They're mainly young, single and urban. They move frequently, usually renting rather than owning their homes. Pollsters call them "cell phone only" because they don't own traditional phones.
As this hard-to-track population grows, so does the problem of accurately incorporating them into polling and scientific surveys that seek to measure everything from health and business practices to political attitudes.
For now, the problem is most pressing in areas of research where the attitudes of young adults are sought, like surveys of radio listener preferences.
Paul Lavrakas, a public opinion researcher at Nielsen Media Research, organized a meeting in New York City this month where pollsters and others involved in survey research discussed ways to address the cell-phone-only issue. One major aim is to learn how to deal with this population before the next presidential election.
"What the industry doesn't know how to do is how to statistically balance those reached by cell phones with those reached on land lines," said Lavrakas.
The cell-phone-only population is growing fast. In 2001, it comprised about a half-percent of the population. Now it's estimated at 7 percent. Among people age 15 to 24, almost one in five have only cell phones, according to Clyde Tucker, a researcher at the Bureau of Labor Statistics.
The industry has to deal with the cell phone issue while it's still manageable or it will be doing that research work when cell phones have become a problem for surveys — "and that's not a good practice," said Mike Brick, a statistician at Westat, which handles scientific surveys for the federal government.
The cell-phone-only issue is just the latest challenge facing survey researchers. People have shown less willingness to talk to survey researchers, especially on the phone. Many polls are conducted on the telephone, the fastest and most efficient way to do a survey. But some surveys are done face to face.
Raj Naik, a 26-year-old employee of a software company who lives in Rosslyn, Va., is a good example of the "cell phone only" crowd.
He got rid of the traditional phone because most of the calls he received "were calls I didn't want to take." Naik said he wouldn't mind getting calls from survey researchers "if those calls aren't costing me anything, but as soon as someone calls me on my cell phone, it's costing me something."
That's part of the problem for pollsters: Cell phone users don't expect to get survey calls. When they do, many are uncooperative.
"People say, 'Do you know you're calling me on my cell phone?'" said Charlotte Steeh, a Georgia State researcher who has done experimental cell phone surveys. "Some people mentioned the privacy issue, saying, 'How did you get my number? I only give it out to family and friends.' Some just hang up immediately."
Rob Stuart, a 23-year-old personal trainer in Washington, said he doesn't see a need for a land line.
"I might get one if I had a family or someone else living with me or if I was working out of my house," he said.
Leading up to the 2004 presidential election there was concern polling would suffer because of difficulties reaching cell phone users. Surveyors accommodated for this by giving extra weight to young adults, and the phone polls turned out to be generally accurate.
"The industry dodged a bullet this time," said Scott Keeter of the Pew Research Center for the People & the Press.
Exit polls indicated the cell-phone-only group was not that different in political attitudes from young people generally. But as the cell-phone-only group grows, that may change.
Federal law doesn't allow researchers to call cell phones with an autodialer, a popular method for quickly reaching random samples for polls and surveys.
But survey researchers plan to start calling cell phone users manually, making those calls on weekends. They want to develop a plan of reimbursement for respondents' time or lost cell phone minutes and come up with a code of ethics and safety so research calls don't endanger cell phone users involved in activities like driving.
Patsy Frase, a resident of Granby, Mo., who is nearing retirement, said she doesn't want to get survey research calls on her cell phone because "we use it more for business than for pleasure."