An Empirical Exploration of Countermeasures in HCI-Based Deception Research
AuthorByrd, Michael David
Psycho-Physiological Deception Detection
Attentional Control Theory
Signal Detection Theory
MetadataShow full item record
PublisherThe University of Arizona.
RightsCopyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction, presentation (such as public display or performance) of protected items is prohibited except with permission of the author.
AbstractThis paper addresses three objectives: first, the extent of the theoretical understanding of countermeasures that is present in the deception detection literature to date was mapped out by conducting a literature review of countermeasures related work in the deception detection discipline. Second, after evaluating and analyzing this literature, Signal Detection Theory (SDT) was leveraged to generate an enhanced and extended theory-based framework for countermeasures. Third, an experiment was designed and conducted to explore the implications of this theory-based framework in the context of an HCI-based deception detection system based on tracking mouse movements and Attentional Control Theory (ACT) in an empirical experiment. The experiment was designed to learn more about what happens when users are aware they are being monitored and identify potential ways to mitigate any such countermeasures they may employ. In the experiment, participants were able to decide to perform and unsanctioned malicious act. In addition, we were able to definitively establish the ground truth about their behavior without imposing monitoring that was too overly invasive to the point of discouraging them from performing the malicious act. Mouse tracking was then used to attempt to detect who chose to perform the act, in a manner similar to how such a system would be deployed in practice. We manipulated the level of user awareness of the tracking and trained the users in strategies that can function as countermeasures to detection. Our analysis let us see how effective the system is at the varying levels of awareness and explore explanations and data analysis techniques to detect and mitigate the countermeasures. Results are discussed and considerations for future research are presented.
Degree ProgramGraduate College
Management Information Systems
Degree GrantorUniversity of Arizona
Showing items related by title, author, creator and subject.
A cultural model of nonverbal deceptive communication: The independent and interdependent self-construals as predictors of deceptive communication motivation and nonverbal behaviors under deceptionKam, Karadeen Y. (The University of Arizona., 2003)Findings from a host of prior cross-cultural studies suggest that those of differing cultural orientations are likely to possess differing motivations for deceiving and truthtelling, and as a consequence, are likely to exhibit differing patterns of behavior when engaging in deceptive communication. Thus, this investigation examined: (a) the impact of cultural identity on one's motivation for deceptive communication, and (b) the impact of cultural orientation on overt manifestations of behavior. In addition, this study investigated the effects of culture and relational familiarity (i.e., strangers versus friends) on truth bias and deception detection accuracy. To test the proposed theoretical relationships, participants from two cultures (i.e., United States and Japan) were employed in an experimental study. Results of the current investigation revealed that degree of independence was the single best predictor of one's motivation to tell the truth and one's motivation to protect the self, whereas degree of interdependence was the best predictor of one's motivation to protect the other. In terms of deceivers'/truthtellers' perceptions of the self under deception, higher interdependence scores were found to be related to self-perceptions of less positive affect, less fluency, and less psychological involvement under truth conditions, but were associated with greater positive affect, greater fluency, and more psychological involvement under conditions of deception. When considering partner perceptions of truthtellers'/deceivers' behavior, higher degrees of independence were found to be associated with less positive affect under deception. When outside-observers viewed the behaviors of truthtellers/deceivers, higher degrees of independence were found to be associated with greater kinesic involvement and pleasantness, less nervousness, and greater vocal pleasantness and vocal relaxation under truth. Conversely, higher scores on independence were found to be related to less kinesic involvement, less pleasantness, greater nervousness, and less vocal pleasantness and vocal relaxation under conditions of deception. Finally, relationship type was not found to be a significant predictor of either accuracy or truth bias, although, higher degrees of interdependence were associated with lower detection accuracy and greater truth bias. The findings of the current investigation strongly suggest that behavioral differences indeed become manifest when research is conducted employing samples of varying cultural orientations.
THE DEFINITION AND FUNCTION OF DECEPTIVE COMMUNICATION IN A LABOR-MANAGEMENT NEGOTIATION SESSION (LYING, BELIEFS, GENUINE).TUCKER, ROBERT E. (The University of Arizona., 1984)This study attempts to bring together the fields of communication, business, philosophy and linguistics. This effort is to make known particular behaviors exhibited in a negotiation setting with a focus on understanding why these behaviors are permitted and even embraced. The communicative dimension of these behaviors warrants study within a Speech Communication Department. The primary purpose of this study was to identify the conditions necessary for untrained observers to predict successful deception in labor-management negotiations. A secondary purpose was to discover observers' descriptions most commonly associated with negotiators. The dependent variable was an observer's judgment that deception would succeed in the negotiations context, measured by a 7-point Likert-type scale. The primary set of independent variables were HEARER-BELIEFS of a speaker's intent, the propositional content of the utterance made and the speaker's sincerity, measured by a 7-point Likert-type scale. The secondary set of independent variables were ten (10) plausible characteristics of a negotiator represented as bi-polar adjectives, measured by a 7-point semantic differential scale. The statistical procedures included a three-way analysis of variance (ANOVA), Scheffe's Test and a factor analysis. The three-way ANOVA allowed for presentation of the three (3) conditions in combination, permitting maximal detection of interactions. Also, the three-way ANOVA allowed for the introduction of and adjustment for each of the four (4) covariates. A single condition emerged as necessary for observers to predict successful deception in a labor-management negotiation session (content, p < .001). No single configuration of HEARER-BELIEFS was significantly more important in a prediction of successful deception than any other configuration. The quality of genuineness emerged as a primary criterion by which negotiators were described. Union leaders were evaluated by a secondary criterion of motivation. The belief-states of negotiators is far from being thoroughly understood. Research on deceptive communication, sensitive to context, should focus on the process of how humans predict successful deception. A cognitive index of deceptive communicative acts is proposed and a framework for future research is discussed.
Indicator of Deception: Response Time When Inquired By an Automated Screening KioskHsu, Irvin (The University of Arizona., 2012-05)An Automated Screening Kiosk (ASK), a non-invasive screening system, is used to conduct an altered Concealed Information Test (CIT), collecting measures to detect deceit. One of the many objectives of using the ASK is to determine whether response time is a viable measure for use in a CIT system. The time it takes for participants to respond to questions asked during the test could signal deception. The results from the coding of response times confirm the results from previous studies showing that guilty individuals take more time to respond when they see a threatening stimulus. More research needs to be done before we can accurately gauge the validity of using response time as a measure for detecting deception.