Functional Analysis

In 1953, B. F. Skinner implicitly introduced the construct of a functional analysis of behavior by positing a causal chain consisting of an operation performed [Page 852]on the organism, a behavioral response by the organism, and some kind of stimulus that follows the behavior that affects the likelihood of the previous behavior being exhibited under the same or similar conditions. This way of construing behavior implied a notion of causality for a particular behavior. Functional analysis refers to a variety of ways of attempting to identify the causes of behavior with the goals of predicting, and controlling it once a proper understanding of the controlling variables occurs. Though Skinner speculated about a variety of complex human behaviors taking place in the natural environment, most of his actual analysis of behavior was confined to highly controlled laboratory situations where the presumed causal variables were well controlled, measured, manipulated, and recorded. In more contemporary times, functional analysis has been attempted in more complex environments where many factors can affect behavior. Humans, having more complex histories and behavioral repertoires, present a greater challenge when attempting to determine the cause of a behavior.

Applied psychologists face the task of understanding and often changing behavior in many situations, including schools, mental health institutions, clinical settings, families, organizations, and communities. Behavioral scientists must assume that behavior is determined by some factor rather than being random, or else it makes no sense to try to bring about change. The search for causal factors for behavior is conducted in a way similar to the traditional scientific method.

In the scientific method, observations about the phenomenon in question are made, hypotheses about factors that control the phenomenon are developed, experimental manipulations of important variables are made, and the observed results provide the data for a test of the hypothesis. If necessary, the hypotheses are modified, data are collected again, and the process goes through these iterations until a workable theory of causality is developed. This heuristic appeals to a variety of researchers, including behavioral scientists.