Frequency measures of behavior for assistive technology and rehabilitation

Abstract

Documenting assistive technology outcomes has grown in importance, but outcome measurement remains problematic. A new approach uses natural science measures and a model (selectionism) from the field of Behavior Analysis. Selectionism defines behaviors by their effects (functional performance) and the environment (including technology) within which they occur, and explicitly treats variation in patterns of behavior over time for individuals (intervention effects). Its basic metric is frequency of behaviors (count per unit time) which is similar to robust engineering measures like centimeters, grams, and seconds. This approach eliminates many of the problems inherent to more traditional psychometrics. Selectionism based on frequencies also provides an empirical structure or taxonomy to organize efforts and outcomes, unified by the notion of fluency. Composite behaviors are combinations of smaller component behaviors that are required for performance of the composite. A frequency above which a component behavior is readily retained, generalized, and recruited into the more complex composite behavior is called fluency; thus individuals fluent on the critical components easily and efficiently demonstrate the composite. This model suggests that when assistive technology interventions raise component behavior frequencies to fluent levels, they will be integrated usefully into an individual's life. This selectionistic approach has been used successfully in the field of education. It has the added benefit of not only empirically defining measurable outcomes, but also of providing useful ongoing measurement of change during treatment. This paper briefly describes this "Precision Measurement" strategy and its data-driven feedback process and makes suggestions for further research and development efforts. The method provides a basis for better documentation, control, and outcomes of assistive technology and related interventions.