992), which is traditionally used to rank stimuli. Additional research should examine prerequisite skills needed for particular preference assessment formats. It remains unclear if one format is more effective for a particular individual, population, or the individual administering the preference assessment (Tullis et al., submitted). Determining the prerequisite skills needed for each preference assessment format will help address this concern. Also, when evaluating responding with an individual with severe to profound cognitive disabilities, a basic requisite skill, such as attending skills, should be evaluated prior to data collection. In this study, Jake‘s attending skills should have been predetermined; his results may have been interpreted more accurately as a result. As it stands, it is unclear if his lack of attending skills, challenging behavior, and/or stimuli were not reinforcing which affected the results. Changing Jake‘s task to one that would teach attending skills, such as making eye contact, may have been more appropriate. Further research is needed to evaluate the utility of moderate- and low-preference stimuli as reinforcers. Several studies support the use of moderate-preference stimuli as reinforcers (e.g., Ciccone et al., 2006) and the use of low-preference stimuli as reinforcers

63

(e.g., Daly et al., 2009; Roscoe et al., 1999; Taravella et al., 2000). However, there are studies demonstrating that the high-preference stimulus produce more reinforcing effects (e.g., Paramore & Higbee, 2005). One of the most important components of preference assessments is the reinforcer assessment. Future research should examine which reinforcer assessment procedures evaluate stimuli as reinforcers most effectively. A plethora of research had evaluated various components, formats, and stimuli used in preference assessments, but comparing experimental designs and reinforcer assessment procedures is arguably one of the most important aspects. Summary Providing choices, from simple to basic, can increase the quality of life for individuals with severe to profound disabilities. Choice interventions are used to increase appropriate behavior and decrease challenging behaviors (Cannella et al., 2005; CannellaMalone et al., 2009; Carlson et al., 2008; Cole & Levinson, 2002; LeBlanc et al., 2006; O‘Reilly et al., 2004; Piazza et al., 1996; Tiger et al., 2006). Identifying preferences of individuals with severe to profound disabilities can be very challenging. Preference assessments have been widely researched and demonstrate efficacy at identifying preferences and reinforcing stimuli. Preference assessments have been investigated with individuals with mild to severe disabilities (e.g., Ahearn et al., 2005; Hanley et al., 2006), individuals with emotional and/or behavioral disorders (e.g., Daly et al., 2009; Taravella et al., 2000; Paramore & Higbee, 2005), physical and medical disabilities (e.g., LeBlanc

64

et al., 2006), in educational settings (Hanley et al., 2007) and with general education students (Resestar & Noell, 2008; Schanding et al., 2008). Three students with mental retardation and/or autism participated in this study. Nine full array MSWO preference assessments were conducted assessing tangible and edible items. The items categorized as moderate-preference using the Ciccone et al. (2005) point weighting scoring method were reassessed. The items categorized as lowpreference were also reassessed using similar procedures. The highest ranked item from the first nine complete preference assessments, the moderate-preference reassessment, and the low-preference reassessment were used as reinforcement during the reinforcer assessment. An alternating treatments design as implemented and the effectiveness of the reinforcers relative to baseline were compared using an A-B design. Reinforcement was delivered contingent upon correct responding of a predetermined task chosen by the classroom teacher. Results indicated the moderate-preference item was effective at increasing responding for two of the three participants. One participant had minimal responding and treatment was terminated prematurely. One participant achieved responding at 100% for four consecutive sessions during the high- and moderate-preference stimuli conditions. Two of the three participants‘ responding increased as the reinforcement contingency was implemented and the low-, moderate-, and high-preference stimuli all functioned as reinforcers. Two of the participants‘ data were undifferentiated throughout the study. Assessment of the effectiveness of the intervention and satisfaction with the outcomes with one consumer of the research indicated high levels of satisfaction with the

