An Evaluation of Choice on Instructional Efficacy and Individual Preferences Among Children With Autism PDF

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2016, 49, 49, 1–6 170–175 NUMBER 1 (SPRING) AN EVALUATION OF CHOICE ON I...

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2016, 49, 49, 1–6 170–175 NUMBER 1 (SPRING) AN EVALUATION OF CHOICE ON INSTRUCTIONAL EFFICACY AND INDIVIDUAL PREFERENCES AMONG CHILDREN WITH AUTISM KAREN A. TOUSSAINT UNIVERSITY OF NORTH TEXAS TIFFANY KODAK UNIVERSITY OF WISCONSIN–MILWAUKEE AND JASON C. VLADESCU CALDWELL UNIVERSITY The current study compared the differential effects of choice and no-choice reinforcement conditions on skill acquisition. In addition, we assessed preference for choice-making opportunities with 3 children with autism, using a modified concurrent-chains procedure. We replicated the experiment with 2 participants. The results indicated that choice-making opportunities increased treatment efficacy for 2 of the 3 participants, and all 3 participants demonstrated a preference for choice-making opportunities. Key words: autism, choice, concurrent-chains arrangement, instructional efficacy, skill acquisition Providing students with choice-making oppor- disruptive behavior during discrete-trial instruction tunities appears to be beneficial during instruc- for children with autism. Newman et al. either tional programs, because it may result in decreases allowed participants to select the task order and in problem behavior and increases in academic reinforcer at the beginning of an instructional engagement (Dunlap et al., 1994). In addition, session (i.e., choice) or the experimenter selected research suggests that the provision of choice both the task and reinforcer (i.e., no choice). increases the frequency of academic responding Results demonstrated no difference in instruc- (Tiger, Toussaint, & Roath, 2010). tional efficacy. However, the effects of choice on Although the provision of choice has been acquisition may be more pronounced by providing shown to increase academic engagement, it has not opportunities to choose during the consequence been demonstrated to improve skill acquisition. In portion of instruction rather than as an antecedent an attempt to address this question, Newman, manipulation (Fenerty & Tiger, 2010). Needelman, Reinecke, and Robek (2002) eval- Thus, one purpose of the current study was to uated whether providing choice-making oppor- evaluate the effects of choice-making opportu- tunities influenced the rate of skill acquisition and nities when they are provided as a consequence. The current analysis also was designed to replicate and extend previous findings that demonstrate The design and data-collection portions of this research preschool children with and without develop- were conducted while the investigators were at the University of Nebraska Medical Center’s Munroe-Meyer mental disabilities prefer choice-making oppor- Institute. tunities (Brigham & Sherman, 1973; Fisher, Address correspondence to Karen Toussaint, Department Thompson, Piazza, Crosland, & Gotjen, 1997; of Behavior Analysis, University of North Texas, Denton, Texas 76203 (e-mail: [email protected]). Thompson, Fisher, & Contrucci, 1998; Tiger, doi: 10.1002/jaba.263 Hanley, & Hernandez, 2006). 1 170 2 KAREN A. EFFECTS OF CHOICE ONTOUSSAINT et al. PREFERENCE EFFICACY AND 171 METHOD total number of trials and converting the result to Participants a percentage. Mean agreement for unprompted correct responses was 99% (range, 83% to 100%) Three preschool-aged children with autism, for Samuel and 100% for both Ethan and Patrick who attended a university-based early intervention during the first evaluation. Initial-link selection program, participated in the evaluation. Samuel agreement was 100% across all participants. was a 5-year-old boy who used simple sentences to A second observer also collected procedural mand and frequently engaged in simple social integrity data for a minimum of 22% of sessions. exchanges. Ethan was a 4-year-old boy who used We calculated procedural integrity for each short phrases to label and request items; he session by dividing the number of correctly engaged in limited social conversation. Patrick was implemented trials by the total number of trials a 3-year-old boy who used a combination of short and converting this ratio to a percentage. Mean phrases and simple sentences to engage in tacting, procedural integrity score was 99.4% (range, manding, and social exchanges. 