Full length articleContext in a bottle: Language-action cues in spontaneous computer-mediated deception
Introduction
As users of computer-mediated communication (CMC) are exposed to an increasing number and variety of risks associated with online deception, it has become more and more important—and increasingly challenging—for users to guard against deceptive activities (e.g., fake review, opinion spam, spear phishing, identity fraud). The key is the ability to correctly interpret and codify the underlying intent of the message sender. In face-to-face (F2F) communication, the assessment of context is informed by words that are accompanied by other physical cues such as body language and facial expressions. CMC, in comparison, is “cue lean.” Thus, to derive context, a message receiver has only words to consider during a message exchange. Context is often difficult to ascertain; however, as Dourish (2004) states, context is a relational—or even an occasioned property. Context is not static, but dynamically defined by activity and the associated interaction. In CMC, we may only derive context (and hence intent) from the interaction itself—including the words used, and any other associated non-verbal cues. Thus, our research question is: How do language-action cues derive context, and further the identification of spontaneous computer-mediated deception?
In this paper, we first discuss the creation of interaction context through language-action cues, and then argue that language-action cues can reveal and visualize communicative intent. Different types of communication modes and media can have an impact on computer-mediated deception, which is further examined. Our research considers and models a sociotechnical system that represents synchronous computer-mediated interaction. An online game was designed and developed to simulate interpersonal computer-mediated deception in pairwise interaction. We collected and analyzed the data to uncover patterns that can provide subtle cues to deceptive intent. Specifically, linear, logistic regression and mixed model ANOVA approaches were deployed to analyze the efficacy and accuracy of certain language-action cues for detecting computer-mediated deception in synchronous, spontaneous communications. The paper concludes with some reflections on implications, limitations, as well as potential directions for future research.
Section snippets
Computer-mediated deception
Online communication is interpersonal and interactive. It involves a sender and one or more receiver(s) who are engaged in an (more-or-less) interactive exchange. Each iteration of an interaction between a sender and a receiver helps to shape the context of communication. Within this exchange, there is often an opportunity for a message sender to influence the receiver(s) actions or beliefs. Accordingly, Miller, Deturck, and Kalbfleisch (1983) described deceptive communication as “… a general
Research design
To investigate context-sensitive and spontaneity in computer-mediated deception, we modeled interaction to better understand the role of context (Dourish, 2004). An online game system was designed and developed to mimic spontaneity in interpersonal deception scenarios, and to capture players’ interactive conversations during pairwise interaction. This system creates a conceptual basis for further exploration of the dynamics of intentional deception. The use and efficacy of language-action cues
Method
A mixed method was employed in this research. Qualitative data (i.e., players’ conversations and interactions) was captured and analyzed quantitatively.
Data analysis
After the data had been cleaned, the linguistic cues from the data were extracted according to the categories established in the Linguistic Inquiry and Word Count (LIWC) tool (Newman et al., 2003; Pennebaker & King, 1999).
Summary and discussion
One of our objectives in undertaking this study was to demonstrate the efficacy that language-action cues can collectively represent in the context of information exchange, which can be indicative to revealing a deceiver. We first performed linear regressions to examine the efficacy of the dataset. Three models were generated, and each model consists of different combinations of language-action cues, thereby demonstrating the efficacy of contextual illustration. Language-action cues are able to
Limitations and future work
Our research demonstrates the effectiveness of adopting Google + Hangout as a communication platform for this experiment; however, technical limitations were also experienced. In particular, players sometimes experienced technical problems in logging into the Google + pseudo accounts created for the game, and when launching the game interface. These difficulties not only confused and distracted players, but also reduced the overall amount of time spent in the game itself, thus reducing our
Contributions and conclusion
Communicators’ actions are embedded in their language-action cues in the iteration of message exchange, and these actions define the interactivity that shapes the dynamic context of computer-mediated communication. From a holistic perspective, the words and other non-verbal language-action cues, as well as the mode and medium used for communication, create the context of that interaction—which can be a viable measure to derive the communicative intent of interacting parties. Rather than a
Acknowledgements
The authors gratefully acknowledge the grants from National Science Foundation (#1347113 and #1347120, 09/01/13–08/31/15), Florida Center for Cybersecurity Collaborative Seed Grant (#2108-1072-00-O, 03/01/15–02/28/16), and Florida State University Council for Research and Creativity Planning Grant (#034138, 12/01/13–12/12/14). The authors acknowledge and appreciate the research efforts and contributions from Cheryl Booth, Sai Surya Shashanka Timmarajus, Kashyap Vemura, and Aravind Hariharan.
Shuyuan Mary Ho is an associate professor at School of Information, at Florida State University. Her research focuses on computer-mediated deception, trusted human computer interaction, cloud forensics, online games for research, and sociotechnical behavioral experiments. Her work appears in over 40 journal articles and conference proceedings and has been funded by U.S. National Science Foundation.
References (48)
- et al.
Cultural determinants fo media choice for deception
Computers in Human Behavior
(2012) - et al.
Website credibility and deceiver credibility: Expanding prominence-interpretation theory
Computers in Human Behavior
(2016) - et al.
Sender credibility and deception detection
Computers in Human Behavior
(2014) - et al.
Cross-cultural deception in social networking sites and face-to-face communication
Computers in Human Behavior
(2008) - et al.
Everyday deception or a few prolific liars? The prevalence of lies in text messaging
Computers in Human Behavior
(2014) - et al.
Would I lie to you? Self-serving lies and other-oriented lies told across different media
Computers in Human Behavior
(2008) Linguistic theories of humor
(1994)- et al.
Deception: Strategic and nonstrategic communication
Strategic Interpersonal Communication
(1994) - et al.
Interpersonal deception theory
Communication Theory
(1996) - et al.
Interpersonal deception: III. Effects of deceit on perceived communication and nonverbal behavior dynamics
Journal of Nonverbal Behavior
(1994)
Interpersonal deception. IV. Effects of suspicion on perceived communication and nonverbal behavior dynamics
Human Communication Research
Automated deception detection of 911 call transcripts
Security Informatics
Message equivocality, media selection, and manager performance: Implications for information systems
MIS Quarterly
Serious lies
Basic and Applied Social Psychology
Everyday lies in close and causal relationships
Journal of Personality and Social Psychology
Lying in everyday life
Journal of Personality and Social Psychology
Cues to deception
Pyschological Bulletin
What we talk about when we talk about context
Journal of Personal and Ubiquitous Computing
Nonverbal leakage and clues to deception
Psychiatry
Who can catch a liar?
American Phychologist
Butler lies: Awareness, deception and design
On lying and being lied to: A linguistic analysis of deception in computer-mediated communication
Discourse Processes
Deception and design: The impact of communication technologies on lying behavior
The truth about lying in online dating profile
Cited by (0)
Shuyuan Mary Ho is an associate professor at School of Information, at Florida State University. Her research focuses on computer-mediated deception, trusted human computer interaction, cloud forensics, online games for research, and sociotechnical behavioral experiments. Her work appears in over 40 journal articles and conference proceedings and has been funded by U.S. National Science Foundation.
Jeffrey T. Hancock is a professor of in the Department of Communication at Stanford University. He works on understanding psychological and interpersonal processes in social media by using computational linguistics and behavioral experiments to examine deception and trust, emotional dynamics, intimacy and relationships, and social support. His research has been published in over eighty journal articles and conference proceedings, and has been supported by funding from the U.S. National Science Foundation and the U.S. Department of Defense. His work has also been featured in the popular press, including the New York Times, CNN, NPR, CBS and the BBC.