65

preference assessment and moderately high levels of satisfaction with the students‘ behavior change. Perhaps the biggest concern was having difficulty determining tasks for some of the participants. The results of this study are consistent with the literature on preference assessments (Carr et al., 2000; Ciccone et al., 2006; Daly et al., 2009; DeLeon & Iwata, 1996; DeLeon et al., 1997; Paramore & Higbee, 2005; Penrod et al., 2008; Roscoe et al., 1999; Taravella et al., 2000). The results add to the literature by demonstrating the moderate-preference stimulus may function more effectively as a reinforcer when compared to the high-preference stimulus. Reinforcement contingencies, no matter the preference, increased responding for two the three participants, suggesting lowpreference stimuli may function as reinforcers. Minimal responding by Jake was likely due to his challenging behaviors and minimal attending skills. Future research should be conducted to evaluate the Ciccone et al. (2005) point weighting scoring method in comparison to the percentage approach method (Fisher et al., 1992) and to further examine moderate- and low-preference stimuli as reinforcers. Teachers and staff working with children with severe to profound disabilities should make use of preference assessments as they are an effective and efficient technique to identify preferences and reinforcing stimuli.

66

References Ahearn, W. H., Clark, K. M., Debar, R., & Florentino, C. (2005). On the role of preference in response competition. Journal of Applied Behavior Analysis, 38, 247-250. Cannella, H. I., O‘Reilly, M. F., & Lancioni, G. E. (2005). Choice and preference assessment research with people with severe to profound developmental disabilities: A review of the literature. Research in Developmental Disabilities, 26, 1-15. Cannella-Malone, H. I., DeBar, R. M., & Sigafoos, J. (2009). An examination of preference for augmentative and alternative communication devices with two boys with significant intellectual disabilities. Augmentative and Alternative Communication, 25, 262-273. Carlson, J. I., Luiselli, J. K., Slyman, A., & Markowski, A. (2008). Choice-making as an intervention for public disrobing in children with developmental disabilities. Journal of Positive Behavior Interventions, 10, 86-90. Carr, J. E., 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. Ciccone, F. J., Graff, R.B., & Ahearn, W.H. (2005). An alternate scoring method for the multiple stimulus without replacement preference assessment. Behavioral Interventions, 20, 121-127. Ciccone, F. J., Graff, R. B., & Ahearn, W. H. (2006). Stimulus preference assessments and the utility of a moderate category. Behavioral Interventions, 21, 59-63. Cole, C. L., & Levinson, T. R. (2002). Effects of within-activity choices on the challenging behavior of children with severe disabilities. Journal of Positive Behavior Interventions, 4, 29-37. Cooper, J. O., Heron, T. E., & Heward, W. L. (2007). Applied behavior analysis (2nd Ed). Upper Saddle River, NJ: Merrill/Prentice Hall. 67

Daly, E. J., Wells, N. J., & Swanger-Gagne, M S., Carr, J. E., Kunz, G. M. & Taylor, A. M. (2009). Evaluation of the multiple stimulus without replacement preference assessment method using activities as stimuli. Journal of Applied Behavior Analysis, 42, 563-574. DeLeon, I. G. & Iwata, B. A. (1996) Evaluation of a Multiple-Stimulus Presentation Format For Assessing Reinforcer Prerferences. Journal of Applied Behavior Analysis, 29, 519-533. DeLeon, I. G., Iwata, B. A., Goh, H., & Worsdell, A. S. (1997) Emergence of reinforcer preference as a function of schedule requirements and stimulus similarity. Journal of Applied Behavior Analysis, 30, 439-449. DeLeon, I. G., Iwata, B. A., & Roscoe, E. M. (1997). Displacement of leisure reinforcers by food during preference assessments. Journal of Applied Behavior Analysis, 30, 475-484. Fisher, W. W., Piazza, C C., Bowman, L. G., & Amari, A. (1996). Integrating caregiver report with a systematic choice assessment to enhance reinforcer identification. American Journal on Mental Retardation, 101, 15-25. Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L. P., Owens, J. C., & Slevin, I. (1992). A comparison of two approaches for identifying reinforcers for persons with severe and profound disabilities. Journal of Applied Behavior Analysis, 25, 491-498. Glover, A. C., Roane, H. S., Kadey, H. J., & Grow, L. L. (2008). Preference for reinforcers under progressive- and fixed-ratio schedules: A comparison of single and concurrent arrangements. Journal of Applied Behavior Analysis, 41, 163-176. Graff, R. B., & Gibson, L. (2003). Using pictures to assess reinforcers in individuals with developmental disabilities. Behavior modification, 27, 470-483. Graff, R. B., Gibson, L., & Galiatsatos, G. T. (2006). The impact of high- and lowpreference stimuli on vocational and academic performances of youths with severe disabilities. Journal of Applied Behavior Analysis, 39, 131-135. Hanley, G.P., Cammilleri, A. P., Tiger, J. H., & Ingvarsson, E. T. (2007). A method for describing preschoolers‘ activity preferences. Journal of Applied Behavior Analysis, 40, 603-618.