90% to 100%) for Samuel, 100% for Ethan, and 99.8% (range, 95% to 100%) for Patrick. Settings and Materials We conducted all sessions in the participants’ Preference Assessments typical learning environment, a room (5 m by We conducted a paired-item preference assess- 5 m) that contained a child-sized table, chairs, ment (Fisher et al., 1992) with small edible items and relevant session materials. We conducted (e.g., M&Ms) and larger items (e.g., Cheetos) three sessions daily for 3 to 4 days per week. The broken into small pieces, approximately 1 cm in child sat at the table, and the experimenter sat diameter. We used the top three items associated next to the participant. A secondary observer sat with the highest selection percentage. We also behind and to one side of the participant during conducted a preference assessment for colors and some sessions. used the three colors associated with the lowest selection percentage to serve as initial-link Measurement, Interobserver Agreement, and stimuli in the concurrent-chains arrangement. Procedural Integrity Observers recorded session data on a trial- Procedure and Design by-trial basis. We collected data on unprompted We evaluated the effect of contingent choice on correct responses, defined as vocalization of the correct responses using a nonconcurrent multiple targeted response before the delivery of the model baseline design embedded within a multiple prompt. Observers also collected data on initial- schedule design. Instructional stimuli previously link selections, defined as participants touching a assigned to the control condition served as card, during the concurrent-chains arrangement. acquisition targets during the preference phase A second observer independently recorded of the evaluation. We conducted the experiment data for a minimum of 32% of all sessions. twice with Samuel and Patrick for replication Agreement for unprompted correct responses purposes. Ethan was no longer enrolled in the was scored if both observers independently program at the time of replication. recorded the same target responses in a trial. There were six targets in each condition in the Agreement for initial-link selections was scored if initial evaluation, and each target was presented both the experimenter and the second observer twice. For Patrick and Samuel, targets consisted independently recorded the same initial-link of intraverbal responses to common “wh–?” selection. We calculated interobserver agreement questions and questions about functions of by dividing the number of agreements by the objects. Ethan’s targets were tacting pictures of 172 KAREN A. EFFECTS OF CHOICE ON TOUSSAINT et al. PREFERENCE EFFICACY AND 3 common objects. During replication of the Choice. The experimenter physically guided evaluation, we included 20 targets that were the participant to touch the relevant initial-link presented once in each condition; these targets stimulus and then presented the antecedent consisted of tacting pictures of popular people or stimulus. After a correct response, the experi- characters (e.g., George Washington, Scooby menter presented an array of three edible items Doo). We attempted to equate targets across and prompted the participant to select one (the conditions by creating sets that contained targets experimenter blocked attempts to select multiple with a similar number of syllables. items if this occurred). The experimenter provided During the treatment efficacy portion of the instructions using a progressive prompt delay to a evaluation, we conducted a series of three sessions vocal model prompt (e.g., the experimenter said, per day. We set the mastery criterion at two “spoon”), and we set prompt delays at 0 s, 2 s, 5 s, consecutive sessions with correct, unprompted 7 s, and 10 s. All participants first experienced one responses at or above 90%. We presented session with trials at a 0-s prompt delay, and the conditions in a pseudorandom and counterbal- second session included trials at a 2-s delay. We anced order with one exception; all participants increased the prompt delay (5 s, 7 s, 10 s) during experienced a choice session before a no-choice the subsequent session if at least 50% of session during the first series of the evaluation. We unprompted incorrect responses were errors of yoked the reinforcers selected by the participant omission in an instructional session. The experi- in the choice condition to the subsequent no- menter delivered reinforcement after both un- choice condition. As a result, it is possible that the prompted and prompted correct responses until selections made in a choice session were yoked to