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

The daily use of a variety of communication technologies, such as email, instant messaging, and the mobile phone, is now a fact of life for an increasing number of people, both at work and in their personal life. At work these technologies are relied upon for a wide range of tasks, such as collaborating with colleagues, reporting to superiors and communicating with customers. At home they are used to stay in touch with family and friends, and for involvement in local communities.

As communication technologies become more ubiquitous in our daily interactions, an important question is raised: how does the design of these technologies affect the phenomenon of lying? Lying is an important, and frequent, part of everyday social interactions.

Research from social psychology suggests as many as one third of typical daily interactions involve some form of deception. This can be defined as a “deliberate attempt, without forewarning, to create in another a belief which the communicator considers to be untrue.” DePaulo and her colleagues, for example, have observed that university students report telling about two lies a day, while non-student populations report about one a day.

The types of lies observed in these studies vary, from small “white” lies, in which inconsequential lies are told to be tactful or polite (such as saying “I love your haircut” when in fact you do not), to more serious lies (such as denying an affair).

How does the increased use of communication technologies affect these kinds of deceptions in our day-to-day social interactions? The design of various technologies creates very different communicative environments that may have important implications for lying behaviour. The telephone, for example, allows people in different physical locations to communicate with vocal and prosodic cues intact, while text-based media such as email and instant messaging, eliminate or distort nonverbal channels. We asked whether speakers were more or less likely to lie on the phone, in an email, or during an instant messaging exchange than they are face-to-face. Are different types of lies more likely to be told in one medium than in others?

We have examined deception in the three most commonly used daily communication media, the telephone, email and instant messaging, in an effort to determine how the design of these technologies affects lying behaviour.

Participants were students drawn from upper-level Communications courses at a north-eastern American university, and they all participated for course credit. There were 28 subjects: 13 males, 17 females, with an average age of 21. They reported lying about 1.6 times a day on average, and about one out of every four of their interactions involved a lie, replicating DePaulo’s original estimates, which indicated that students lied 1.9 times a day on average, and that a third of their social interactions involved some deception.

The primary objective of our study, however, was to determine the effect of the design of different communication media on lying behaviour during everyday social interactions. Although the total number of lies was greatest in the face-to-face setting, the highest proportion of lies occurred in telephone conversations, with 37 per cent of phone interactions involving some deception, significantly higher than lies in face-to-face conversations (27 per cent).

Two prominent theories suggest that a single underlying dimension influences the probability of deception. Media Richness Theory predicts speakers will choose the richest media, (specifically face to face) to lie most frequently. However, we found this was not the case; telephone interactions involved significantly more lies than face-to-face interactions, suggesting that a media’s richness is not the primary factor operating in lying behaviour across media.

The Social Distance Hypothesis argues that speakers will choose less-rich media when engaging in deception in order to avoid the discomfort associated with lying. Contrary to the Social Distance Hypothesis, in our study significantly fewer lies were reported in the least rich media, email, and no difference was observed between instant messaging and face-to-face. It seems the social distance of communication technology, and the relief from the discomfort of being deceptive that it may provide, does not predict everyday lying behaviour.

Both theories are overly simplistic. Communication media can be differentiated along a number of design features that are not captured by either richness or social distance.

Our model predicts that according to the degree in which a medium 1) allows for synchronous interaction, 2) is recordless, and 3) is distributed (that is, the parties are not in the same place), the greater the frequency of lying that should occur in that medium.

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.