ИСТИНА |
Войти в систему Регистрация |
|
ИСТИНА ИНХС РАН |
||
Nowadays new methods of lie detection are actively developed, including those based on the eye-tracking technology. It was shown that the count and duration of fixations, blinks and saccades could be used as reliable indicators in the detection of concealed information. In our study we investigated the dwell time and the number of revisits to detect whether a subject is lying or not. All observers performed the task to tell the truth about one familiar face (FF1) and to conceal the information about another familiar face (FF2). We hypothesized that in the situation of concealing the information about FF2 the subject’s eye movements during the face scanning would differ in comparison to the situation of perceiving FF1 or unfamiliar faces. Thirty three subjects took part in the experiment (22F, 11M, age range 18—22). All of them were familiar with the persons shown in the photos (FF1 and FF2). Our stimuli were a 3x3 matrix consisted of 9 photos of faces. Three types of matrix were constructed comprising: the familiar face FF1 among 8 unfamiliar faces (type I); the familiar face FF2 among 8 unfamiliar faces (type II); 9 unfamiliar faces (type III). The experimental session involved 54 trials. On every trail the subjects were asked to look for a familiar face and then tell truth or lie about him. Stimuli were presented for 5000 ms. Results showed that the eye movements during the recognition of familiar and unfamiliar faces were different: the subjects looked at familiar faces longer, made more fixations and more revisits during their perception. But there were significant differences in some eye movement parameters during FF1 and FF2 perception. So, significant differences in the dwell time were revealed for the familiar faces FF1 and FF2 (p<0,01) in 30 observers (91%); for the familiar face FF2 and unfamiliar faces (p<0,01) in 22 observers (67%). The number of revisits were significantly higher for the familiar face FF1 in comparison to FF2 in 25 participants (76%) and in comparison to unfamiliar faces (p<0,01) in 16 participants (48%). Thus the obtained data allow considering the dwell time and the number of revisits as reliable measures for detecting deception. These characteristics may be used in combination with other eye movement and psychophysiological parameters to improve lie-detection method based on eye movements.