| Abstract/Results: | ABSTRACT:
OBJECTIVES:
From mid-twentieth century, the detection of deception has become recognized as an object of scientific study, especially in the case of guilty knowledge. Nevertheless, the interpretations depend on the human factor. Thus, we tried to demonstrate that you can automate the detection of deception from psychophysiological records in order to avoid bias attributable to the researcher. For this we identify differences in SCR between "truth" and "deception". We will use other psychophysiological measures to increase the effectiveness of this algorithm.
METHOD:
Participants: 17 participants, aged between 20 and 30 years of age (M = 23.18, SD = 2.74), 9 of them female.
Materials: A Biofeedback I-330-C2+ (J & J Engineering) system, connected to a laptop for recording digital data of skin conductance. A computer with a standard monitor. Soap, paper cloth, cup electrodes Ag / AgCl.
Procedures: We used a test-retest methodology. To this end we resort to a presentation of a sequence of 20 letters for 30s. each, without intervals between them. The presentation was performed using Microsoft © PowerPoint 2002 SP3. The subject had selected one of the cards before the trial. Participants had been instructed to "lye" throughout the procedure, saying "this is not the chosen card" before ALL presentations.
RESULTD:
The results show statistically significant differences between the situations in which the subject is telling the truth and those in which he "deceits". Differences were found for the rise-time t(542)=-2.444, p=0.015 and for the amplitude t(33.889)=-3.041, p=0.005. It was constructed a decision algorithm in order to distinguish “truth” from “deceit”. This provided significant results in relation to the difference between truth and deception (t (33.403) =- 10.409, p = 0.000). As a test, we used the same values, but this time trying to predict the card in which he would be deceiving us. Individually we were able to identify 71% of the times in which the subject was deceiving. Combining test data with retest data we were able to identify 82% of the time in which he tries to deceive.
CONCLUSIONS:
The SGR signal collected from the created situation allows distinguishing the true situation of the deceit situation. It was also possible to create an algorithm that allowed the automatic detection of situations of deception. This algorithm will be expanded and improved to integrate other psychophysiological measures in order to become more efficient.
PUBLICATIONS:
1. Rodrigues, P. (2007). Psicologia experimental: Um olhar psicofisiológico. Comunicação apresentada na conferência «Um olhar pedagógico sobre a psicologia», Universidade da Beira Interior – Covilhã.
2. Rodrigues, P., Silva, C., Moura, J., Paixão, R., Nascimento, C. (2008). Psicofisiologia e Engano: Efeito do conhecimento culpado (estudo piloto) no congresso anual da Associação Portuguesa de Psicologia Experimental: Universidade de Faro
3. There are still being prepared two papers.
|
Reference:
| Silva, C. F., Paixão, R., Rodrigues, P., & Silvério, J. (2010). Psicofisiologia e detecção do engano: paradigma do conhecimento culpável [Psychophysiology and detection of delusion: paradigm of the guilty knowledge]. In Aquém e além do cérebro. Behind and beyond the brain. Proceedings of the 8th Symposium of Fundação Bial (pp. 193-194). Porto: Fundação Bial.
|