68

Hanley, G. P., Iwata, B. A., Lindberg, J. S., & Conners, J. (2003). Response-restriction analysis: I. assessment of activity preference. Journal of Applied Behavior Analysis, 36, 47-58. Hanley, G. P., Iwata, B. A., & Roscoe, E. M. (2006). Some determinants of changes in preference over time, Journal of Applied Behavior Analysis, 39, 189-202. Hanley, G. P. Iwata, B. A., Roscoe, E. M., Thompson, R. H., & Lindberg, J. H. (2003). Response-restriction analysis: II. alteration of activity preferences. Journal of Applied Behavior Analysis, 36, 59-76. Hanley, G. P., Piazza, C.C., Fisher, W.W., & Maglieri, K.A. (2005). On the effectiveness of and preference for punishment and extinction components of function-based interventions. Journal of Applied Behavior Analysis, 38, 51-65. Horner, R. H., Carr, E. G., Halle, J., McGee, G., Odom, S., & Wolery, M. (2005) The use of single-subject research to identify evidence-based practice in special education. Exceptional Children, 71, 165-179. Horrocks, E. L., & Morgan, R. L. (2009). Comparison of a video-based assessment and a multiple stimulus assessment to identify preferred jobs for individuals with significant intellectual disabilities. Research in Developmental Disabilities, 30, 902-909. Kodak, T., Fisher, W. W., Kelley, M. E., & Kisamore, A. (2009). Comparing preference assessments: Selection- versus duration-based preference assessment procedures. Research in Developmental Disabilities, 30, 1068–1077. Lancioni, G.E., O‘Reilly, M.F. & Emerson, E. (1996). A review of choice research with people with severe and profound developmental disabilities. Research in Developmental Disabilities, 17, 391-411. LeBlanc, L. A., Cherup, S. M., Feliciano, L., & Sidener, T.M. (2006). Using ChoiceMaking Opportunities to Increase Activity Engagement in Individuals with Dementia. Journal of Alzheimer’s Disease and Other Dementias, 21, 318-325. O‘Reilly, M. F., Lancioni, G. E., & Sigafoos, J. (2004). Using Paired-Choice Assessment to Identify Variables Maintaining Sleep Problems in a Child with Severe Disabilities. Journal of Applied Behavior Analysis, 37, 209-212. Pace, G. M., Ivancic, M. T., Edwards, G. L, Iwata, B. A., & Page, T. A. (1985). Assessment of Stimulus Preference and Reinforcer Value with Profoundly Retarded Individuals. Journal of Applied Behavior Analysis, 18, 249-255. 69

