• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

The Use of Concurrent-Operants in Assessing Preference for Instructional Treatment in Children with Escape-Maintained Problem Behavior

ProQuest Dissertations and Theses, 2011
Dissertation
Author: Carrie Marie Brower-Breitwieser
Abstract:
Conducting preference assessments for behavioral treatments could add social validity to the treatments that are used, as the direct consumers of these treatments often are unable to report preference for treatment by traditional social validity methods (e.g., self-report). The purpose of the present study was to use a concurrent-operant preference assessment to evaluate preference for two instructional approaches, Functional Communication Training (FCT) and a reinforcement-based intervention that used a "work, then play" card. Participants included 11 children, ages 6 to 12 years, who were diagnosed with intellectual disabilities. First, a screening condition was conducted to evaluate whether the participants could discriminate between the two alternatives used throughout the experiments. In Experiment 1, participants chose between a treatment that varied in the reinforcement densities that were provided with each alternative (medium vs. high quality). Here, the same treatment and tasks were available at two desks, but the reinforcement for completing tasks varied in terms of quality and duration. In Experiment 2, participants were given choices between the "work, then play" card vs. FCT + the "work, then play" intervention. In this experiment, the reinforcer densities for each treatment were equal, so preference was based solely on the differences between the two treatments. It was hypothesized that in Experiment 1, participants would allocate greater responding to the desk with the greatest density of reinforcement. In Experiment 2, it was hypothesized that participants would allocate responding to the desk containing the FCT + "work, then play" intervention, as this approach allowed the participant to gain access to reinforcement through task completion, or to request a break (e.g., escape from task). For nine of the eleven participants, the concurrent operant preference assessment was able to detect participant preference for one of the two available treatment interventions. The preferences were robust and consistent, spanning more than one session. In the first experiment, preference was supported for interventions that lead to higher sources of reinforcement. In the second experiment, preference for interventions that allow for an opportunity for a break was assessed, and only some participants demonstrated this preference. Implications, limitations, and future directions for research are discussed.

PREFERENCE ASSESSMENTS TABLE OF CONTENTS PRELIMINARY PAGES: i List of Figures xi List of Tables xiii ABSTRACT xv CHAPTER ONE: REVIEW OF THE LITERATURE 1 Concurrent Operant Preference Procedures 3 Concurrent Schedules of Reinforcement in Applied Behavior Analysis 6 The Use of Concurrent Operants Procedures as an Alternative to Extinction-Base Procedures 14 Three-Choice Procedures 19 Purpose of Research 20 Research Questions 26 CHAPTER TWO: THE USE OF CONCURRENT-OPERANTS IN ASSESSING PREFERENCE FOR INSTRUCTIONAL TREATMENTS IN CHILDREN WITH ESCAPE-MAINTAINED PROBLEM BEHAVIOR 27 Introduction 27 Experiment 1 31 Method 31 Participants 31 Setting 32 Participant Materials 33 Data Collection Materials 35 Dependent Variables and Measurements 35 Choices 35 Appropriate Behaviors 35 Task Engagement 36 Socially Appropriate Behavior 36 Other Appropriate Behavior 36 Within-Session Problem Behaviors 37 Observers and Interobserver Agreement 37 Treatment Integrity 38 Procedures and Experimental Design 38 Discrimination Test 38 Phase 1 (Therapist guided) 39 vn

PREFERENCE ASSESSMENTS Phase 2 40 Phase 3 40 Preference Assessment 41 Phase 1 (Therapist guided) 41 Phase 2 42 Phase 3 43 Data Analysis 43 Results 43 Discrimination Test 43 Choice Data 45 In-session behavior data 48 Preference Assessment 50 Choice Data 51 In-session behavior data 54 Discussion 56 Experiment 2 59 Method 59 Participants 59 Setting 59 Participant 59 Materials 60 Dependent Variables and Measurements 61 Procedures and Experimental Design 61 Discrimination Test 61 Phase 1 (Therapist guided) 61 Phase 2 61 Phase 3 61 Preference Assessment 62 Phase 1 (Therapist guided) 62 Phase 2 62 Phase 3 62 Data Analysis 63 Results 63 Discrimination Test 63 Choice Data 64 In-session behavior data 66 Preference Assessment 67 Desk choice data 69 viii

