Signal-detection analysis of recognition memory of obese subjects.

Obese and normal-weight subjects were run on a series of recognition-memory tasks with low and high meaningful CVC trigrams. Subjects received feedback or no feedback regarding the appropriateness of their answers during testing for recognition memory. A signal-detection analysis was used to examine a relatively pure index of memory (d') as compared to response-bias factors [Ln (beta)]. No significant main effect on memory of obese and control subjects was noted although a significant interaction indicated that obese subjects' memory increases over trials in a differential fashion from normal-weight individuals with poorer performance among the obese after 3 trials. Feedback facilitated memory for both groups. Analysis of response criteria [Ln (beta)] showed no significant difference between groups. Highly meaningful trigrams had a significantly lower response criterion than low meaningful trigrams and there was a significant interaction of meaningfulness by trials. An interaction of feedback by meaningfulness was also present. Advantages of using a signal-detection analysis in memory studies comparing obese and normal weight subjects are discussed. Ramifications of the present data for the internality/externality hypothesis of obesity are also discussed.