|
| Variations/Enhancements of Established Preference Assessment Procedures |
| Saturday, May 29, 2004 |
| 4:00 PM–5:20 PM |
| Constitution B |
| Area: DDA; Domain: Applied Research |
| Chair: Richard B. Graff (New England Center for Children) |
| Abstract: Learning Objectives
The participant will be able to describe: the advantages of the points scoring method for the MSWO. one procedure for overcoming position biases during paired-stimulus preference assessments. the effects of the contingent use and restriction of stimuli on relative preferences on subsequent preference assessments, as described by Sidener, DeLeon, Christensen, & Wallace. |
| |
| An Alternate Scoring Method for the MSWO |
| FRANCIS J. CICCONE (New England Center for Children), Richard B. Graff (New England Center for Children), William H. Ahearn (New England Center for Children) |
| Abstract: DeLeon and Iwata (1996) described the difficulties in applying the percentage approach scoring method to the MSWO due the fact that with a perfectly consistent pattern of selection, an item ranked second can only be selected on 50% of trials. In the present study, 19 individuals diagnosed with autism, mental retardation or developmental delay, ages 14-22, participated. The results of 58 MSWO preference assessments were scored using both the traditional percentage approach method and a novel point weighting system. The dependent variable in measured with both scoring systems was approach, defined as the student passing the item through the plane of the lips. Interobserver agreement and procedural integrity data were collected during 46% of preference assessment sessions; mean IOA and PI was above 99%. When the preference assessments were scored using the percentage approach method, 21 items were identified as highly preferred across all assessments; however, when the assessments were scored using the point system, 61 items were identified as highly preferred. Several items that were scored as high preference using the point scoring system (high-pts) were scored as moderate- or low-preference on the percentage approach method (low-%). Reinforcer assessments indicated that high-pts/low-% stimuli functioned as reinforcers for participants. |
| |
| Methods for Eliminating Position Bias During Preference Assessment and Discrimination Training |
| STEPHEN T. NORTH (University of Florida), Brian A. Iwata (University of Florida) |
| Abstract: Selection responses for mast individuals during preference assessment or discrimination training typically are under stimulus control, in that the participant identifies and selects an item primarily based on physical properties of the stimulus presented. However, some individuals may habitually show a position bias, in which responses do not reflect control by the experimental stimuli, but, rather, by the locations in which these stimuli arc placed when they are presented. We evaluated the effects of several procedures for eliminating a position bias exhibited by a 41-year-old woman diagnosed with moderate mental retardation. Following baseline paired-stimulus preference assessments in which the participant exclusively selected items presented on her left side, the quality and magnitude of items included in the preference assessment were systematically manipulated in an effort to facilitate discrimination during subsequent assessments and generalization probes. The inclusion of nonpreferred items (identified via single-stimulus preference assessments) enhanced discrimination in the training context (i.e., the participant made right-sided selections to avoid nonpreferred items) but did not produce the same result during ongoing baseline preference assessments. The magnitude manipulation produced increases in discrimination in both training and baseline preference assessment sessions, as well as during discrimination test probes. |
| |
| The Influence of Contingent Use and Restriction of Reinforcers on Changes in Relative Preferences |
| DAVID W. SIDENER (Kennedy Krieger Institute), Iser Guillermo DeLeon (Johns Hopkins University School of Medicine), Amy L. Christensen (University of Nevada, Reno), Michele D. Wallace (University of Nevada, Reno) |
| Abstract: The identification of preferred stimuli is a critical component of successful training and treatment programs for individuals with developmental disabilities. The success of these programs also hinges on the durability of reinforcers over time. Although a small number of studies have examined general changes in stimulus preferences over time, none have directly explored the variables that influence the long-term durability and/or stability of reinforcer value. In the current study, we examined the role of repeated, contingent use on changes in relative preferences among stimuli. Following initial preference assessments, stimuli from the middle of preference hierarchies were delivered contingently on a daily basis in skills acquisition programs. Other stimuli from the same assessment were restricted. Following a history of contingent use, paired-choice preference assessments were repeated to examine the effects of contingent use on changes in relative preference. Results for three individuals revealed that stimuli that were used contingently generally moved up in the preference hierarchy- By contrast, stimuli that were restricted showed unsystematic fluctuations in relative preference. Interobserver agreement on selection responses during the assessments averaged 100%. These results are discussed in terms of their implications for the repeated, long-term use of reinforcers for this population. |
| |
| Evaluating Preschool Children’s Preferences for Instructional Contexts |
| NICOLE HEAL (University of Kansas), Gregory P. Hanley (University of Kansas) |
| Abstract: Although providing care and safety for young children have been primary roles of preschool teachers, many teach a variety of academic skills. We used a concurrent chains procedure in the current study to evaluate the effectiveness of and preference for several instructional contexts with two preschool children. The relative effectiveness was determined by comparing levels of problem behavior and correct responding during terminal links (where the children experienced the contexts) while preference was determined by observing relative response rates in the initial links (where children chose the contexts). Interobserver agreement was collected for a minimum of 30% of sessions and mean agreement was 80% or higher for all measures. The instructional contexts differed along two empirically-determined dimensions involving the value of the instructional materials and consequences for correct responding. Three instructional contexts resulted: highly preferred materials and less preferred consequences (HP/LP), less preferred materials and highly preferred consequences (LP/HP), and less preferred materials and less preferred consequences (LP/LP; control). We found no difference in correct responding across instructional contexts, and slightly elevated rates of undesirable behaviors correlated in the LP/LP context for one participant. Nevertheless, both participants showed a preference for the use of LP/HP condition. |
|
| |