PREFERENCE ASSESSMENTS Choices made at desk data 70 In-session behavior data 72 Parent/Teacher Questionnaire 73 Discussion 74 General Discussion 79 REFERENCES: 87 FIGURES Figure 1 33 Figure 2 34 Figure 3 45 Figure 4 46 Figure 5 48 Figure 6 49 Figure 7 51 Figure 8 52 Figure 9 54 Figure 10 55 Figure 11 60 Figure 12 61 Figure 13 64 Figure 14 65 Figure 15 66 Figure 16 67 Figure 17 69 Figure 18 70 Figure 19 71 Figure 20 72 Figure 21 73 TABLES Table 1 34 Table 2 61 Table 3 74 APPENDICES 94 Appendix A: Set-up of the screening condition 94 Appendix B: Set-up of Experiment 1 96 Appendix C: Set-up for Experiment 2 98 Appendix D: Consent Form 100 Appendix E: Script for Obtaining Informed Consent 106 Appendix F: Participant Data Sheets 109 IX

PREFERENCE ASSESSMENTS Appendix G: Treatment Integrity Data Sheets 113 Appendix H: Parent/Teacher Questionnaire 119 Appendix I: Interobserver Agreement for all Participants 121 Appendix J: Treatment Integrity for all Participants 133 x

PREFERENCE ASSESSMENTS LIST OF FIGURES Figure Page 1 Photograph depicting the desks used in the screening condition and preference assessment 33 2 Communication cards for behavior-consequence relations: first, "work", then, play 34 3 Percentage of choices made during the screening condition for Ellis, Lewis, and Sally 45 4 Percentage of choices made during the screening condition for Richard, Ben, and Austin 46 5 Percent of 10-s intervals engaged in behavior for Ellis, Lewis, and Sally 48 6 Percentage of 10-s intervals engaged in behavior for Richard, Ben, and Austin 49 7 Percentage of choices made during the preference assessment for Ellis, Lewis, and Sally 51 8 Percentage of choices made during the preference assessment for Richard, Ben, and Austin 52 9 Percent of 10-s intervals engaged in behavior for Ellis, Lewis, and Sally 54 10 Percentage of 10-s intervals engaged in behavior for Richard, Ben, and Austin..55 11 Sample "break" communication card 60 12 Photograph depicting the desks used in the preference assessment for Experiment 2 61 13 Percentage of choices made during the screening condition for Dalton, Kayla, and Erin 64 14 Percentage of choices made during the screening condition for Damon and Amy 65 15 Percent of 10-s intervals engaged in behavior for Dalton, Kayla, and Erin 66 16 Percentage of 10-s intervals engaged in behavior for Damon and Amy 67 17 Percentage of desk choices to the 3-choice desk 69 XI

PREFERENCE ASSESSMENTS 18 Percentage of choices made during the preference assessment for Dalton, Kayla and Eria 70 19 Percentage of choices made during the preference assessment for Amy and Damon 71 20 Percentage of intervals engaged in behavior for Dalton, Kayla, and Erin 72 21 Percentage of 10-s intervals engaged in behavior for Damon and Amy 73 xn

PREFERENCE ASSESSMENTS LIST OF TABLES Table Page 1 Individualized participant work tasks 34 2 Individualized participant work tasks 61 3 Parent and teacher predictions of the participant's preference for intervention....74 xm

