The ability to detect deception is of critical value in criminal and investigative contexts. This study has investigated beliefs about deception detection held by police officers (N = 63) and police students (N = 130). The results show that there are inconsistencies when comparing the beliefs to empirical research findings. One example is the belief that liars avert their gaze. The results are discussed and contrasted with the Truth-Default Theory. Instead of a focus on cues that are probabilisticallyassociated with deception, the Truth-Default Theory focuses on contextualized communication content. The theory recognizes that people are truth biased. Truth-Default Theory proposes that reliance on cues pushes the accuracy of deception detection to the level of chance.

Deception detection has always been of subject of interest. After all, determining if a person is telling the truth or not could be detrimental in many real-world cases. Current methods to discern deceptions require expensive equipment that need specialists to read and interpret them. In this article, we carry out an exhaustive comparison between 9 different facial landmark recognition based recurrent deep learning models trained on a recent man-made database used to determine lies, comparing them by accuracy and AUC. We also propose two new metrics that represent the validity of each prediction. The results of a 5-fold cross validation show that out of all the tested models, the Stacked GRU neural model has the highest AUC of.9853 and the highest accuracy of 93.69% between the trained models. Then, a comparison is done between other machine and deep learning methods and our proposed Stacked GRU architecture where the latter surpasses them in the AUC metric. These results indicate that we are not that far away from a future where deception detection could be accessible throughout computers or smart devices. / Revisión por pares

No description available.

No description available.

While much research has shown that some linguistic features can indicate a person is lying, this line of research has led to conflicting results. Furthermore, very little research has been done to verify that these supposed linguistic features of deception are universal. In addition, few studies have researched the cross-cultural perceptions of deception, which knowledge could greatly improve the detection of deception across cultures. The current study addresses these gaps in the literature by analyzing and comparing truthful and deceptive discourse of eight native English-speaking Americans and eight non-native English-speaking Ghanaians. The discourse was elicited in one-on-one interviews where each interviewee spontaneously responded to questions about themselves. Later, interviewee responses were judged by 47 native English-speaking Americans and 35 non-native English-speaking Ghanaians. The results showed that Americans and Ghanaians lie differently—Americans' lies were more superfluous and redundant; had more pronoun inconsistencies, adjectives, adverbs, and modal verbs; and had fewer negative emotion words than their truths. Ghanaians' lies, on the other hand, also had more pronoun inconsistencies but had fewer negations than their truths. Furthermore, the groups' baseline speech differed in superfluousness, positive emotion words, word count, and response latency. Regarding detecting deception, Ghanaians were slightly more accurate and significantly more confident in detecting lies than Americans. Both groups were slightly more accurate and confident in judging the veracity of statements within their own cultures. Neither group, however, demonstrated truth- or lie-bias cross-culturally. These results have implications for law enforcement investigators and analysts who can learn the differences between Americans' and Ghanaians' truthful and deceptive speech so as to more accurately detect deception through language. In addition, cross-cultural deception perception research can improve cross-cultural communication and understanding.

This dissertation thesis called 'Lie Detection Through Verbal and Nonverbal Behavior' deals with lying and lies from the viewpoint of its types, functions and considering the development of an individual and it describes the role of lying in our lives. Lying itself is not considered as a negative term, but as a type of communication, a social phenomenon, which is a part of our everyday existence. Attention is aimed to verbal behaviour, but especially to nonverbal aspects of lying persons. Then it examines the use of information on lying and lies in the system of forensic psychology during the deception detection tests. It describes historical approaches to lie detection and its current appearance as well as some other aspects, that may be considered lying even if they are not lying at all - false memories or suggestions. The empirical part compares two research designs, one of them trying to find possible personal determinants of successful lie detectors on a group of psychology students and connecting lie detection ability with the concept of emotional inteligence. The second research analyzes the ability to detect deception among Czech police officers as a specific population group, that uses lie detection methods as a means of investigation. Keywords: lying, deception detection, nonverbal...

