Psychologie a její kontexty, Vol.5, No.2

Detekce lži na základě neverbálních projevů - studie zaměstnanců Policie České republiky
Lie detection based on nonverbal expressions - study of the Czech Republic Police employees

Hedvika Boukalová, Lenka Mynaříková

Detekce lži prostřednictvím neverbálního chování není standardní metoda, ale jde o intuitivní proces, aplikovaný laiky i profesionály. V rámci naší studie nás zajímá, jak se bude dařit v rámci detekce lži na základě neverbálních projevů specifické profesní skupině - příslušníkům Policie ČR. Výsledky byly získány od 197 zaměstnanců Policie ČR v různých funkcích. Byla jim prezentována sada 21 videozáznamů dlouhých 30 vteřin a byli požádáni o hodnocení ukázky z hlediska pravdivosti či lživosti sdělovaného obsahu. Výsledky ukázaly, mimo jiné, výskyt tzv. lie bias, neboli zkreslení obvyklého právě u policejních a bezpečnostních pracovníků, které zahrnuje zvýšené očekávání lži u osob v rámci trestního řízení. Tato a další zjištění by mohla být zahrnuta do policejního výcviku zaměřeného na komunikaci a lež v průběhu trestního řízení, přičemž obdobný systematický výcvik v České republice chybí. Informace bychom rády do budoucna srovnaly s výsledky dalších profesních skupin.

Klíčová slova:
detekce lži, policie, neverbální projevy, paraverbální projevy

Lie detection based on nonverbal behavior is not a standard method, it is an intuitive process, applied by lay persons, but also professionals. Some of the major sources (e.g. widespread Interrogation Manual by F. Inbau et al., 2004) offer clear recommendations about the nonverbal behavior of liars to investigators of serious crime. These findings are not supported by the research, moreover they can lead to lowering the ability to detect lie (Blair, Kooi 2004). Another topic is mapping the skills of professionals (police officers, members of the secret services) and non-specialists to detect lies by nonverbal signs. Across the studies (with few exceptions) a low performance in the task of detecting lies by nonverbal expressions (Ekman P., 1996; Vrij, 2004 and others) is found. The levels of success are usually around the level of chance. The potential reasons for such results are analyzed (e.g. Blair, Kooi, 2004). However a group of psychologists led by P. Ekman and M. O'Sullivan (O'Sullivan, 2007) managed to find in their years lasting research a group of people whose ability to detect lies is well above the population average. This group is diverse in terms of age, interests and professions, all of them come from the USA. There were certain common features found in this group and also a focus on similar phenomena in the detection of lying.
The main goal and research question is to find out: what is the success rate of differentiation between lies and truths in this specific professional group of Czech population, is it the same or different from the results reported in the context of available resources. The research will focus on the ability of respondents to determine the truth or deceit on the basis of non-verbal and paraverbal expressions of observed subjects, with focus on specific professional groups - mainly police workers. We assume, that the police officers are frequently in the contact with people, who are not willing to reveal critical information or who are motivated to lie.
The procedure: The material consisted of short video sections originating from real investigation of serious crime. After the pilot study the stimuli material has been created consisting of 21 video-sections lasting 30 seconds, the sound of the videos was filtered, so the paraverbal characteristics were maintained, but the content of the speech was unclear. The research sample consisted of 197 police officers from 5 departments.
Selected results: The TRUTH detection accuracy - (the accuracy of correct evaluation of the truth-telling people on the video) was situated between 0,25 (25%) to 0,75 (75%). The LIE detection accuracy - was situated between 0,44 (44%) to 1 (100%). The overall accuracy (means combined accuracy of lie and truth detection) was situated between 0,33 (33%) and 0.86 (86%). 5 respondents obtained results, that were above the average of the group in all the three categories (truth, lie, overall).
The discussion: The results indicate that (among others) we can find so called lie bias among the police workers, which is common in the similar professional settings and it involves higher expectations of lie and deceit on the side of the communication partner. In the common population the opposite tendency - so called truth bias - tendency to evaluate others (situations, behavior, statements) as truthful is found. The study managed to find police workers, who obtained better results than guessing and also better than the average of the tested group. The main limit of the study is probably the lack of repeated retest of the sample. This and other findings were involved into the police training focused on communication and lie during criminal investigation, which is not systematic in Czech Republic. The plan is to test also other professional groups in the future.

Keywords:
lie detection, police, nonverbal behavior, paraverbal behavior