Paramore, N. W. & Higbee, T. S. (2005). An Evaluation of a Brief Multiple-Stimulus Preference Assessment with Adolescents with Emotional-Behavioral Disorders in an Educational Setting. Journal of Applied Behavior Analysis, 38, 399-403. Penrod, B., Wallace, M. D., & Dyer, E. J. (2008). Assessing potency of high- and lowpreference reinforcers with respect to response rate and response patterns. Journal of Applied Behavior Analysis, 41, 177-188. Piazza, C. C., Fisher, W. W., Hagopian, L. P. Bowman, L. G., & Toole, L. (1996). Using a choice assessment to predict reinforcer effectiveness. Journal of Applied Behavior Analysis, 29, 1-10. Reid, D. H., Parsons, M. B., Towery, D., Lattimore, L. P., Green, C. W., & Brackett, L. (2007). Identifying work preferences among supported worked s with severe disabilities: Efficiency and accuracy of a preference-assessment protocol. Behavioral Interventions, 22, 279–296. Resetar, J. L. & Noell, G. H. (2008). Evaluating preference assessments for use in the general education population. Journal of Applied Behavior Analysis, 41, 447-451. Roane, H. S., Vollmer, T. F., Ringdahl, J. E., & Marcus, B. A. (1998) Evaluation of a brief stimulus preference assessment. Journal of Applied Behavior Analysis, 31, 605-620. Roscoe, E. M., Iwata, B. A., & Kahng, S. W., (1999). Relative versus absolute reinforcement effects: Implications for preference assessments. Journal of Applied Behavior Analysis, 32, 105-108. Schanding, G. T., Tingstrom, D. H., & Sterling-Turner, H. E. (2008). Evaluation of stimulus preference assessment methods with general education students. Psychology in the Schools, 46(2), 89-99. Sparrow, S., Balla, D., & Cicchetti, D. (1985). Vineland Adaptive Behavior Scales. Circle Pines, MN: American Guidance Service. Taravella, C. C., Lerman, D. C., Contrucci, S. A., & Roane, H. S. (2000). Further Evaluation of Low-Ranked Items in Stimulus-Choice Preference Assessments. Journal of Applied Behavior Analysis, 33, 105-108. Tiger, J. H., Hanley, G. P., & Hernandez, E. (2006). An Evaluation of the Value of Choice with Preschool Children. Journal of Applied Behavior Analysis, 39, 1-16.

70

Tullis, C. A., Cannella-Malone, H. I., Basbigill, A. R., Yeager, A., Fleming, C. V., Payne, D., Wu, F. A review of the choice and preference assessment literature for individuals with severe to profound disabilities (submitted; 2010) Wilder, D. A., Schadler, J., Higbee, T. S., Haymes, L. K., Bajagic, V., & Register, M. (2008). Identification of olfactory stimuli as reinforcers in individuals with autism: A preliminary investigation. Behavioral Interventions, 23, 97–103. Windsor, J., Piche, L. M., & Locke, P. A., (1994). Preference testing: A comparison of two presentation methods. Research in Developmental Disabilities, 15, 439-455.

71

Appendix A: Student Information Letter

72

Special Education School of Physical Activity & Educational Services College of Education and Human Ecology PAES Building, Third Floor 305 West 17th Avenue Columbus, OH 43210-1224 Phone: 614-292-8709 FAX: 614-292-4255 http://ehe.osu.edu/paes/

Dear Parents, We would like to include your child in a study that will examine methods to assess your child‘s preferences. The purpose of this project is to determine whether moderate and low preferred items can function as reinforcers. If your child participates in this study, sessions will be conducted 4-5 times weekly for twenty to thirty minutes. Three sets of preference assessments will be conducted with your child. First, we will conduct eleven preference assessments in which a series of nine items will be lined up and your child will be given the opportunity to choose one of the items. A choice will be defined as your child making physical contact with the item. This process will be continued until all the items are selected or until your child opts to not make a choice. The items that were categorized at moderate preferred will be reassessed using the same procedures. The items categorized as low preferred will also be reassessed using the same procedures. Once we have identified those items that are preferred based on the results of all three assessments, we will assess whether or not those items are reinforcing to your child. In other words, can we use those items to increase skills in your child. This will be done by asking your child to do a task, such as cutting paper or writing his/her name and providing access to the preferred item. If that behavior increases, it would suggest that the items are indeed preferred and reinforcing. If this study is successful, your child‘s teacher will have a way to systematically determine what your child does and does not like. They can then use these items to reinforce and teach new skills. I will be leading this project with the assistance of OSU graduate students. If you would like more information, please feel free to contact me at the phone number or email address below. If you would like for your child to participate in this study, please sign the attached consent form and return it to your child‘s teacher. Please know that your consent for your child‘s participation is voluntary, you can refuse to answer questions that you do not wish to answer, and you can refuse your child‘s participation or withdraw your child at any time without penalty or repercussion. Thank you for your time and attention. Helen I. Malone, Ph.D Assistant Professor Diane M. Sainato, Ph.D. Associate Professor Special Education  Sport & Exercise Education, Humanities, Management & Science Counselor Education & School Psychology  Workforce Development & Education

