We are more likely to lie over the phone than in an email

Our study shows that telephone interactions, which are distributed, synchronous and recordless, had the greatest rate of deception. Face-to-face interactions, which are synchronous and recordless but not distributed, involved the next highest rate of deception, while email, which is distributed but not synchronous or recordless, had the lowest rate of deception. Instant messaging, which did not differ in the rate of lying from face-to-face interactions, is distributed and nearly synchronous, but is recorded in a log file.

Our study suggests that the design of communication technology, such as email, instant messaging and the telephone, has an impact on everyday lying behaviour. It is important to note that these technologies are not obscure, or limited to only highly sophisticated users, but instead are used by millions of people across the globe on a daily basis. As such, these data have important implications for those of us who use these technologies to accomplish our everyday communicative activities at work and at home. In particular, the results indicate that we are more likely to lie (and to be lied to) on the telephone than in any other medium, and to lie the least in email.

What guidance does this research offer for developers and managers? Our findings suggest that specific design features can be used to influence rates of everyday forms of deception. For example, if the objective in a given situation is to reduce overall deception during social interactions, then a communication medium should be used that is asynchronous and recordable. Also, technologies that specify the user’s current physical activity, such as videoconferencing, should reduce deception by reinstating the constraints of co-presence. If, for whatever reason, the goal is to facilitate deception, then our model suggests these factors should be reversed.

According to recent studies, mutual trust develops most slowly in text-based interactions. Our experiment suggests that, somewhat ironically, participants lied least frequently in text-based interactions, especially email. How this irony will resolve itself over time is an interesting question. Will people begin to lie more frequently as they become more comfortable with text-based media? Or, as communication via text becomes increasingly ubiquitous, will people come to trust others more readily in their text-based interactions? Research that examines how the design of everyday communication technologies affects these types of interpersonal processes will become increasingly important as these technologies continue to support more and more of our social interactions.