PREFERENCE ASSESSMENTS ABSTRACT Conducting preference assessments for behavioral treatments could add social validity to the treatments that are used, as the direct consumers of these treatments often are unable to report preference for treatment by traditional social validity methods (e.g., self-report). The purpose of the present study was to use a concurrent-operant preference assessment to evaluate preference for two instructional approaches, Functional Communication Training (FCT) and a reinforcement-based intervention that used a "work, then play" card. Participants included 11 children, ages 6 to 12 years, who were diagnosed with intellectual disabilities. First, a screening condition was conducted to evaluate whether the participants could discriminate between the two alternatives used throughout the experiments. In Experiment 1, participants chose between a treatment that varied in the reinforcement densities that were provided with each alternative (medium vs. high quality). Here, the same treatment and tasks were available at two desks, but the reinforcement for completing tasks varied in terms of quality and duration. In Experiment 2, participants were given choices between the "work, then play" card vs. FCT + the "work, then play" intervention. In this experiment, the reinforcer densities for each treatment were equal, so preference was based solely on the differences between the two treatments. It was hypothesized that in Experiment 1, participants would allocate greater responding to the desk with the greatest density of reinforcement. In Experiment 2, it was hypothesized that participants would allocate responding to the desk containing the FCT+ "work, then play" intervention, as this approach allowed the participant to gain access to reinforcement through task completion, or to request a break (e.g., escape from task). For nine of the eleven participants, the concurrent operant preference assessment xiv

PREFERENCE ASSESSMENTS was able to detect participant preference for one of the two available treatment interventions. The preferences were robust and consistent, spanning more than one session. In the first experiment, preference was supported for interventions that lead to higher sources of reinforcement. In the second experiment, preference for interventions that allow for an opportunity for a break was assessed, and only some participants demonstrated this preference. Implications, limitations, and future directions for research are discussed. xv

PREFERENCE ASSESSMENTS CHAPTER 1 REVIEW OF THE LITERATURE In the field of applied behavior analysis, research has been conducted that uses formal preference assessment methodology to examine participant preference for specific stimuli that are used in behavioral interventions. The ability to evaluate preference for stimuli is important, as being able to increase the quality of reinforcement may be a fundamental component of the efficacy of a behavioral intervention. Determining a participant's preferred stimuli provides researchers with the knowledge necessary to be able to change behavior effectively. There are several methods for assessing preference for stimuli. One technique that is commonly used is an interview of caregivers of the individual. These interviews typically involve asking the caregiver questions about items or activities that the participant likes and dislikes. This technique has its limitations, as it relies on conjectural statements from another about an individual's preference, which may be inaccurate (Wolfe, 1978). Furthermore, when the participant is an individual diagnosed with an intellectual disability, research suggests that there is poor correspondence between the results of interviews with caregivers and direct preference assessments (Green et al., 1988). There are also techniques that involve more direct measurement of the participants' preferences. This may involve asking the individuals directly about preferred reinforcers. Under these conditions, a reinforcer inventory form is used. The form is a measure that provides a list of potential reinforcers in which participants are asked to rate each reinforcer on some scale (e.g., 3-point rating scale with ratings of 1

PREFERENCE ASSESSMENTS dislike, neutral, like), and the highest rated reinforcer is the one chosen as the reinforcing stimulus for the treatment (Miltenberger, 2008). This approach, however, is limited when working with participants who are non-verbal, as these participants may not readily be able to report preferred stimuli through these indirect measures. A single-stimulus preference assessment (Pace et al., 1985) is one example of a direct preference assessment technique used in the applied field that allows non-verbal individuals to provide information about what might be effective reinforcers. A single- stimulus preference assessment in an approach-based procedure involves presenting an individual with a series of single stimuli, and each is presented individually. Observers record whether or not the individual approaches each stimulus, and approach to each stimulus is considered a measure of preference. A limitation of single-stimulus preference assessments is that some individuals approach all of the stimuli presented, thus providing no evidence about the rank-order preference of the presented stimuli (Hagopian et al, 2001). One of the most frequently used direct preference assessment methods was developed by Fischer et al (1992). This method of assessing preference is known as a paired stimulus (PS) preference assessment, and it involves simultaneously presenting potential reinforcers in a forced-choice situation and requiring the participant to choose between them. Each potential reinforcer is presented multiple times against different potential reinforcers. This preference assessment procedure differentiates between highly preferred and less-preferred stimuli, and provides the researcher with a rank order of preferred items. Highly preferred stimuli are those that are most frequently selected. 2