73

Appendix B: Consent Form

74

PARENTAL PERMISSION Behavioral/Social Science

IRB Protocol Number: IRB Approval date: Version:

2008B0366 1/26/09 1.3

The Ohio State University Parental Permission For Child’s Participation in Research

Study Title:

Assessing preference in students with severe intellectual impairments: Multiple Stimulus Without Replacement

Researcher:

Helen I. Malone & Diane M. Sainato

Sponsor:

FCBMRDD

This is a parental permission form for research participation. It contains important information about this study and what to expect if you permit your child to participate. Your child’s participation is voluntary. Please consider the information carefully. Feel free to discuss the study with your friends and family and to ask questions before making your decision whether or not to permit your child to participate. If you permit your child to participate, you will be asked to sign this form and will receive a copy of the form. Purpose: The purpose of this study is to determine preferred items and assess if those items can function as reinforcers for students with severe intellectual disabilities. Procedures/Tasks: If you allow your child to participate in this study, s/he will participate in three preference assessments and a reinforcer assessment. In the first preference assessment, nine items will be selected for inclusion in this assessment based on a survey (RAISD) filled out by you. The seven items will be aligned in an array in front of your child so that each item is an equal distance from your child. Your child will then be instructed to select one item by saying, pick one. If your child selects an item, they will be provided with 10 s access to that item. After 10 s, the item will be removed and the experimenter will randomly present the remaining items and repeat the above procedures. If your child does not make 75

a selection within 10 s, the instruction will be given one more time before that trial will be terminated. Following nine presentations of the full array, the items categorized at moderately preferred will be reassessed using the same procedures as above. The items categorized at low preferred will also be reassessed using the same procedures as above. The items that score the highest (determined by a systematic point weighting system) from each preference assessment will be selected to be used in the reinforce assessment. After completing the preference assessments, a reinforcer assessment will be conducted to determine the effectiveness of the items identified as highly preferred, moderately preferred, and low preferred to act as reinforcers and these items will be counterbalanced randomly across sessions. In this assessment, the experimenter will select a task that the participant performs inconsistently. These tasks will be selected either from your child‘s IEP or from classroom staff report. At the beginning of each trial, the experimenter will deliver prompt to your child. If they comply with the instruction, they will be given immediate access to the item (identified as highly preferred, moderately preferred, or low preferred) in the MSWO for 10 s or until the reinforcer is consumed. If they do not comply within 10 s, a hand over hand prompt will be provided to aid the participant in completing the task. Following completion of the task, access to the item (identified as highly preferred, moderately preferred, or low preferred) in the MSWO will be provided for 10 s or until the reinforcer is consumed. In addition to participating in the preference and reinforcer assessments, we will collect information from your child‘s educational file that is not publically available, including your child‘s disability and standardized assessment scores (where available). Duration: This study will last approximately 6 months. During this study, we expect to work with your child four to five days per week for twenty to thirty minutes per day. Your child may leave the study at any time. If you or your child decides to stop participation in the study, there will be no penalty and neither you nor your child will lose any benefits to which you are otherwise entitled. Your decision will not affect your future relationship with The Ohio State University. Risks and Benefits: We do not anticipate any risks as a result of participating in this study. Participants will be working with OSU students they are familiar with, so they should be comfortable in the study sessions. One potential risk is that the study is not successful in systematically identifying preferred items for individuals with significant intellectual and physical disabilities.

