Psychologist Justin Kruger, PhD, has seen plenty of e-mails gone awry. Kruger, a professor at New York University, was once a member of a psychology departmental e-mail list at a different university. A job candidate came into town to interview for a faculty position. The faculty member responsible for organizing a meet-and-greet dinner sent around an e-mail invitation that read "talking to the candidate is not required; just don't embarrass us."

"She meant it as a joke, but much to her surprise some people were really upset," says Kruger. "It was a comical miscommunication."

Now, Kruger and his colleague Nicholas Epley, PhD, of the University of Chicago, have published research that helps explain why these electronic misunderstandings occur so frequently. In a study in the December Journal of Personality and Social Psychology (Vol. 89, No. 5, pages 925–936), they find that people overestimate both their ability to convey their intended tone–be it sarcastic, serious or funny–when they send an e-mail, as well as their ability to correctly interpret the tone of messages others send to them.

The reason for this communication disconnect, the researchers find, is egocentrism–the well-established social psychological phenomenon whereby people have a difficult time detaching themselves from their own perspectives and understanding how other people will interpret them.

And as e-mail has become more prevalent, Epley says, the opportunities for misunderstanding have increased.

"Of course there's nothing new about text-based communication; people have been writing letters for centuries," he explains. "But what's different in this medium is…the ease with which we can fire things back and forth. It makes text-based communication seem more informal and more like face-to-face communication than it really is."

What do you mean?

Despite this ease, though, e-mail can have some serious disadvantages. In their recent study, Kruger and Epley found that people are better at communicating and interpreting tone in vocal messages than in text-based ones. In one experiment, the researchers tested 30 pairs of undergraduate students. Each participant received a list of 20 statements about topics like campus food or the weather.

One member of each pair read their statements into a tape recorder–taking either a sarcastic or serious tone–while the other member e-mailed the statements. The participants also noted whether they thought their partners would correctly interpret each statement's tone.

The participants then listened to or read their partners' statements, guessed the intended tone and indicated how confident they were in their answers.

Both the e-mailers and those who recorded their messages were highly confident that their partners would correctly detect their tone–both groups predicted about a 78 percent success rate. The speakers weren't too far off–their partners got the tone correct about 75 percent of the time. The partners who read the statements over e-mail, though, had only a 56 percent success rate–not much better than chance.

What's more, the participants who received the messages were no better at predicting their own success–both the listeners and the readers guessed that they had correctly interpreted the message's tone 90 percent of the time.

"I think people do have some intuition, abstractly, about the limits of e-mail," Epley says. "But I don't think that in specific instances people realize that a particular message is unclear."

In a follow-up experiment in the same paper, Kruger and Epley dug deeper to uncover the reasons behind e-mailers' overconfidence. They suspected that it might be because e-mailers assume that other people have the same inside information about their intentions and motivations that they do–what social psychologists call egocentrism.

Such an effect was found in a 1990 study by psychologist Elizabeth Newton, PhD. For her dissertation, Newton asked participants to tap the rhythm of a well-known tune. The tappers predicted that listeners would be able to identify the songs 50 percent of the time, whereas in reality the listeners could only figure out the tune about 3 percent of the time. The reason for the disconnect, Kruger says, is that tappers would inevitably "hear" the whole, orchestrated tune in their minds as they tapped, whereas listeners heard only an irregular series of taps.

"It's impossible not to hear the song as you're tapping away," says Kruger. "So you have a hard time separating yourself from your own perspective and realizing how impoverished the listeners' data really are."

Similarly, he says, e-mailers might inevitably "hear" the tone they intend their e-mail to convey, while forgetting that receivers don't have access to that extra information.

To test this, he and Epley repeated their first experiment, but this time asked the e-mailers to read their statements aloud before sending them. Half the participants read the statements as they intended them, while half read them using the opposite tone: seriously for a sarcastic statement, and sarcastically for a serious statement. The idea, Epley says, was to force participants to step outside their own perspective, perhaps negating some of the effects of egocentrism.

And indeed, that's what happened: Participants who read the statements as they intended them still overestimated receivers' ability to guess the e-mails' meanings, but participants who read the statements using the opposite tone no longer did.

Pick up the phone

Kruger and Epley's research adds a new level of rigor to previous speculations about e-mail communications, says Lee Sproull, PhD, a sociologist who studies technology and communication at New York University.

"What I like about this study is that it applies well-understood ideas about egocentrism and social judgment," she says. "To the best of my knowledge, they are the first to apply those findings to e-mail, and they really reinforce and explain many of the phenomena found in previous studies."

Given these findings, then, what's the average e-mailer to do? Well, perhaps just pick up the phone, says Epley. "E-mail is fine if you just want to communicate content, but not any emotional material," he explains.

Or, suggests Sproull, try the researchers' manipulation on your own–read your message out loud in various tones of voice, to see how your recipient might interpret it.

Overall, she says, she finds the study very informative.