Abstract
An experiment was conducted to determine whether highly depressed and non-depressed psychiatric inpatients and non-patient normal controls react differently on a concept attainment task to different reinforcement contingencies. Specifically, it was hypothesized that:
1) Depressed and non-depressed psychiatric patients and normal controls would perform at approximately the same levels on a concept attainment task when they received informational feedback only.
2) Depressed patients would show less improvement than normals and non-depressed patients when positive reinforcement (money) was given in addition to informational feedback in a concept attainment task.
3) Depressed patients would show markedly poorer performance than non-depressed patients and normals when punishment was administered for incorrect responses according to certain deductions from Freud's theory, but would be predicted to perform as well or better according to Beck's theory.
4) Following Freud's theory, depressives might initially attain concepts because of the demands of their ‘duty morality’ but would subsequently give incorrect responses to avoid positive reinforcement and/or gain punishment, since they reject ‘pleasure morality.’ This wolld not be predicted by Beck's theory.
5) If, as Beck contends, depressives are more sensitive to fear and frustration than non-depressed persons, depressed patients should have greater response latencies under punishment conditions than non-depressed persons.
A 3 x 4 factorial design was used with three diagnostic categories and four reinforcement conditions with four replications (4 Ss in each cell). The four reinforcement conditions were information feedback only, positive reinforcement only, punishment only, and positive reinforcement and punishment. The three diagnostic categories were high-depressed patients scoring greater than 24 on the Beck depression inventory (DI); low-depressed patients scoring less than 15 on the Beck DI; and normal controls.
The results did not strongly support any of the five hypotheses. In general, effects attributable to the different reinforcement conditions were weak and did not interact with the diagnostic categories as required to . support the hypotheses. Normals tended to perform better than psychiatric patients, and punishment tended to be associated with longer response latencies than positive reinforcement. Implications of these findings are discussed.