Association for Behavior Analysis International

Conclusions derived from current functional assessment practices heavily rely on indirect methods for gathering data (e.g. FAST, MAS). When a function is experimentally tested, current practices pose ethical, practical and theoretical concerns. Both approaches are problematic in that indirect data produces inaccurate and imprecise data, and experimental methods are typically not driven by a hypothesis, directly reinforce problematic behaviors, and do not involve simultaneous establishment of appropriate alternative behaviors. The first presentation will focus on a comparison between two indirect and one direct data collection method to generate hypothesis regarding function that is more accurate and efficient. The second will propose alternative experimental methods to test a subset of hypothesized functions and involve teaching alternative responses and do not reinforce problematic behaviors. The symposium will conclude with an argument to support (1) direct observation of consequences be used in place of indirect data to develop hypothesis and (2) use of use of alternative experimental methods such a free-operant and trial-based functional analysis procedures. The proposed methodology provides a more ethical, conceptually systematic, and practical assessment of function.

Indirect data obtained through interviews such as the Functional Analysis Screening Tool (FAST) and Motivational Assessment Scale (MAS) are commonly used to develop hypothesis regarding function of problematic behavior despite the known inaccuracies produced by indirect data. While researchers use the FAST and MAS as a simple way to raise hypothesis, such use can be problematic if (1) the instrument fails to include the actual function as part of the hypothesis (miss) and (2) does not filter enough possibilities (false alarms). Those outcomes may mislead or waste assessors’ and clients’ time and efforts. The current research aimed to compare the FAST and MAS to a direct observation of consequences that follow problematic behaviors (Beacon Consequence Analysis Form - BCAF). Data from twelve children whose function of problematic behaviors were confirmed by a trial-based or free-operant functional analysis were used in this study. A comparison of the instruments hypothesis and confirmed function was conducted and rates of hits, correct rejection, misses, and false alarms calculated. The results supported the use of the BCAF which had the highest rates of hits (100%) and correct rejections (93.3%) and lowest rates of misses (0%) and false alarms (6.7%) to raise hypothesis regarding potential function.

Skinner defined functional analysis as the identification contingencies of reinforcement responsible for the acquisition of maintenance of responses. Iwata et al (1994) introduced a procedure that involved direct manipulations of the antecedent and consequences in order to experimentally determine the function of the problem behavior. While its approach was and has been now widely accepted some ethical, practical and theoretical concerns may be raised. For example, the appropriateness of its wide use may be questionable in some situations due to its directly reinforcing specific topographies of problematic behavior, and it assumes an invariable relationship between antecedent conditions and the consequences maintaining problematic behavior (e.g., problematic behaviors occurring under demand conditions are always reinforced by escape) which is not always the case. For example, a demand condition may function as a discriminative stimulus that attention (follow through with the demand) will be delivered. The present study attempts to identify a trial-based and free-operant alternative to conducting a functional analysis that involves teaching functional communication responses rather than reinforcing problematic behaviors. This method will be described and examples concerning functional analysis of multiple topographies such as prompt dependency, aggressions, and tantrums will be reviewed.