the participant demonstrated two consecutive two subsequent no-choice sessions if a no-choice sessions with at least 50% unprompted correct session randomly preceded a choice session in a responses. Thereafter, reinforcement was provided given series. We arranged for at least 10 min only for unprompted correct responses. Instruc- between each condition to account for satiation tion continued until performance reached mastery associated with food items. We paired each criterion. instructional condition with a colored index No choice. This procedure was identical to the card, which served as the discriminative stimulus choice condition except that after a correct during the instructional efficacy evaluation. response, the experimenter delivered an edible Colored cards later served as initial links of a item that was yoked to selections in the previous concurrent-chains arrangement in the instruc- choice condition. In the event that there were tional preference evaluation. Previous research more reinforced trials in a no-choice session than suggests that an effective method for identifying in the previous choice session (due to discontin- consumer preference for instructional conditions is uation of reinforcement for prompted respon- to measure selection among concurrently available ses after the preset criterion), the experimenter responses (selection of initial links) that determine repeated the delivery of items based on the the contingencies in place in the terminal link sequence of item selections. (choice, no choice, or no reinforcement; Hanley, Control. The control condition was identical Piazza, Fisher, Contrucci, & Maglieri, 1997). to baseline except that the therapist presented the relevant initial-link stimulus. Efficacy Evaluation Baseline. The experimenter presented the Preference Evaluation antecedent stimulus and allowed the participant During the subsequent instructional preference 5 s to respond. No consequences were provided evaluation, the experimenter placed the three for correct or incorrect responses. initial-link stimuli in front of the participant and 4 KAREN A. EFFECTS OF CHOICE ONTOUSSAINT et al. PREFERENCE EFFICACY AND 173 provided the instruction, “pick one.” After an low levels of correct responding during baseline initial-link selection, the experimenter presented sessions, and correct responding increased for all the antecedent stimulus, and responding resulted participants after implementation of the instruc- in access to the respective terminal-link con- tional procedures. Samuel demonstrated mas- tingencies for that session. tery-level responding in the choice condition in five training sessions compared to 14 training sessions in the no-choice condition. Differential RESULTS AND DISCUSSION efficacy was replicated, in that Samuel met The results of the differential efficacy portion mastery criterion in one fewer session in the of the evaluation are presented in the first two choice condition than in the no-choice condition panels of Figure 1. All participants demonstrated in his second evaluation. Ethan’s performances in Figure 1. The percentage of correct responses during baseline (first phase) and instructional conditions (second phase) for Samuel (first and second panels), Ethan (middle panel), and Patrick (fourth and bottom panels) during the instructional efficacy evaluation. The dashed line represents the mastery criterion. The percentage of unprompted correct responses and condition selection during the instructional preference evaluation are shown in the third phase. The number in parentheses denotes the first or second evaluation for Samuel and Patrick. 174 KAREN A. EFFECTS OF CHOICE ON TOUSSAINT et al. PREFERENCE EFFICACY AND 5 the choice and no-choice conditions were equal; Previous research has demonstrated that he met mastery criterion within an equal number individuals with typical and atypical development of choice and no-choice sessions. Patrick’s correct prefer choice-making opportunities (Tiger et al., responding reached the mastery criterion in three 2006). Providing choice of reinforcement is one fewer training sessions in the choice condition of many ways in which therapists may provide than in the no-choice condition in the first choice-making opportunities and promote per- evaluation and in four fewer training sessions in sonal liberties for individuals with autism and his second evaluation. related developmental disabilities (Bannerman, We saw a distinct advantage, in that two of Sheldon, Sherman, & Harchik, 1990). three participants (Samuel and Patrick) required