A guessing task was performed in three conditions of feedback sequence. One sequence was random; the other sequences were nonrandom. Each subject made 300 responses per session for four sessions. The frequency distribution of the differences between response/feedback for the immediately preceding trial and the current response was analyzed per session and for the pooled data. The results suggest that there are two response processes operating in this task: a static process and a dynamic process. This model has two advantages. First, it can explain the sequential dependencies in the present study and in Ward and Lockhead's (1971) Study 3. Second, each of the two processes proposed here has a theoretical foundation: the static process is based on the subject's tendency to generate a random sequence with many alternatives; the dynamic process is a modified version of a learning model in a signal detection task (Dorfman, Saslow, & Simpson, 1975).
All Science Journal Classification (ASJC) codes