William V. Dube - US grants

This proposal seeks a FIRST award to support a five-year study of discrimination learning in mentally retarded individuals. The research will make use of a programmed method for generating one-trial discrimination learning (OTDL). Successful completion of the OTDL program establishes the following two-trial performance: On the first trial, a single visual stimulus Is displayed and Its selection is reinforced. On the second trial, two stimuli are displayed, one from the previous trial and a different stimulus: selection of the previously displayed stimulus is defined as correct. Different stimuli appear on each two-trial problem. Consistently accurate performance on the second trial of each problem demonstrates acquisition of new discriminations after exposure to a single training trial. The proposed studies will apply the programmed OTDL method to several areas of interest in the analysis and remediation of learning problems in mental retardation. The specific aims are: (1) To compare the effects of programmed instruction and verbal instruction on several characteristics of the OTDL performance. By including subjects across a range of mental retardation, the comparisons will test the prediction that acquisition of such performance in lower-functioning subjects will be more successful with programmed procedures, and that higher-functioning subjects will learn equally well with either procedure. Subsequent exposure to changes in reinforcement contingencies will examine reported differences in "contingency shaped" and "rule governed" behavior. (2) To develop broadly applicable procedures for studying "breadth of attention" in mentally retarded individuals. including those classified as severely retarded. Breadth of attention refers to discrimination of the individual elements of complex stimuli after very limited exposure (e.g., a single trial) to each complex. The studies will (a) examine effects of training history on element discrimination, (b) examine for the first time breadth of attention in low-functioning mentally retarded individuals, and (c) determine whether attention can be broadened by extension of the programmed methods. (3) To determine if one-trial learning can be extended to conditional discrimination in the arbitrary matching-to-sample format. The study will investigate conditions under which a single exposure to a two-element complex will lead to selection of one element conditionally upon the other. (4) To determine if one-trial learning can be extended to the formation of classes of equivalent stimuli. The study will ask if conditional relations established in one trial are also relations of equivalence, according to widely accepted criteria established by Sidman & Tailby (1982).

mental retardation education /training; statistics /biometry; biomedical facility;

sensory discrimination; learning disorders; mental retardation education /training; reinforcer; visual stimulus; mental retardation; stimulus /response; visual tracking; stimulus generalization; computer assisted instruction; programmed instruction; attention; performance; behavioral /social science research tag; adolescence (12-20); middle childhood (6-11); clinical research; human subject; behavior test; psychometrics;

Stimulus overselectivity, also known as restricted stimulus control, is a widely acknowledged problem in the education of individuals with developmental disabilities. Overselectivity refers to learning that is atypically limited with respect to range, breadth, or number of stimuli or stimulus features. In the education of individuals with developmental disabilities, overselectivity may be related to inefficient, incomplete, or inconsistent performance with multi-element stimuli (e.g., printed words). The research will investigate variables that affect overselectivity and the development of stimulus control by multiple or multielement stimuli. From the existing literature on our study population of individuals with moderate to severe developmental disabilities, one may conclude that the number of remedial studies is remarkably small, given the prevalence of the problem. Importantly, no generally applicable approach toward solving the problem has yet emerged. Our preliminary studies have set the stage for such an effort by providing compelling evidence that overselectivity (1) is open to manipulation by reinforcement contingencies, (2) may be directly related to deficiencies in observing behavior, and (3) can be greatly reduced by interventions that control observing behavior. The goals of the proposed research are: (1) To compare the prevalence and degree of overselectivity in children and adolescents who have mental retardation and autism, versus those who have mental retardation only, and to examine the relation between overselectivity and measured mental age (MA) scores. (2) To investigate environmental variables that may affect the behavioral aspects of stimulus overselectivity, including reinforcement variables that have been little studied in past research. (3) To develop effective remedial procedures that reduce or eliminate overselectivity. Approaches include stimulus control shaping procedures (e.g., fading) to shape effective observing, and the use of response shaping procedures to teach participants to make explicit responses that guide effective observing. (4) To examine the generalization of remedial training across tasks in a laboratory study of overselectivity. (5) To examine generalization across tasks and settings in field studies to be conducted in special-education classrooms with tasks that have been identified in participating students' educational plans.