fewer instructional sessions to reach the mastery criterion under choice conditions. Efficiency REFERENCES ranged from a savings of one to nine instructional Bannerman, D. J., Sheldon, J. B., Sherman, J. A., & sessions per instructional comparison, which Harchik, A. E. (1990). Balancing the right to may be viewed as relatively minor. However, this habilitation with the right to personal liberties: The practice may yield a substantial time savings if rights of people with developmental disabilities to eat too many doughnuts and take a nap. Journal of Applied implemented across several instructional pro- Behavior Analysis, 23, 79–89. doi: 10.1901/jaba.1990. grams over longer periods of time. 23-79 All participants demonstrated a preference for Brigham, T. A., & Sherman, J. A. (1973). Effects of choice and immediacy of reinforcement on single response choice-making conditions (Figure 1, third panel). and switching behavior of children. Journal of the Samuel showed an exclusive preference for the Experimental Analysis of Behavior, 19, 425–435. doi: choice condition. Ethan demonstrated preference 10.1901/jeab.1973.19-425 for the choice condition by selecting the choice Dunlap, G., dePerczel, M., Clarke, S., Wilson, D., Wright, S., White, R., & Gomez, A. (1994). Choice making to condition for seven of eight sessions. Patrick promote adaptive behavior for students with emo- exclusively selected the choice condition during tional and behavioral challenges. Journal of Applied both evaluations. His correct responding did not Behavior Analysis, 27, 505–518. doi: 10.1901/jaba. 1994.27-505 reach mastery criterion in the preference evalua- Fenerty, K. A., & Tiger, J. H. (2010). Determining tion phase of his replication evaluation because preschoolers’ preferences for choice-making opportu- his family relocated and he left the program nities: Choice of task versus choice of consequence. Journal of Applied Behavior Analysis, 43, 503–507. doi: midevaluation. 10.1901/jaba.2010.43-503 A yoking procedure was used in an attempt to Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L. P., isolate choice as a variable and control for item Owens, J. C., & Slevin, I. (1992). A comparison of two preference. However, it is possible that our approaches for identifying reinforcers for persons with severe and profound disabilities. Journal of Applied current results may still be attributed to the Behavior Analysis, 25, 491–498. doi: 10.1901/jaba. differential consequences associated with choos- 1992.25-491 ing. That is, a selected item in the choice Fisher, W. W., Thompson, R. H., Piazza, C. C., Crosland, K., & Gotjen, D. (1997). On the relative reinforcing condition may have served as the most preferred effects of choice and differential consequences. Journal item at the moment it was selected but not other of Applied Behavior Analysis, 30, 423–438. doi: 10.19 moments, such as when it was provided in the no- 01/jaba.1997.30-423 Hanley, G. P., Piazza, C. C., Fisher, W. W., Contrucci, S. A., choice condition (Fisher et al., 1997). One & Maglieri, K. A. (1997). Evaluation of client solution has been to use identical reinforcers in preference for function-based treatment packages. both conditions (Thompson et al., 1998). Journal of Applied Behavior Analysis, 30, 459–473. However, we elected to provide a choice among doi: 10.1901/jaba.1997.30-459 Newman, B., Needelman, M., Reinecke, D. R., & Robek, different items to mimic an arrangement that is A. (2002). The effect of providing choices on skill more likely to occur during clinical practice. acquisition and competing behavior of children with 6 KAREN A. EFFECTS OF CHOICE ONTOUSSAINT et al. PREFERENCE EFFICACY AND 175 autism during discrete trial instruction. Behavioral Tiger, J. H., Toussaint, K. A., & Roath, C. T. (2010). An Interventions, 17, 31–41. doi: 10.1002/bin.99 evaluation of the value of choice-making opportunities Thompson, R. H., Fisher, W. W., & Contrucci, S. A. in single-operant arrangements: Simple fixed- and (1998). Evaluating the reinforcing effects of choice in progressive-ratio schedules. Journal of Applied Behavior comparison to reinforcement rate. Research in Devel- Analysis, 43, 519–524. doi: 10.1901/jaba.2010.43-519 opmental Disabilities, 19, 181–187. doi: 10.1016/ S0891-4222(97)00050-4 Tiger, J. H., Hanley, G. P., & Hernandez, E. (2006). An evaluation of the value of choice with preschool Received November 13, 2013 children. Journal of Applied Behavior Analysis, 39, 1–16. Final acceptance July 9, 2015 doi: 10.1901/jaba.2006.158-04 Action Editor, Michael Kelley