Tant lors de procès en droit criminel que lors de procès en droit administratif, civil ou familial, le mensonge peut contribuer à la mise en place de trames narratives qui embrouillent les faits que les juges devraient connaitre afin d’appliquer adéquatement le droit. De plus, si les juges ne sont pas capables de distinguer le vrai du faux lors des témoignages, la confiance du public indispensable au bon fonctionnement du système de justice pourrait être compromise. Par ailleurs, même si des chercheurs s’intéressent, depuis les années 1960, à la détection du mensonge, en particulier lors d’interrogatoires policiers reproduits en laboratoires, ces expérimentations ignorent plusieurs attributs des systèmes de justice accusatoires (p. ex., les procédures judiciaires, les éléments de preuve incriminante, les déclarations assermentées, les interrogatoires, les contre-interrogatoires et les plaidoiries). Par conséquent, puisque la détection des témoignages mensongers par les juges lors de procès est rarement explorée, les juges des tribunaux municipaux, provinciaux et fédéraux sont pratiquement laissés à eux-mêmes. Dans le cadre de ma thèse, je présente un travail de type descriptif, analytique et explicatif pour observer et mieux comprendre le raisonnement ou, du moins, une partie du raisonnement par lequel des juges, en contexte naturel, en sont venus à déterminer, selon leur point de vue, que des témoins ont menti alors qu’ils témoignaient. Cette recherche vise, entre autres, à bonifier la validité écologique des expérimentations en laboratoire et à s’attaquer plus adéquatement à la problématique des mensonges lors de procès. D’une part, afin d’étudier la détection du mensonge dans le contexte concret où elle a lieu plutôt que dans le cadre d’une représentation expérimentale de cette activité, j’ai fait le choix de me tourner vers des jugements écrits de tribunaux canadiens comme données de recherche, plus spécifiquement, quatre jugements canadiens portant sur une accusation criminelle de parjure résultant d’une déclaration incriminante où les accusés ont été déclarés coupables. D’autre part, afin de décrire, analyser et expliquer de façon minutieuse et approfondie les mécanismes communicationnels sous-jacents aux déclarations de culpabilité pour parjure, j’ai fait le choix de mobiliser une approche contemporaine d’analyse du discours s’inscrivant dans les travaux de l’École de Montréal—l’approche ventriloque de la communication (Cooren, 2013)—dans la mesure où celle-ci permet d’identifier les éléments qui apparaissent comme faisant la différence dans la constitution d’un jugement. Hormis la rareté alarmante des jugements rendus à la suite de procès sur une accusation criminelle de parjure résultant d’un témoignage, les résultats de mes analyses permettent d’observer le rôle substantiel des idées reçues et l’apport considérable des précédents dans le processus décisionnel des juges lors de procès. Comme je le montre dans ma thèse, ces idées reçues peuvent exposer des croyances sur le comportement humain véhiculées ou exprimées, explicitement ou non, par les juges dans leur jugement. Quant aux précédents, ils peuvent avoir l’effet de voiler le rôle substantiel de ces mêmes idées reçues dans le processus décisionnel des juges. De plus, les résultats de mes analyses suggèrent que la détection des témoignages mensongers par les juges lors de procès dépend, entre autres, de nombreux éléments du contexte tels que mis en scène dans le raisonnement des juges et n’a pas grand-chose à voir avec le scénario expérimental typique mis en place pour étudier la détection du mensonge. Finalement, bien qu’ils n’offrent pas de « nouvelle technique » pour distinguer l’honnête justiciable du menteur, les résultats de mes analyses s’inscrivent en continuité avec d’autres travaux de recherche qui démontrent l’utilité a priori limitée des indicateurs comportementaux de mensonge et qui suggèrent que les praticiens du droit, en particulier les juges, devraient être systématiquement formés sur la nature faillible du processus décisionnel lors de procès. / Both in criminal law trials and in administrative, civil or family law trials, deception can lead to the development of narrative patterns that obscure facts that judges should know in order to properly apply the law. In addition, if judges are unable to distinguish truth from falsehood in testimonies, the public trust essential to the proper functioning of the justice system may be compromised. Furthermore, even if researchers have, since the 1960s, been interested in detecting deception, especially during police interrogations conducted in laboratories, these experiments ignore several attributes of adversarial justice systems (e.g., court proceedings, incriminating evidences, sworn statements, examinations, cross-examinations and pleadings). As a result, since the detection of deceptive testimony by judges during trials is hardly ever addressed, judges in municipal, provincial and federal courts are practically left on their own. In my thesis, I present a descriptive, analytical and explanatory work to observe and better understand the reasoning, or at least part of the reasoning, by which judges, in a natural setting, have come to determine, from their perspective, that witnesses have lied while testifying. This research aims, among other things, to improve the ecological validity of laboratory experiments and to address more adequately the problem of deception during trials. On the one hand, in order to study the detection of deception in the actual context in which it takes place rather than in the context of an experimental representation of this activity, I have chosen to turn to written judgments of Canadian courts as research data, more specifically, four Canadian judgments pertaining to a criminal charge of perjury resulting from an incriminating statement in which the accused was found guilty. On the other hand, in order to describe, analyze and explain in a thorough and in-depth manner the underlying communicative mechanisms of perjury convictions, I have chosen to use a contemporary approach to discourse analysis that is part of the work of the Montreal School—the ventriloquist approach to communication (Cooren, 2013)—to the extent that this approach allows to identify the elements that appear to make a difference in the constitution of a judgment. Aside from the alarming scarcity of judgments rendered following trials on a criminal charge of perjury resulting from a testimony, the results of my analyses make it possible to observe the substantial role of popular beliefs and the considerable contribution of precedents in the judges’ decision-making process during trials. As demonstrated in my thesis, these popular beliefs may expose assumptions about human behaviour conveyed or expressed, explicitly or not, by judges in their judgments. As for precedents, they may have the effect of veiling the substantial role of these very beliefs in the judges’ decision-making process. Furthermore, the results of my analyses suggest that the detection of deceptive testimony by judges during trials depends, among other things, on several contextual elements as staged in the judges’ reasoning and has little to do with the typical experimental design set up for studying deception detection. Finally, although they do not offer a “new technique” for distinguishing the honest litigant from the liar, the results of my analyses are consistent with other research that demonstrates the a priori limited usefulness of behavioural indicators of deception and suggests that legal practitioners, particularly judges, should be systematically trained on the fallible nature of the decision-making process during trials.