PREFERENCE ASSESSMENTS Research exists that compares the paired-stimulus preference assessment to other preference assessment techniques (Windsor, Piche, & Locke, 1994). Windsor and colleagues compared Fischer et al's (1992) compared stimulus preference assessment to a technique called multiple stimulus (MS) preference assessment, in which participants approach a set of multiple stimuli (instead of two), and choose one item during each trial. Each preference assessment consists of 10 20-s trials. The results of the Windsor et al. (1994) study suggest that while both procedures identified the same stimuli as preferred, the MS procedure was a more efficient method, as it required less time to complete. In response to these findings, DeLeon and Iwata (1996) developed a preference assessment technique that incorporates the strengths of the PS (Fischer et al.) and MS preference (Windsor et al.) assessment techniques. This technique, the multiple-stimulus without replacement (MSWO) involves presenting multiple stimuli from trial to trial, but once an item is selected by an individual, it is not replaced in the subsequent trial. The results of the DeLeon and Iwata study suggest that the MSWO procedure identifies more reinforcers than the MS procedure. Furthermore, this technique also provides a rank- order of preferences in less time than is required to complete a PS preference assessment. Concurrent-Operant Preference Procedures The concurrent operant preference assessment builds on the PS and the MS methods. In a traditional concurrent-operants assessment, two or more different responses are possible for an individual to perform, and each response is associated with an independent schedule of reinforcement or different reinforcing outcome (Hanley, Piazza, Fisher, Contrucci, & Maglieri, 1997). Thus, allocation of responses (i.e., choice) to each outcome is a sensitive measure of the individual's preference for concurrently 3

PREFERENCE ASSESSMENTS available reinforcers (Fisher & Mazura, 1997), in that choices with higher response allocation (i.e., percentage of responses) indicate preference. The matching law is a well-documented principle of behavior that has been developed from almost 50 years of basic and applied research with non-human and human research (e.g., de Villiers, 1977; see Fisher & Mazur, 1997, for a review; Graff & Libby, 1999; Hermstein, 1961, 1964,1970). The matching law, which uses a concurrent- operants arrangement, states that when an individual is presented with two or more choices at the same time, they will choose the one that has the highest rate of reinforcement in relation to the total amount of reinforcement offered by both options (Hermstein, 1964), and the allocation of behavior will be in direct proportion to the reinforcement derived from each response (Hermstein, 1964). For example, if two reinforcers that were equally valued by a child were delivered in different quantities (3 M&Ms for choice A vs. 1 M&M for choice B), the child would likely allocate 75% of its responses to choice A, because of every 4 reinforcers, three (or 75%) come from choice A. This theory generalizes also to those operants that produce not only the highest rate of reinforcement, but also the longest duration, quality, and/or immediacy of reinforcement rate (Hermstein, 1961, 1970; McDowell, 1988; Myerson & Hale, 1984; Pierce & Epling, 1995). There is a wealth of research on the matching law in the basic behavior analysis field with nonhuman animals (Mace, 1994). According to Davison and McCarthy's (1998) review of matching law research, the majority of research in the basic field has involved examining the effect of various schedule-controlled parameters on response allocation, such as reinforcer frequency. In the applied field, other schedule parameters, 4