This proposal requests continuation funding for a project investigating relationships between cognitive and behavioral flexibility and reinforcement processes. The approach is motivated primarily by an extensive program of prior research from our laboratories at the Shriver Center and behavioral momentum, theory which relates behavioral persistence to the rates of reinforcement signaled by contextual cues. The overall goal of this project is to examine environmental variables that determine behavioral flexibility and to develop a principled, broadly applicable approach to designing behavioral interventions to make adjustments-both increases-in flexibility. Results thus far have shown that: (1) behavioral allocations of individuals with mental retardation conform to relationships predicted by Herrnstein's "matching law" in choice situations; (2) behavioral momentum theory consistently describes the behavior of individuals with mental retardation in laboratory tests; and (3) discriminati on learning procedures designed in accordance with behavioral momentum theory can reduce maladaptive persistence of stimulus control when contingencies changes (e.g., in reversal learning). The specific aims of the continuation project are to conduct basic and applied research that: (1) Documents the extent to which behavioral momentum effects demonstrable in the laboratory will also be found within special-education settings; (2) develops intervention procedures to help individuals with mental retardation respond more effectively in situations that typically occasion distraction or inflexibility in behavioral transitions; (3) Extends functional-analysis diagnostic procedures to include evaluations of behavioral persistence and resistance to change; (4) defines the temporal context in which momentum analyses apply to individuals with mental retardation; (5) applies the principles of behavioral momentum to improve procedures for establishing discrimination skills in individuals with int ellectual disabilities; and (6) compares the effectiveness of momentum-based interventions for behavioral transitions, self-stimulatory behavior, and prompt dependency in individuals who do and do not have autism diagnoses.

This project will investigate a behavioral economics approach to the management of motivational variables for persons with intellectual disabilities. In behavioral economics, reinforcers are treated as commodities, and the environmental contingencies that govern access to those commodities as the price. To conduct the analysis, we will first determine individual demand functions for specific commodities. Demand functions relate consumption rate to price; they can be obtained by a parametric assessment of consumption over a range of prices. The demand function provides a more comprehensive picture of performance dynamics than the point estimate produced by traditional/current methods of evaluating reinforcer preference and effectiveness. Demand is termed elastic if consumption falls rapidly as price increases, and inelastic if it does not (i.e., output increases as price rises). We will characterize variables that affect demand functions and the subsequent effects on dependent measures that include both acquisition and performance. These variables include (1) price increases due to amount vs. difficulty of work; and (2) substitutability of commodities. Commodities are substitutes if the concurrent availability of one increases the elasticity of demand for another, and complements if increased consumption of one increases demand for the other. Substitutability analyses will identify classes of reinforcers defined by similar economic functions. (3) We will also examine the effects of open vs. closed economies. This dimension refers to the degree to which commodities are available outside of the economy under study. (4) The project will also include research to examine the economic dynamics of token exchange systems, or "token economies." We will study the effects of schedule manipulations for earning and exchanging tokens on performance accuracy, response rate, and off-task or disruptive behavior, and we will investigate the potential for "virtual tokens" for computer-based instruction. We expect the results of our studies to define a technology for obtaining individualized economic profiles that will be useful in designing more effective motivational systems, and to inform designs of motivational systems for persons with intellectual disabilities by producing algorithms to support a technology for "tuning" such systems to accommodate individual differences. A series of applied intervention studies will demonstrate the application of the research to clinical and educational problems in the special-education classroom.

This project will apply quantitative analyses of behavioral choice to situations that involve social reinforcers, defined here as positively reinforcing events that are mediated by the actions of another person (e.g., attention, approval, affection); and social contexts, defined here as those marked by stimuli that are reliable cues for the availability of social reinforcers (e.g., the immediate presence of a familiar person). The goals of the project are to develop methods for maximizing adaptive sensitivity to social reinforcement contingencies and minimize maladaptive behavioral allocations in social contexts. One set of basic studies will examine sensitivity to reinforcement contingencies in social and nonsocial contexts. Sensitivity refers to the degree to which the individual's behavioral allocation among concurrently available alternatives matches the relative amounts of reinforcement that are obtained from those alternatives. If sensitivity is high, reinforcement is maximized as behavior is re-allocated in response to changes in reinforcement contingencies. The basic studies will be conducted in automated testing environments where experimental sessions in which the presence/absence of an adult in the immediate environment will be rigorously controlled as experimental variable. These studies will develop methods for determining individual profiles of contingency sensitivity in social vs. nonsocial contexts. Our hypothesis is that adaptive behavior is maximized when sensitivity is unaffected by context shifts. We will test this hypothesis by examining the acquisition of new behavior in social vs. nonsocial contexts with individuals who have a range of social-reinforcer profiles. We will also investigate procedures for improving maladaptive profiles in both social and nonsocial contexts, and the subsequent effects of improvements on learning. A second set of studies will examine social reinforcer function in relation to the well-documented deficit in the development of joint attention in young children with autism and related disorders (PDD-NOS). Joint attention refers to the use of gestures, verbalizations, and gaze shifts to coordinate attention with another person in order to share the experience of an object or event. The proposed research will examine correlations between quantitative measures of preference and sensitivity for social reinforcers and the results of assessments of joint attention initiation and develop training procedures that target specific behavioral prerequisites for joint attention initiation. These prerequisites are derived from a functional analysis of joint attention initiation. We will monitor concurrent changes in preference, contingency sensitivity, and joint-attention initiation as training progresses.

