Abstract

Eight clinically anxious and eight non-anxious subjects were exposed to a stimulus equivalence training procedure. Each matching-to-sample training involved threatening situations (A), nonsense syllables (B), and pleasant-state adjectives (C). All subjects met criteria for learning A-B relationships and B-C relationships in a matching-to-sample procedure, but in the critical test phase (where C sample elements are matched to A elements) the non-anxious group differed from the anxious group on two of the three C-A tests. Responding appropriately on all three C-A tests, thus indicating that the stimuli had become members of an equivalence class, occurred in six out of eight non-anxious subjects and only one out of eight anxious subjects. This striking between-group difference suggests that the presence of clinical anxiety can affect stimulus equivalence class formation, and that the procedures used here could be developed as a diagnostic procedure.