76

The main anticipated benefit of this study is that we will identify a means of systematically identifying preference and reinforcers for individuals with severe developmental disabilities. Knowing whether the identified items are actually preferred and whether or not they can act as reinforcers for this sample of students would be extremely beneficial. If we are successful, we will be able to provide systematic instruction AND be able to reinforce the new behaviors with things that are actually reinforcing to the student, rather than using something that we think might be something the student likes. Confidentiality: Efforts will be made to keep your child‘s study-related information confidential. However, there may be circumstances where this information must be released. For example, personal information regarding your child‘s participation in this study may be disclosed if required by state law. Also, your child‘s records may be reviewed by the following groups (as applicable to the research): 

Office for Human Research Protections or other federal, state, or international regulatory agencies;



The Ohio State University Institutional Review Board or Office of Responsible Research Practices;



The sponsor, if any, or agency (including the Food and Drug Administration for FDA-regulated research) supporting the study.

Incentives: There are no incentives for participating in this study. Participant Rights: You or your child may refuse to participate in this study without penalty or loss of benefits to which you are otherwise entitled. If you or your child is a student or employee at Ohio State, your decision will not affect your grades or employment status. If you and your child choose to participate in the study, you may discontinue participation at any time without penalty or loss of benefits. By signing this form, you do not give up any personal legal rights your child may have as a participant in this study. An Institutional Review Board responsible for human subjects research at The Ohio State University reviewed this research project and found it to be acceptable, according to applicable state and federal regulations and University policies designed to protect the rights and welfare of participants in research.

77

Contacts and Questions: For questions, concerns, or complaints about the study you may contact Diane Sainato at 614-292-8709 or [email protected]. For questions about your child‘s rights as a participant in this study or to discuss other study-related concerns or complaints with someone who is not part of the research team, you may contact Ms. Sandra Meadows in the Office of Responsible Research Practices at 1-800-678-6251. If your child is injured as a result of participating in this study or for questions about a study-related injury, you may contact Diane Sainato at 614-292-8709 or [email protected].

78

Signing the parental permission form I have read (or someone has read to me) this form and I am aware that I am being asked to provide permission for my child to participate in a research study. I have had the opportunity to ask questions and have had them answered to my satisfaction. I voluntarily agree to permit my child to participate in this study. I am not giving up any legal rights by signing this form. I will be given a copy of this form.

Printed name of subject

Printed name of person authorized to provide permission for subject

Signature of person authorized to provide permission for subject

AM/PM Relationship to the subject

Date and time

Investigator/Research Staff I have explained the research to the participant or his/her representative before requesting the signature(s) above. There are no blanks in this document. A copy of this form has been given to the participant or his/her representative.

Printed name of person obtaining consent

Signature of person obtaining consent

AM/PM Date and time

79

Appendix C: Caregiver Questionnaire

80

The Reinforcer Assessment for Individuals with Severe Disabilities (RAISD) CHILD‘S NAME: ______________________________

DATE: _____________

NAME OF REPORTER: _________________________ The purpose of this structured interview is to get as much specific information as possible form the parent (or caregiver) as to what they believe would be useful reinforcers for the client. Therefore, this survey asks parents questions about categories of stimuli (e.g., visual, auditory, etc.). After the parent has generated a list of preferred stimuli, ask additional probe questions to get more specific information on his/her preferences and the stimulus conditions under which the object or activity is most preferred (e.g., What specific TV shows are his favorite? What does she do when she plays with a mirror? Does she prefer to do this alone or with another person?)

We would like to get some information on ___________________‘s preference for different items and activities.

1.

Some children really enjoy looking at things such as a mirror, bright lights, shiny objects, spinning objects, TV, etc. What are the things you think _____________________ most likes to watch? __ ____________________________________

RESPONSE TO PROBE QUESTIONS:

_______________________________

2.

Some children really enjoy different sounds such as listening sounds such as listening to music, car sounds, whistles, beeps, sirens, clapping, people singing, etc. What are the things you think ___________________________ most likes to listen to?

81

___________________________

RESPONSE TO PROBE QUESTIONS:

_________________________

3.

Some children really enjoy different smells such as perfume, flowers, coffee, pine trees, etc. What are the things you think _____________________________ most likes to smell?

___________________________________________

RESPONSE TO PROBE QUESTIONS:

_________________________________________

4.

Some children really enjoy certain foods or snacks such as ice cream, pizza, juice, graham crackers. McDonald‘s hamburgers, etc. What are the things you think ____________________ most likes to eat?

