Would you protect your computer’s feelings? Clifford Nass says yes

[From the Chronicle of Higher Education’s ProfHacker blog]

September 22, 2010

Would You Protect Your Computer’s Feelings? Clifford Nass Says Yes.

By Jason B. Jones

What if there was a book that explained how to write end comments on student papers or exams; why peer review processes often avoid, rather than facilitate, sound judgment; how to encourage meaningful group work; and why academic events feature so much ritual flattery? Clifford Nass’s The Man Who Lied to His Laptop: What Machines Teach Us About Human Relationships doesn’t restrict itself to academe—indeed, it claims to offer social rules for almost any situation—but it has a wealth of provocative experiments that any professor might want to reflect upon.

Clifford Nass is the Thomas M. Storke Professor at Stanford University; his home department is communications, but he has numerous courtesy appointments, and, crucially, is the founder of the Communication between Humans and Interactive Media Lab. He has helped many companies design interactive elements—including, as he recounts here, an attempt to make Clippy less reviled! The Man Who Lied to His Laptop condenses for a popular audience an argument that Nass has been making for at least 15 years: humans do not differentiate between computers and people in their social interactions.

At first blush, this sounds absurd. Everyone knows that it’s “just a computer,” and of course computers don’t have feelings. And yet. Nass has a slew of amusing stories—and, crucially, studies based on those stories—indicating that, no matter what “everyone knows,” people act as if the computer secretly cares. For example: In one study, users reviewed a software package, either on the same computer they’d used it on, or on a different computer. Consistently, participants gave the software better ratings when they reviewed in on the same computer—as if they didn’t want the computer to feel bad. What’s more, Nass notes, “every one of the participants insisted that she or he would never bother being polite to a computer” (7).

Nass was called in by BMW, as the company was tormented by support calls from German men who could not accept GPS directions from a female voice, even though, obviously, the directions were being calculated by a computer.

In another study, Nass found that users given completely random praise by a computer program liked it more than the same program without praise, even though they knew in advance the praise was meaningless. In fact, they liked it as much as the same program, if they were told the praise was accurate. (In other words, flattery was as well received as praise, and both were preferred to no positive comments.) Again, when questioned about the results, users angrily denied any difference at all in their reactions.

I can, though, think I see Nass’s point. I’m far more scandalized by the concept of pornography on the iPad or iPhone, for example, than by the concept of pornography as such. Over here in my head—which is, I’ll stipulate in advance, a bad place to be, it seems more provocative to share those sorts of desires on those sorts of devices.

In The Man Who Lied to His Laptop, Nass reports on numerous studies involving social rules or expectations that were studied using human-computer interaction. Rather than focus on the implications of this research for software/hardware design, Nass here focuses on what we might learn about interacting with others.

For example, anyone’s who ever graded a paper or exam is familiar with the so-called “criticism sandwich,” where you start with some good things, then move on to the criticism, and then conclude on a more positive note. Nass demonstrates, however, that people internalize praise and criticism differently—while we welcome the former, we really dwell on and obsess over the latter. In the criticism sandwich, then, “the criticism blasts the first list of positive achievements out of listeners’ memory. They then think hard about the criticism (which will make them remember it better) and are on the alert to think even harder about what happens next. What do they then get? Positive remarks that are too general to be remembered” (32).

And because we focus so much on the negative, having a similar number of positive and negative comments “feels negative overall” (33). The best strategy, he suggests, is “to briefly present a few negative remarks and then provide a long list of positive remarks…You should also provide as much detail as possible within the positive comments, even more than feels natural, because positive feedback is less memorable” (33).

The entire chapter about praise and criticism is fascinating from an academic perspective. It makes a difference whether praise or blame is delivered with an expectation that people are basically fixed, or capable of growth. (If the latter, then it’s possible to turn criticism to productive use. If the former, then what’s the point?) On the one hand, self-praise is bad—no one likes a braggart—but, on the other, so is modesty—people believe you when you’re self-critical.

Other chapters are almost equally relevant: discussions of team-building, for example, have insights about group work, and the discussion of persuasion is at least indirectly relevant to anyone trying to convince an indifferent student that their subject matter’s worth studying. There are excellent experiments that demonstrate the ways in which stereotyped expectations about gender, race, and accent shape our reception of information. All of these have important implications for classroom performance, for academics’ ability to communicate with the outside world, and for our ability work together.

A few caveats: Nass is probably a bit more confident than non-social scientists about the validity/applicability of sociological insights into personality and character. (NB: I’m not saying these insights don’t exist!) To take just one example, people who aren’t already persuaded of the interpretative utility of, say, Meyers-Briggs ratings probably won’t be convinced by a personality grid that maps everyone on two axes. Also, the first fifty pages in the book seem to have overlooked its own strictures against self-praise.

The Man Who Lied to His Laptop is both almost as fun as its title and far more practical for academics than I’d expected. And it is a very quick read, to boot—perfect for in-semester reading.

Related Links:

This entry was posted in Presence in the News. Bookmark the permalink. Both comments and trackbacks are currently closed.
  • Find Researchers

    Use the links below to find researchers listed alphabetically by the first letter of their last name.

    A | B | C | D | E | F| G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z