BOB GARFIELD: Recent polls tell us that American media habits are changing -- turning away from broadcast TV and newspapers to cable and the internet. But how exactly do they know? Researchers at Ball State University set out to find how much media, including TV, radio, computers, books, even the telephone, their Indiana neighbors were consuming. But they were equally interested in testing one key methodology used by other pollsters -- the phone interview. Telecommunications professor Bob Papper, who co-authored the Middletown Media Studies report, says his team followed up its own series of phone interviews by going a step further.

BOB PAPPER: Indeed, we went two steps further. We did a diary study where we asked people to keep a log of their media use for a day, and we also did an observational study.

BOB GARFIELD: You had people in their homes?

BOB PAPPER:In their homes, from just after they woke up in the morning, as early as we could get there, until bedtime at night - we followed them to work, we followed them, you know, in the car, to parties -- whatever they did, we were there.

BOB GARFIELD:And the ultimate discovery was that the respondents dramatically under-reported in the telephone interviews and even in their diaries the amount of media that they were consuming.

BOB PAPPER: Yeah. In fact, in the observational study, people used more than twice as much media as they said they did on the phone -- 129.7 percent more media use per day than they reported to someone on the phone.

BOB GARFIELD: To what do you attribute the under-reportage?

BOB PAPPER:There is a body of research, mostly in much smaller studies, that says people have a lot of trouble telling you about their behavior. It's one thing for someone to answer - this is how I feel about something - it's entirely different when you ask someone what did you do or not do? Now the other factor I suspect that's involved is that we all know that there's certain media that's judged as "good," and other media that it may be judged as "bad," and so there's some indications that we do a whole lot better job of estimating our time with what might be considered "good media," than we do with "bad media."

BOB GARFIELD:People understand they're, they're supposed to say they eat broccoli; not supposed to say they eat the junk food. On newspaper readership, people have been pretty accurate in their reporting?

BOB PAPPER: Well, what they told us on the phone was very, very close to what we actually observed in practice. The worst medium, as far as phone estimation is, is TV. It wasn't simply that people couldn't estimate how much time they spent with a medium. They weren't even very good at identifying whether they even used TV yesterday, when they did.

BOB GARFIELD: In your conclusion, you observed that media usage is probably less than the sum of its parts. What did you mean by that?

BOB PAPPER:Well, one of the things that we found is that almost a quarter of the media day, people are using two or more media at the same time. And, and we think about multiple media use among kids, but this population were all adults, so while we found that the average person spent 11.7 hours a day in media, if we summed the individual media, we'd be at 15.4 hours a day.

BOB GARFIELD:Now this has to be good news for the people in network television who have been observing what they have believed to be a steady erosion in their audience. Is this going to turn the advertising market upside down?

BOB PAPPER: Advertisers are interested because they always want to know what are they buying and what are they getting, and we've got some very different news, and media companies are obviously interested, because we're saying that a lot of the media measurement that we rely on isn't very accurate.

BOB GARFIELD:But what about the methodology of having people in the respondents' homes? Wouldn't that tend to skew the data as well by people being so aware of being observed that they find themselves gravitating to the use of media to help these nice folks who are doing the study?

BOB PAPPER: Yeah, it may alter it. One researcher who reviewed this, however, suggested you know people may be doing a lot less media because they're so conscious of it, and they're worried that they don't want to do too much. Margaret Mead inadvertently taught us that you cannot observe a population without altering its behavior at least some. As an example, no one in our population that we observed watched any pornography. Could be clean Midwestern living, but my guess is there's something else going on. I would still argue it's a considerably better measure of behavior than asking 'em on the phone.

BOB GARFIELD:If people's responses in telephone polls on the questions of their media consumption are skewed downward, it gets to a larger issue of, of how telephone polling is done on a whole range of subjects including who do they intend to vote for in the presidential race.

BOB PAPPER: You know, I suspect that one of the kind of dirty little secrets in research is that we always have known that there are problems in phone research, but it's so easy and it's so cheap that we keep doing it, without ever testing, you know, how valid is this? And so, you know, we've really opened up that can of worms.

BOB GARFIELD: All right. Well, Bob, thanks very much.

BOB PAPPER: My pleasure.

BOB GARFIELD:Bob Papper is professor of telecommunications at Ball State University and co-author of the report titled "Middletown Media Studies."

copyright 2004 WNYC Radio