Effects of certain training parameters on detection of simulated breast cancer.

Forty-two female subjects, constituting 14 groups, palpated a silicone breast model in a lump detection task. Three variables, number of trials (amount of practice), search pattern requirement (restricted vs. unrestricted), and the presence or absence of lumps, were manipulated according to 3 x 2 x 2 design. The effect of all modes of practice on detection performance in posttest 1 was evidenced by an increase in percent of lumps detected and a decrease in detection threshold, but no form of practice was superior to the others. Following posttest 1 a brief search-training procedure was administered to 39 subjects which further lowered detection threshold, increased percent of lumps detected, and increased trial duration in posttest 2. The results indicate the effectiveness of the training procedure for improving detection ability and suggest the need for a more complex model to determine more specifically the effects of the practice models employed.