Abstract:
In an attempt to replicate an existing study of recall and response bias, 77 students, divided into groups of depressed, formerly depressed and non depressed subjects were presented with word lists comprising equal numbers of positively and negatively weighted words. In addition to the original study a group of 20 clinically depressed outpatients was included in the study. Subjects were asked first to rate the words according to their level of self-descriptiveness and then to select the 10 most self-descriptive from both the positive and negative lists. They were given recall tests for their 20 most self-descriptive words at 1 hour, 2 day, and 7 day intervals. Measures were taken of both correct recall and errors (intrusions). The subjects were also given a 7 day recognition test for the original lists plus 10 positive and 10 negative distractor words (words not previously seen). The depressed subject groups recalled less than other groups, and produced most errors (intrusions). All groups recalled more positive than negative adjectives, both correctly and incorrectly. All groups made more positive hits (correct recognitions) than negative hits. The clinically depressed subjects produced more positive than negative false alarms (incorrect recognitions). Signal detection analysis was considered, however because of methodological problems was not included. Failures to confirm results of the original study are discussed.
