It was all in the name of research. Students in David Rappaport's social studies classes had been studying effects of opinion polls on the American consciousness, and Rappaport invited Frankovic to share her expertise with his eighth-graders at a lunchtime question-and-answer session.

One student was keen to be on the receiving end of a call from one of Frankovic's interviewers, who conduct telephone surveys used by The New York Times and by CBS news programs such as 60 Minutes and 48 Hours.

"Do you ever survey kids?" he asked.

"Being called for a survey is pretty rare, and you'll have to be a little older," Frankovic told him, adding that respondents must be at least 18 years old.

CBS did poll children for their reactions after the space shuttle Challenger exploded in 1986, killing teacher Christa McAuliffe and the shuttle's six other crewmembers.

The news organization also surveyed high school students for its Class of 2000 project, tracking the students from their freshman year through graduation. In response to another student's question, Frankovic said this survey is among her favorites.

"In an election year, you're always asking people how they're going to vote. It gets a little tiring," she explained. "It's very exciting when you get to explore a new topic with a group you don't usually talk to."

As part of their scrutiny of opinion polls, Bret Harte students examined whether respondents given a multiple-choice question are likely to choose their answers based on the order of the options presented. Frankovic said phone survey respondents are indeed more likely to choose either the first or the last option they hear. To counteract this "primacy and recency effect," she added, interviewers will scramble the order of the options from survey to survey.

CBS interviewers are also instructed to ask open-ended questions. When asking people how they're going to vote, they always ask respondents to give their approval rating of the president first.

"Don't ask, 'You do approve of the way the president is doing his job, don't you?'" Frankovic said. "We don't want to influence the thinking of the person we're talking to."

Frankovic, who has been with CBS since 1977, said her team conducts an average of 35 surveys a year. "In my first year [on the job], we did five," she told the students. "We're rarely doing two things at once."

The most surveys CBS ever conducted in a year was 50 in 1998. "That was the year of Monica Lewinsky," Frankovic explained, adding that news organizations wanted to ensure that their finger was on the pulse of public opinion on that issue. The percentage of respondents who favored impeaching former President Clinton versus those who didn't "never changed in the course of that year," she said.

One student was as concerned with the financial aspects of polling as he was with their impact on society. "You make a lot of long-distance calls. Doesn't that cost a lot of money?" he asked.

Frankovic said her team spends less than $1 per phone call when conducting a survey. "It's actually a lot cheaper than sending a camera crew out," she added.