KALSBEEK: So does that possibly mean that the convergence that you anticipate—and I tend to agree with you—might be more a matter of the technology generation getting older and beginning to move into the population. Is age a factor in the things Mick is talking about; do you expect these effects more because use of these tools is more of a way of life for younger people?

TOURANGEAU: I like that speculation, too … I would anticipate convergence, or even a cross-over, happening in the younger age groups. Yes?

PARTICIPANT: What is the length of these Web surveys before you start to see a lot of break-off?

TOURANGEAU: I don’t think that there’s a magic formula here. I think a lot of it has to do with what you anticipate. For instance, we did a study—this is actually using interactive voice response, telephone Audio-CASI. Well, actually, we used the long-form questions from the census, the decennial. And people broke off whenever they got to a new person; you know, a light went off, “[Uh-oh], 35 dull questions about to come.” And then they hung up. And I think you’ll see similar phenomena when it comes to the Web; when people can anticipate a lot of tedious questions, you’ll have break-offs. On the other hand, I’ve also done some surveys where you had no progress indicator, and it was just one item after another, using the interactive format, and you’re just totally clueless about how far along you are, and I—at some point, that’s not a successful strategy either, and there, it just might be a function of time rather than anticipation. Miron?

STRAF: If I could ask another question, on the collection of sensitive information … some of what you’re showing seems to show that there aren’t major desirability effects and the like. But are there alternatives, and here I’m thinking of things like randomized response, using the Web?

TOURANGEAU: I’m not sure. All of our results tend to come from comparisons of different Web interfaces. It’s conceivable that people didn’t, weren’t sensitive to the different interfaces because in every condition they’re underreporting sensitive information. You know what I mean? In fact, there’s one study done by Moon, one of the Nas collaborators, who asked the same questions but portrayed the computer that the respondents were using as either a standalone computer, not hooked up to any kind of network, or part of a network and the questionnaire resided on a computer across town. Or on a network and the questionnaire resided across the country. And you got more socially desirable responding when it was viewed as being across country, hooked up to a computer across country. And it might be that people, when they’re using the Internet, are very sensitive to the fact that people could be