En la vida cotidiana, detectar una falacia puede tener importantes implicaciones en distintas situaciones sociales. Descifrar mentiras, puede ser determinante en situaciones que impliquen consecuencias graves o moderadas; como el caso de investigaciones policiales. El trabajo expuesto en las siguientes paginas tiene como fin la realización de un sistema de detección de mentiras que utilice una cámara web como medio único para la detección. Además de esto, se busca realizar la investigación correspondiente a las subáreas relacionadas al problema. Estas son la de detección de mentiras, Deep learning y visión computacional. En este trabajo expuesto, se asumirá al acto de mentir como cualquier acto que busque comunicar información falsa o trastornada, de forma deliberada con la finalidad de engañar a otros. La investigación realizada, se hará presente en el desarrollo de un proyecto cuyo alcance consiste en la creación de una aplicación capaz de detectar si una persona dice la verdad a partir de su reconocimiento facial. Para ello, se utilizarán técnicas de visión computacional y machine learning con el fin de dar otra opción más económica y accesible ante las otras metodologías (polígrafo, ERPs, fMRI) que se basan en analizar el estado cerebral requieren de maquinaria extremadamente costosa y tienden a tener la misma precisión que el uso de polígrafos. / In everyday life, detecting a fallacy can have important implications in different social situations. Deciphering lies can be decisive in situations that involve serious or moderate consequences, as in the case of police investigations. The work presented in the following pages is aimed at the realization of a lie detection system that uses a web camera as the only means for detection. In addition to this, it seeks to carry out the investigation corresponding to the subareas related to the problem. These subareas are lie detection, deep learning, and computer vision. In this exposed work, the act of lying will be assumed as any act that seeks to communicate false or disturbed information, deliberately with the purpose of deceiving others. The research carried out will be present in the development of a project whose scope consists of the creation of an application capable of detecting if a person is telling the truth from their facial recognition. To do this, computer vision and machine learning techniques will be used in order to provide another cheaper and more accessible option compared to other methodologies (polygraph, ERPs, fMRI) that are based on analyzing the brain state, require extremely expensive machinery and tend to have the same precision as the use of polygraphs. / Trabajo de investigación

Financial fraud rampages onwards seemingly uncontained. The annual cost of fraud in the UK is estimated to be as high as £193bn a year [1] . From a data science perspective and hitherto less explored this thesis demonstrates how the use of linguistic features to drive data mining algorithms can aid in unravelling fraud. To this end, the spotlight is turned on Financial Statement Fraud (FSF), known to be the costliest type of fraud [2]. A new corpus of 6.3 million words is composed of102 annual reports/10-K (narrative sections) from firms formally indicted for FSF juxtaposed with 306 non-fraud firms of similar size and industrial grouping. Differently from other similar studies, this thesis uniquely takes a wide angled view and extracts a range of features of different categories from the corpus. These linguistic correlates of deception are uncovered using a variety of techniques and tools. Corpus linguistics methodology is applied to extract keywords and to examine linguistic structure. N-grams are extracted to draw out collocations. Readability measurement in financial text is advanced through the extraction of new indices that probe the text at a deeper level. Cognitive and perceptual processes are also picked out. Tone, intention and liquidity are gauged using customised word lists. Linguistic ratios are derived from grammatical constructs and word categories. An attempt is also made to determine ‘what’ was said as opposed to ‘how’. Further a new module is developed to condense synonyms into concepts. Lastly frequency counts from keywords unearthed from a previous content analysis study on financial narrative are also used. These features are then used to drive machine learning based classification and clustering algorithms to determine if they aid in discriminating a fraud from a non-fraud firm. The results derived from the battery of models built typically exceed classification accuracy of 70%. The above process is amalgamated into a framework. The process outlined, driven by empirical data demonstrates in a practical way how linguistic analysis could aid in fraud detection and also constitutes a unique contribution made to deception detection studies.

This study examined the ability of students to correctly discern between truthful and deceptive messages from a group of second language English speakers. Recent studies have found a ‘lie bias’ when making veracity judgments towards second language speakers. This lie bias may be problematic in a country such as South Africa where the majority of the population communicate, to a greater or lesser extent, in their second language. In this study participants (n=64) made classifications of 24 messages as either truthful or deceptive. The messages were created by second language English speakers who were asked to describe a truthful or deceptive event. The results revealed that there was a significant difference between the way that first and second language participants made their judgments. It is argued that this difference can be attributed to the notion that second language participants require more cognitive effort, than their first language counterparts, to understand and classify messages. / Psychology / M. A.(Psychology)