____________________________________

82

RESPONSE TO PROBE QUESTIONS:

____________________________

5.

Some children really enjoy physical play or movement such as being tickled, wrestling, running, dancing, swinging, being pulled on a scooter board, etc. What activities like this do you think ___________________________ most enjoys?

_____________________________________

RESPONSE TO PROBE QUESTIONS:

_____________________________________

6.

Some children really enjoy touching things of different temperatures, cold things like snow or an ice pack, or warm things like a hand warmer or a cup containing hot tea or coffee. What activities like this do you think ________________________ most enjoys?

_________________________________

RESPONSE TO PROBE QUESTIONS:

_________________________________

83

7.

Some children really enjoy feeling different sensations such as splashing water in a sink, a vibrator against the skin, or the feel of air blow on the face from a fan. What activities like this do you think __________________________ most enjoys?

_______________________________

RESPONSE TO PROBE QUESTIONS:

_______________________________________________

8.

Some children really enjoy it when others give them attention such a hug, a pat on the back, clapping, say ―Good job‖, etc. What forms of attention do you think _____________________ most enjoys?

_________________________________________

RESPONSE TO PROBE QUESTIONS:

____________________________________

9.

Some children really enjoy certain toys such as puzzles, toy cars, balloons, comic books, flashlights, bubbles, etc. What are _________________________‘s favorite toys or objects?

84

________________________________

RESPONSE TO PROBE QUESTIONS:

________________________________

10.

What are some other items or activities that _____________________ really enjoys?

After completion of the survey, select all the stimuli which could be presented or withdrawn contingent on target behaviors during a session or classroom activity (e.g., a toy could be presented or withdrawn, a walk in the park could not). Write down all of the specific information about each selected stimulus on a 3‖ x 5‖ index card (e.g., ―Having a female adult ready him the ―Three Little Pigs‘ story‖). Then have the parents select the top 16 stimuli and rank them using the cards. Then list the ranked stimuli below.

1.

___________________

9.

___________________

2.

___________________

10.

___________________

3.

___________________

11.

___________________

4.

___________________

12.

___________________

5.

___________________

13.

___________________

85

6.

___________________

14.

___________________

7.

___________________

15.

___________________

8.

___________________

16.

___________________

BC FORMS:RAISD-APEND.DOC: 03/2000

86

Appendix D: Data Sheet for Experimenter and IOA for Preference Assessments

87

Moderate Item 1

Date: 0123

Item 2 Item 3

Low Item 1

Date: 0123

0123

Item 2

0123

0123

Item 3

0123

88

Appendix E: Ranking Stimuli Data Sheet

89

Item 1: # of times selected Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Trial 6 Trial 7 Trial 8 Trial 9

Rank Order X9 X8 X7 X6 X5 X4 X3 X2 X1

= = = = = = = = =

Item 2: # of times selected Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Trial 6 Trial 7 Trial 8 Trial 9

Rank Order X9 X8 X7 X6 X5 X4 X3 X2 X1

= = = = = = = = =

Item 3: # of times selected Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Trial 6 Trial 7 Trial 8 Trial 9

Rank Order X9 X8 X7 X6 X5 X4 X3 X2 X1

= = = = = = = = =

90

Item 4: # of times selected Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Trial 6 Trial 7 Trial 8 Trial 9

Rank Order X9 X8 X7 X6 X5 X4 X3 X2 X1

= = = = = = = = =

Item 5: # of times selected Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Trial 6 Trial 7 Trial 8 Trial 9

Rank Order X9 X8 X7 X6 X5 X4 X3 X2 X1

= = = = = = = = =

Item 6: # of times selected Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Trial 6 Trial 7 Trial 8 Trial 9

Rank Order X9 X8 X7 X6 X5 X4 X3 X2 X1

= = = = = = = = =

91

Item 7: # of times selected Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Trial 6 Trial 7 Trial 8 Trial 9

Rank Order X9 X8 X7 X6 X5 X4 X3 X2 X1

= = = = = = = = =

Item 8: # of times selected Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Trial 6 Trial 7 Trial 8 Trial 9