PREFERENCE ASSESSMENTS such as reinforcer quality, immediacy of reinforcement, and response effort in natural environmental settings (Neef, Mace, Shea, & Shade, 1992) may affect behavioral allocation more than relative frequency of reinforcement, most likely because quality of the reinforcer is more salient than frequency. According to Mace (1994), in applied work, it appears to be the combination of rate of reinforcement, along with these other variables, affects the choices humans make in concurrent-operant paradigms. Understanding the differences between basic and applied matching law research findings provide applied researchers with the knowledge to develop effective behavioral interventions that involve arranging contingencies to bias responding to adaptive behaviors instead of aberrant behaviors. Research in the applied field has indicated that matching is an effective way to assess an individual's preference for different schedules of reinforcement, types of reinforcement (e.g., quality) (Catania & Sagvolden, 1980; Hanley, et al., 1997; Hermstein, 1964), or for different approaches to treatment (Hanley et al., 1997). There are several advantages of using a concurrent-operants arrangement. One main advantage of using this arrangement over less objective preference assessment methods (e.g., self- report) is that it allows the experimenter to study an individual's choice of one operant over another at different periods of time (Fisher & Mazur, 1997). Self-report assessments can be confounded due to the social contingencies provided by the experimenter (i.e., raters may "fake good" or "fake bad"; Schwartz & Baer, 1991). A concurrent-operants arrangement allows an experimenter to observe a participant's behavior without the need to rely on indirect consumers for information. Another advantage of using a concurrent- operants arrangement is that it allows for a much more sensitive means of examining the 5

PREFERENCE ASSESSMENTS preference for one reinforcer over another, as multiple responses, instead of one discrete choice, is used as a measure (Fisher & Mazur, 1997). Concurrent Schedules of Reinforcement in Applied Behavior Analysis The use of concurrent operants and matching theory has been applied to research with humans in an attempt to understand preference for particular learning environments. Some studies, for example, use a concurrent-operant arrangement with typically developing populations to examine preference for academic tasks (e.g., Neef, Mace, Shea, & Shade, 1992) or in evaluating preference for teaching styles (e.g., Heal & Hanley, & Layer, 2009). Others have used the concurrent operant procedure to develop treatments designed to decrease problem behavior. Using the theoretical framework provided by matching theory, researchers can develop behavioral interventions that increase response allocation to appropriate alternative behaviors, and decrease allocation to inappropriate behaviors. Piazza, Fisher, Hanley, Remick, Contrucci, and Aitken (1997), for example, used a concurrent operant arrangement to examine the effects of providing reinforcement contingencies for three participants who displayed escape-maintained problem behaviors (engaged in inappropriate behavior because in the past, they were reinforced by escaping the task). This study examined the effects of three reinforcers contingent on compliance to various academic tasks (e.g., completing math problems). The reinforcers for completing tasks were: 1) access to preferred stimuli with experimenter attention, 2) taking a break from tasks with no experimenter attention, or 3) a combination of breaks and preferred stimuli with experimenter attention. Participants who engaged in task compliance (one response option) received a 30-s break with highly preferred tangible 6

PREFERENCE ASSESSMENTS items and experimenter attention. Participants who engaged in problem behavior (a second response option) received a 30-s break without access to tangible items or experimenter attention. Results showed that participants allocated responding in favor of appropriate behavior (task compliance) instead of problem behavior (e.g., throwing task materials) when task breaks include preferred stimuli. These results indicated that when preferred stimuli were combined with a task break, this combination of reinforcers appear to compete with the previously negatively reinforced problem behavior. Harding et al. (1999) also used a concurrent operant arrangement design to examine the effects of reinforcement on choice behavior for participants who demonstrated escape-maintained problem behavior. The researchers in this study allowed participants to allocate their time to one of two concurrently available choices. Problem behavior was placed on extinction. Five different conditions were conducted during this concurrent operant assessment that pitted the presence or absence of parent attention with and without instmctions (e.g., "play my way"), preferred stimuli, or neutral stimuli. Participants were given a choice between two desks. At each desk, there was some type of reinforcer present. For example, Desk 1 may have included parent attention without instmctions and preferred stimuli, and Desk 2 may have included parent attention without instmctions and neutral stimuli. Results indicated that the participants allocated responding to receive parent attention, regardless of how highly preferred the stimuli associated with parent attention was—i.e., parent attention was the aspect of the reinforcer that was most salient and preferred. However, when the parent attention involved the parent giving instmctions, the participants allocated responding to receive parent attention only when the stimuli present were highly preferred. If the option with 7