DESCRIPTION (provided by applicant): These P01 projects are unified by a shared focus on behavioral processes involved in the allocation of behavior - choice-making - among individuals with intellectual disabilities. They are also bound by a common effort to translate extensive basic laboratory research findings into effective, individualized motivational support to individuals with intellectual disabilities. Through application of quantitative approaches to behavioral allocation, our goal is to advance understanding of the variables that underlie the types of choices made by persons with intellectual disabilities. A longterm goal is to develop a technology for teaching choice-making that benefits the individual, the family, and the larger social context. Toward this end, each project will include both basic research on advanced quantitative methods to problems relevant to mental retardation and applied research to validate the resulting techniques in ecologically valid teaching settings. These themes are explored in four projects and supported by three cores. Project 1 emphasizes basic choice processes in behavioral allocation and persistence in persons with severe intellectual disabilities. Project 2 adopts a behavioral economics approach to the management of motivational variables for persons with intellectual disabilities. Project 3 utilizes signal detection theory in the quantitative analyses of motivational variables. Project 4 applies quantitative analyses of behavioral choice to situations that involve social reinforcers (those mediated by another person) and social contexts (those marked by cues for of social reinforcers). The projects are further unified by three research cores that provide essential, cost-effective services. TheTechnical Support core centralizes and renders more efficient the common technical support needs of the program (e.g., computer programming, database construction and management, statistical analysis, apparatus maintenance). Throughout these projects, the Participant Services core assists in subject recruitment, diagnostic and psychometric evaluation, scheduling, and continuity of participation.

DESCRIPTION (provided by applicant): These P01 projects are unified by a shared focus on behavioral processes involved in the allocation of behavior - choice-making - among individuals with intellectual disabilities. They are also bound by a common effort to translate extensive basic laboratory research findings into effective, individualized motivational support to individuals with intellectual disabilities. Through application of quantitative approaches to behavioral allocation, our goal is to advance understanding of the variables that underlie the types of choices made by persons with intellectual disabilities. A longterm goal is to develop a technology for teaching choice-making that benefits the individual, the family, and the larger social context. Toward this end, each project will include both basic research on advanced quantitative methods to problems relevant to mental retardation and applied research to validate the resulting techniques in ecologically valid teaching settings. These themes are explored in four projects and supported by three cores. Project 1 emphasizes basic choice processes in behavioral allocation and persistence in persons with severe intellectual disabilities. Project 2 adopts a behavioral economics approach to the management of motivational variables for persons with intellectual disabilities. Project 3 utilizes signal detection theory in the quantitative analyses of motivational variables. Project 4 applies quantitative analyses of behavioral choice to situations that involve social reinforcers (those mediated by another person) and social contexts (those marked by cues for of social reinforcers). The projects are further unified by three research cores that provide essential, cost-effective services. TheTechnical Support core centralizes and renders more efficient the common technical support needs of the program (e.g., computer programming, database construction and management, statistical analysis, apparatus maintenance). Throughout these projects, the Participant Services core assists in subject recruitment, diagnostic and psychometric evaluation, scheduling, and continuity of participation.

This project will investigate a behavioral economics approach to the management of motivational variables for persons with intellectual disabilities. In behavioral economics, reinforcers are treated as commodities, and the environmental contingencies that govern access to those commodities as the price. To conduct the analysis, we will first determine individual demand functions for specific commodities. Demand functions relate consumption rate to price; they can be obtained by a parametric assessment of consumption over a range of prices. The demand function provides a more comprehensive picture of performance dynamics than the point estimate produced by traditional/current methods of evaluating reinforcer preference and effectiveness. Demand is termed elastic if consumption falls rapidly as price increases, and inelastic if it does not (i.e., output increases as price rises). We will characterize variables that affect demand functions and the subsequent effects on dependent measures that include both acquisition and performance. These variables include (1) price increases due to amount vs. difficulty of work; and (2) substitutability of commodities. Commodities are substitutes if the concurrent availability of one increases the elasticity of demand for another, and complements if increased consumption of one increases demand for the other. Substitutability analyses will identify classes of reinforcers defined by similar economic functions. (3) We will also examine the effects of open vs. closed economies. This dimension refers to the degree to which commodities are available outside of the economy under study. (4) The project will also include research to examine the economic dynamics of token exchange systems, or "token economies." We will study the effects of schedule manipulations for earning and exchanging tokens on performance accuracy, response rate, and off-task or disruptive behavior, and we will investigate the potential for "virtual tokens" for computer-based instruction. We expect the results of our studies to define a technology for obtaining individualized economic profiles that will be useful in designing more effective motivational systems, and to inform designs of motivational systems for persons with intellectual disabilities by producing algorithms to support a technology for "tuning" such systems to accommodate individual differences. A series of applied intervention studies will demonstrate the application of the research to clinical and educational problems in the special-education classroom.