Rank Order X9 X8 X7 X6 X5 X4 X3 X2 X1

= = = = = = = = =

Item 9: # of times selected Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Trial 6 Trial 7 Trial 8 Trial 9

Rank Order X9 X8 X7 X6 X5 X4 X3 X2 X1

= = = = = = = = =

92

Appendix F: Procedural Integrity Checklist for Preference Assessment

93

Procedural Integrity; Preference Assessments IOA? Y N [circle one]

Data Collector______________ Student‘s Name______________

Exposes participant to each item for 10 s: Y N Line all stimuli in a linear array, approx. 5 cm apart and .3 m in front of participant: Y N

Item #:

1

2

3

4

5

6

7

8

9

Instructs participant to “pick one”

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Waits 10 s for student to make a choice

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

If selection is made, student receives 10 s access or allows student to consume edible

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Experimenter repeats instruction after 10 s if no selection is made

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

After selection, experimenter takes item on the far left and moves it the right end, repositioning the stimuli

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Selected item was not replaced

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

If “pick one” was stated twice and the student opts not to make a choice, the session is terminated

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

94

Appendix G: Procedural Integrity Checklist for Reinforcer Assessment; Baseline

95

Procedural Integrity Reinforcer Assessment; Baseline Data Collector: _________________ Student: _________________ Task: _________________________

A white placemat is used with the supplies needed.

Y N N/A

Directs student to sit down by stating, “Hi (student’s name). Come sit at the table.”

Y N N/A

Student is seated at the table with the placemat in front of him/her.

Y N N/A

Experimenter states, “Thanks for working with me today.”

Y N N/A

There are no reinforcement contingencies

Y N N/A

Discrete Trial #:

1

2

States the SD for the task.

Y N N/A

Y N N/A

Student responds correctly, no reinforcer is given

Y N N/A

Student responds incorrectly, no hand-over-hand prompt is provided

Y N N/A

96

3

4

5

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Appendix H: Procedural Integrity Checklist for Reinforcer Assessment

97

Procedural Integrity Reinforcer Assessment Data Collector: _________________ Student: _________________ Task: _________________________ Appropriate placemats: 1st phase (HP)cafeteria tray 2nd phase (MP)stone textured placemat 3rd phase (LP)red corduroy placemat The appropriate placemat is placed on the table, with the supplies needed.

Y N N/A

Directs student to sit down by stating, “Hi (student’s name). Come sit at the table.”

Y N N/A

Student is seated at the table with the placemat in front of him/her.

Y N N/A

Shows reinforcer to student and states “Thanks for working with me today. You will be earning _____ (shows item) if you do your work.”

Y N N/A

Discrete Trial #:

1

2

3

4

States the SD for the task.

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

If participant completes task, he/she is given 10 s access or allowed to consume reinforcer.

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

If he/she does not complete the task, experimenter provides a hand-over-hand prompt to aid the participant in completing the task, restating the SD. There are no additional prompts.

Y N N/A

Y N N/A

Y N N/A

Y N N/A

Y N N/A

After five trials, states, “We are finished. Thanks for working with me.”

Y N N/A

98

5

Appendix I: Social Validity Questionnaire

99

1. How satisfied are you with the changes in the students’ behavior regarding the assigned tasks (i.e., Luke: cutting paper, Andy: matching letters to name)? 1 Not at all satisfied

2

3

Somewhat not

4

Neutral

5

Somewhat satisfied

Very Satisfied

Satisfied

2. How effective do you believe the preference assessments were in identifying reinforcers? 1 Not at all effective

2

3

Somewhat not

4

Neutral

5

Somewhat Effective

Very Effective

Effective

3. How likely are you to use preference assessments in your classroom? 1

2

3

Not at all likely Somewhat not

4 Neutral

5

Somewhat likely

Very likely

likely

4. How likely are you to recommend preference assessments? 1

2

Not at all likely Somewhat not

3

4 Neutral

5

Somewhat likely

Very likely

likely

5. General overall comments regarding the preference assessments and the students’ tasks:

________________________________________________ ________________________________________________ ________________________________________________ __________________ 100