PREFERENCE ASSESSMENTS parent attention contained neutral stimuli, the participants allocated responding to the option with no parent attention and highly preferred stimuli. This study shows that a reinforcer's efficacy can change with regard to what other alternatives are concurrently available, meaning that a reinforcer's efficacy is relative. Concurrent operant arrangements have also been used to assess choice for preferred properties of the learning environment in children with special education needs. Mace and colleagues (1994) used a concurrent-operant arrangement to assess participant choice for sets of math problems. The researchers provided three adolescent special education students with a choice between completing one of two sets of math problems. The researchers provided three different concurrent schedules arrangements with reinforcer ratios (nickels) that varied (in the first condition, the ratios paid off 2:1; in the other two conditions, 6:1, and 12:1). Similar to other applied research with concurrent schedules of reinforcement, participants allocated responding for sets of math problems in direct response to the reinforcement available for each set. The results of this study showed that undermatching was observed for all participants. Undermatching occurs when the proportion of time an individual spends on a richer schedule of reinforcement is less than the proportion of reinforcers derived from that schedule. In other words, individuals allocate fewer responses than the reinforcement ratios would predict. Furthermore, according to the authors, the participants in this study showed consistent preference for one of the two response alternatives available, independent of the schedules correlated rate of reinforcement. More specifically, when changes were made to the schedules (e.g., schedules were thinned), results suggested that participants persisted in their original response allocation and did not alter response allocation. The 8

PREFERENCE ASSESSMENTS findings of this study may suggest that concurrent-operant arrangements in which relative frequency of reinforcement is manipulated may not be well suited or efficient for use with individuals diagnosed with intellectual disabilities. Results from the aforementioned studies generally support the use of concurrent operants procedures for assessing participant preference for stimuli. One interesting application of these procedures may be used to assess preference for a particular behavioral intervention. Such analyses would allow behavior analysts to increase participant involvement in the planning and evaluating of their behavioral interventions, instead of relying on parent or teacher ratings of acceptability— an important benefit for behavior analysts (Schwartz & Baer, 1991). The following section of this proposal will describe a method for incorporating participant involvement in assessing social validity of behavioral interventions for participants who are unable to utilize the methods (e.g., self-report, interview, etc.) described earlier in this proposal. In a study conducted by Hanley et al. (1997), a concurrent-operant preference assessment was used to evaluate preference for functional communication training (FCT) or non-contingent reinforcement (NCR) as a behavioral intervention for reducing problem behavior in children in an inpatient setting that engaged in severe dismptive behaviors. Functional Communication Training (FCT; Carr & Durand, 1985) is an intervention developed to reduce escape-maintained problem behaviors exhibited by individuals diagnosed with intellectual disabilities. Using FCT, an individual is taught to emit an appropriate behavior to obtain a reinforcer (e.g., break). Non-contingent reinforcement involves delivering the reinforcer identified for maintaining the behavior of interest (e.g., problem behavior) on a time-based schedule, independent of an 9

PREFERENCE ASSESSMENTS individual's responding (Hagopian, Fischer, & Legacy, 1994). In other words, the participant's behavior has no control over the delivery of reinforcement. Instead, after a set period of time (e.g., 30-s) a reinforcer is delivered. In Hanley et al's study, training sessions were conducted with each participant for the purpose of exposing him/her to the different contingencies that were associated with each operant (displacing two different switches). The participant could press a switch, which resulted in 2 minutes of the contingency associated with the chosen response. After the 2-minute treatment session was completed, the participant was once again prompted to choose a switch. After the completion of the training session, the treatment preference session was conducted for a duration of 20 minutes. Results indicated that both participants had a higher relative response rate for FCT than NCR, thus suggesting that the participants preferred the ability to control the rate of reinforcement during treatment. Some research (Brower-Breitwieser, Miltenberger, Gross, Orizondo-Korotko, Fuqua, and Breitwieser, 2008) may suggest that a concurrent-chains operant preference assessment may not be an effective tool for determining preference for treatments for children diagnosed with autism. Brower-Breitwieser et al. (2008) evaluated the use of a concurrent-operant preference assessment for determining preference between two training programs frequently used with children diagnosed with autism: Discrete Trials Training and the Treatment and Education of Autistic and Communication Handicapped Children (Project TEACCH). Discrete Trial Training (DTT) is an applied behavior analysis (ABA) procedure that is commonly used in home and school-based programs with children diagnosed with autism. Discrete Trial Training uses stimulus control, and 10

