Following the logic of a prior experiment (Seligman et al., 1990) with vars
ity collegiate swimmers, sixty student volunteers performed a brief, but hi
ghly stressful vigilance task. Half were then given false feedback indicati
ng poor performance; the others experienced a non-evaluative display at tha
t point. All were then asked to repeat the vigil. Half the observers were a
ssessed as high optimists and half as high pessimists. The pessimists showe
d a steeper vigilance decrement than the optimists, consistent with a model
(Scheier and Carver, 1987) that proposes that pessimists are more emotion-
focused under stress than optimists, and hence would be less attentive to t
he vigilance display, regardless of feedback condition. The false negative
feedback actually turned out to be ineffective; hence, based on Seligman's
account of the swimmer study, which requires effective, negative feedback,
there should have been no vigilance performance difference between optimist
s and pessimists, either pre- or post-feedback.