|
| Variations of Functional Analysis Methodology |
| Sunday, May 24, 2009 |
| 3:00 PM–4:20 PM |
| North 128 |
| Area: DDA; Domain: Applied Behavior Analysis |
| Chair: Jeff Tiger (Louisiana State University) |
| Abstract: Functional analysis is a widely researched behavioral assessment procedure used to identify the operant mechanisms responsible for the maintenance of problem behavior. There have been a number of procedural variations and refinements described in the published literature since the initial publication describing this assessment technique in 1982. The 4 talks in this symposia evaluate some of the variations that may result in more rapid and accurate determinations of behavioral function. |
| |
| Evaluation of false positive (tangible) functional analysis outcomes |
| GRIFFIN W. ROOKER (University of Florida), Brian A. Iwata (University of Florida), Erin Camp (University of Florida) |
| Abstract: Functional analysis (FA) methodology has been extremely effective in identifying the maintaining variables for problem behavior. However, the results of a FA may be influenced by idiosyncratic sensitivities to aspects of the assessment conditions. For example, Shirley, Iwata, and Kahng (1999) demonstrated a false positive FA outcome associated with exposure to the tangible condition. However, the extent to which tangible reinforcement routinely produces such outcomes is unknown The purpose of this study was to examines susceptibility to tangible reinforcement by determining (a) whether a new response is acquired more readily when exposed to tangible reinforcement than when exposed to other contingencies commonly used in a functional analysis, and (b) whether problem behavior known not to have a social function nevertheless emerges quickly when exposed to tangible reinforcement. Results suggest that the inclusion of items in the tangible condition should be done with care and be based on those items in the individual’s environment. |
| |
| Using discriminative stimuli to facilitate condition differentiation during brief functional analyses |
| KELLY MCCONNELL (New England Center for Children), Eileen M. Roscoe (New England Center for Children), William H. Ahearn (The New England Center for Children), Jason C. Bourret (New England Center for Children) |
| Abstract: Previous research has shown that including salient stimuli (e.g., different colored rooms or different therapists associated with each condition) may enhance differential outcomes during a functional analysis (FA). However, clinicians may not have the resources necessary for using discriminative stimuli when conducting a FA. Therefore, the purpose of this study was to evaluate the effects of using practical stimuli, (e.g., different colored shirts, poster boards, and photos of the participant contacting antecedent environmental events associated with FA conditions) on FA outcomes. In addition, because the inclusion of salient stimuli may be most helpful when conducting brief or nonstandard FAs, we evaluated the effects of using such stimuli during brief or latency-based FAs. Four individuals with autism, who exhibited severe problem behavior, participated. A multielement design was used to demonstrate experimental control, and a block of 4 conditions (alone, attention, play, and demand) paired with discriminated stimuli was alternated with a block that was not paired with these stimuli. Results indicate that the inclusion of salient stimuli may serve as a practical enhancement when conducting brief functional analyses. Interobserver agreement was conducted for over 30% of sessions and averaged over 90%. |
| |
| Expediting the Brief Functional Analysis by Using Hypotheses Derived from Descriptive Assessments |
| MATTHEW O'BRIEN (The University of Iowa Children's Hospital), Brenda J. Engebretson (University of Iowa Children's Hospital), David P. Wacker (University of Iowa Children's Hospital), Kelly M. Schieltz (University of Iowa), Melanie H. Bachmeyer (University of Iowa), Patrick Romani (University of Iowa) |
| Abstract: The Behavioral Pediatrics Clinic at the University of Iowa Children’s Hospital is a clinic for young children who are typically developing and who engage in disruptive behaviors (e.g., noncompliance, aggression). Because of appointment time constraints, an alternative to the extended functional analysis was needed so that response-reinforcer relations could be assessed within a 90-minute evaluation. In the early 1990’s, the brief functional analysis (BFA) was developed as this alternative, testing the response-reinforcer relationship between different sources of negative and positive reinforcement on target behaviors via a multi-element design. Studies have since demonstrated the BFA to be efficient and often correlated with extended functional analyses. The current clinic assessment protocol consists of descriptive assessment (e.g., A-B-C interview), a BFA, and treatment evaluation. Inter-observer agreement is collected during assessment and treatment evaluations. The sequence of test conditions (e.g., demand, attention) in the BFA is based on the hypothesized response-reinforcer relation identified via the descriptive assessment. Using hypothesis-based analyses allows us to test and confirm response-reinforcer relationships in an expeditious fashion and initiate treatment more quickly. This presentation will include a discussion of the assessment protocol, a summary of the outcomes from utilizing the assessment protocol, and a case example demonstrating this approach. |
| |
| A comparison of fixed and random session sequences during functional analyses |
| JEFF TIGER (Louisiana State University), Megan L. Kliebert (Louisiana State University), Karen A Toussaint (Louisiana State University), Joslyn N. Cynkus (Louisiana State University), Carolyn Barahona (Louisiana State University), Kristen Abbondante (Louisiana State University) |
| Abstract: Multielement designs are the most common format for conducting functional analyses. While highly efficient, these designs introduce potential of carry-over effects between conditions which researchers have accounted for in 2 ways. One approach has been to randomize session order such that carry-over effects are likely to be equally distributed across conditions. The other approach has been to conduct sessions in a fixed sequence designed to capitalize on the impact of carry over; possibly expediting the analysis (i.e., alone conditions are likely to establish the value of attention; so they are beneficial to conduct prior to attention sessions). The current study was designed to determine if functional analysis outcomes may be determined more rapidly when sessions are conducted fixed relative to a random sequence. Each participant experienced 2 simultaneous functional analyses of their problem behavior. During one analysis sessions were sequenced in a random order, and during the other analysis, sessions were sequenced in a fixed order (alone-attention-toy play-escape). The resultant data from each analysis were then inspected with structured criteria to determine how quickly a determination of behavioral function could be made. Interobserver agreement was collected for at least 25% of each analysis and averaged above 90%. |
|
| |