Abstract

This project examines the impact of communication technology on lying behavior, and whether automated analysis techniques can be used to detect online deception.

Issue

Deception is one of the most significant and pervasive social phenomena of our age. On average, people tell one to two lies a day, and these lies range from the trivial to the more serious, including deception between friends and family, in the workplace, and in politics. At the same time, information and communication technologies have pervaded almost all aspects of human communication and interaction, from everyday technologies that support interpersonal interactions, such as email and instant messaging, to more sophisticated systems that support organizational interactions.

Given the prevalence of both deception and communication technology in our personal and professional lives, an important set of questions have recently emerged about how technology affects digital deception, which refers to any intentional control of information in a technologically mediated message to create a false belief in the receiver of the message. The first question is concerned with how communication technologies affect the production of deception. Do people use different media to lie about different types of things, or to different types of people? The second question concerns our ability to detect deception across various media. Are we worse at detecting a lie in a text-based interaction than we are in face-to-face? Finally, can automated techniques be developed to analyze the linguistic properties of deceptive and truthful messages?

Response

To address these issues I have conducted three types of studies. First, I have conducted a series of series of diary based studies to examine the first question of how, when, and about what people lie about in different media. Second, I've conducted several experimental laboratory studies that examine the detection of deception during truthful and deceptive conversations in the lab. Third, I have conducted automated linguistic analyses of text-based conversations in an effort to identify linguistic patterns that differentiate truthful from deceptive conversations. To this end, I have begun to develop the Deceptive Message Corpus, a large-scale, database of several hundred thousand messages coded for their deceptiveness. The ultimate objective is to develop automated tools to assist in the real-time detection of digital deception based on statistical learning models and linguistic analyses of deceptive communication in online environments

Impact

This research has increased our understanding of how communication technology affects deception. This research has also had a large impact in the popular press that has perhaps affected the way that people think about the Internet and its relationship to deception. In particular, this line of research was featured in the New York Times Magazine ("The honesty virus," Mar 21, 2004), and was covered in New York Times (Mar 2, 2004), the New Scientist (Feb 12, 2004), ABC News (Feb 24, 2004) and the Washington Times (Feb 20, 2004) among others. I was interviewed about the research on CNN, and the research has been covered in over a dozen countries, including Canada, England, Chile, Scotland, Denmark, Holland, Wales, Germany, South Africa, India, Pakistan, Japan, and Australia, in over 100 print and web pieces. The research has been discussed in over 20 radio interviews, including the BBC and NPR, and has been reported on over 450 stations in the U.S.

Funding Sources

• Federal Formula Funds - Research (e.g., Hatch, McIntire-Stennis, Animal Health)
• start up funds

Collaborators

• Michael Woodworth - Okanagan University College

Key Personnel

• Jeffery Hancock, Department of Communication, Cornell University