PREFERENCE ASSESSMENTS reinforcement and punishment contingencies to teach individuals a variety of skills such as imitation, language acquisitions, and educational skills. Treatments are individualized to each participant's unique behavioral problem. Project TEACCH is a training approach developed by Schopler (Tissot & Evans, 2003) that uses environmental stimuli to cue participants about the expectations within their environment. It incorporates non- transient visual cues (picture schedules, social stories, or flashcards) which are necessary for the child to organize and conceptualize the stimuli, but does not utilize a one-on-one intervention style often seen in a discrete trials training. Participants in the Brower-Breitwieser et al. study were all children diagnosed with autism who had a history of receiving both Discrete Trials Training and Project TEACCH interventions. Participant preference for intervention was assessed using a concurrent-operant preference assessment. The participants were positioned on a "choice line" (a taped line on the floor equidistant between the two treatment rooms) and were instructed to make a choice between the two therapy rooms. The therapy rooms were identical to one another and the same task was completed in each room (e.g., sorting colors), thus the only difference between the two rooms was the treatment approach used. The results of this study showed that the three participants did not show a clear preference for either training program, despite the fact that the Discrete Trials Training program had a higher rate of reinforcement provided than the Project TEACCH protocol. The overall percentage of ABA choices was slightly higher than the percentage of Project TEACCH choices (53.6% vs. 46.4%). Furthermore, the results of that study suggested that the participants engaged in "switching" or alternating between operants, a 11

Full document contains 161 pages
Abstract: Conducting preference assessments for behavioral treatments could add social validity to the treatments that are used, as the direct consumers of these treatments often are unable to report preference for treatment by traditional social validity methods (e.g., self-report). The purpose of the present study was to use a concurrent-operant preference assessment to evaluate preference for two instructional approaches, Functional Communication Training (FCT) and a reinforcement-based intervention that used a "work, then play" card. Participants included 11 children, ages 6 to 12 years, who were diagnosed with intellectual disabilities. First, a screening condition was conducted to evaluate whether the participants could discriminate between the two alternatives used throughout the experiments. In Experiment 1, participants chose between a treatment that varied in the reinforcement densities that were provided with each alternative (medium vs. high quality). Here, the same treatment and tasks were available at two desks, but the reinforcement for completing tasks varied in terms of quality and duration. In Experiment 2, participants were given choices between the "work, then play" card vs. FCT + the "work, then play" intervention. In this experiment, the reinforcer densities for each treatment were equal, so preference was based solely on the differences between the two treatments. It was hypothesized that in Experiment 1, participants would allocate greater responding to the desk with the greatest density of reinforcement. In Experiment 2, it was hypothesized that participants would allocate responding to the desk containing the FCT + "work, then play" intervention, as this approach allowed the participant to gain access to reinforcement through task completion, or to request a break (e.g., escape from task). For nine of the eleven participants, the concurrent operant preference assessment was able to detect participant preference for one of the two available treatment interventions. The preferences were robust and consistent, spanning more than one session. In the first experiment, preference was supported for interventions that lead to higher sources of reinforcement. In the second experiment, preference for interventions that allow for an opportunity for a break was assessed, and only some participants demonstrated this preference. Implications, limitations, and future directions for research are discussed.