被験者の表象と質問内容のずれが虚偽検出成績に及ぼす影響

  Two experiments examined the influence of differences between memory representation and the content of critical items presented visually for detection accuracy in a guilty knowledge test.<br>   In Experiment 1, skin conductance responses (SCR) of 12 subjects were measured under three conditions. In each of two conditions, one of two pictures that differed in the degree of match to the stimulus memorized previously was used as a critical item, while the identical picture was used in another condition. The results indicated that detection accuracy on the item relatively similar to the memorized material was lower than that of the item identical to representation. Both items could be detected. However, when the critical item was dissimilar to the previously memorized stimulus, the rate of correct detection did not indicate a significant increase above chance expectancy.<br>   In Experiment 2, two types of pictures, identical and dissimilar pictures, were used as critical items and names of which were also presented orally. There was another condition where the critical item was presented only orally. Skin conductance responses of 12 subjects were measured. In the dissimilar picture condition, detection accuracy was lower than that of the identical picture condition, but the rate of correct detection was significantly greater than a chance level. There was no statistically significant difference on detection accuracy between the dissimilar picture and oral conditions.<br>   The results of these studies showed that 1) examiners shoud use critical items which were as similar to memory representation of the subjects as possible, and 2) there could be a case that was successful to detect deception even if the critical items were not identical to memory representation.<br>