The primary goal of Project 3 is to contribute to the development of an explicit technology of treatment generalization for chronic aberrant behavior (CAB). The past 20 years have seen significant advances in the development of intervention and treatment strategies. Because CAB is often treated in specialized clinical or educational settings, long term success of these interventions hinges on the successful generalization of treatment outcomes to non-treatment settings. Technologies for engineering the generalization of treatment gains obtained under highly controlled circumstances remain underdeveloped. Project 3 will conduct translational research that takes a stimulus-control perspective as its conceptual foundation. This project will be the first comprehensive effort to examine the problem of treatment generalization from the perspective of stimulus equivalence, including both acquired equivalences among discriminative stimuli that set the occasion for behavior and the related issue of the functional equivalence among the positive/negative reinforcing stimuli that are the consequences of behavior. Although equivalence research has advanced significantly in the past 30 years, there has been relatively little translation of these findings to the analysis of cross-setting transfer of treatment effects. The proposed studies are derived from the assumption that successful generalization and maintenance of treatment effects depends upon equivalence between treatment and generalization settings. We propose research aimed at two stages of the treatment process. The first is the initial diagnostic stage in which the function of problem behavior is determined by functional analysis (FA). From our perspective, future generalization will depend on part to the extent that the FA environment creates a model that is functionally equivalent to the environment in which the problem behavior occurs. Undifferentiated and false-positive FA results may reflect lack of equivalence. The second stage of treatment that we will address is the transfer of treatment gains achieved within a controlled environment to the everyday settings in which problem behavior occurs. From our perspective, generalization is facilitated when the FA-inspired treatment environment and the everyday environment are equivalent along discriminative and consequential dimensions. The specific aims of the project are: (1) To develop and validate a protocol for identifying relevant variables following undifferentiated outcomes with standard FA procedures;(2) To develop and validate a protocol for identifying both positive and negative reinforcement in FA Demand conditions;(3) To define and validate a process for applying stimulus-control shaping procedures to transfer the effects of successful CAB-reduction procedures from the clinical environment to the individual's usual daily environment;and (4) To define and validate a process for (a) systematically identifying equivalence relations between therapists, the settings, and the materials in treatment and generalization environments;and (b) prescribing precisely targeted equivalence training procedures that promote treatment generalization.

DESCRIPTION (provided by applicant): This application seeks support for research to address the problem of stimulus overselectivity as it may impact stimulus control in functional academics and augmentative/alternative communication (AAC). Overselectivity refers to maladaptive narrow attending that is a common learning problem in children with intellectual disabilities. Overselectivity is often associated with autism spectrum disorders (ASD), but the research foundation for this relation is inconclusive and largely based on procedures that fail to capture the complex and dynamic relational learning aspects of special-education curricula. Recent research has produced methodologies to study overselectivity in contexts that model teaching situations, as well as promising remedial procedures that can reduce or eliminate overselective attending by corrective therapy and/or behavioral prostheses. We propose a formal comparison of stimulus overselectivity in four study populations: MA- and CA-matched children with ASD, Down syndrome, and intellectual disabilities of mixed/unknown etiology; and MA-matched typically developing children. Tests will include stimuli developed for basic research, as well as stimulus sets of clinical/educational interest (AAC icons, photos of faces, and printed words). We will determine whether an ASD diagnosis is related to (1) increased prevalence or severity of overselective stimulus control; (2) a deficit in the disengagement of attention and/or indifference to perceptual coherence of stimuli; and (3) the effectiveness, durability, and net gain resulting from intervention and remedial training. We will also apply behavior-analytic quantitative models of attention (4) to determine whether strategic manipulations of reinforcement parameters can be used to identify and ameliorate overselectivity that emerges from attention biases interacting with the uncontrolled reinforcement contingencies of teaching procedures typically used in special-education settings. Finally, (5) we will conduct a series of applied studies to examine generalization and durability of remedial interventions for academic tasks in special-education classrooms. Results of the proposed studies will contribute to a better characterization of the learning problems associated with the study populations and to increased understanding of basic stimulus control processes in learning. Application of the resulting knowledge seems highly likely to improve current methods for teaching and evaluating individuals with intellectual disabilities. PUBLIC HEALTH RELEVANCE: The proposed research is highly relevant to the public health interests of NICHD and its Mental Retardation Developmental Disabilities Branch because it will contribute to the improvement of evaluation and teaching methods for individuals with autism and other intellectual disabilities. Specific areas of application include improvements in functional communication capabilities for children with moderate to severe disabilities.