Study secretly tracked cellphone users

Researchers secretly tracked the locations of 100,000 people outside the United States through their cellphone use and concluded that most people rarely stray more than a few miles from home.

The first-of-its-kind study by Northeastern University raises privacy and ethical questions for its monitoring methods, which would be illegal in the United States.

It also yielded somewhat surprising results that reveal how little people move around in their daily lives. Nearly three-quarters of those studied mainly stayed within a 20-mile-wide circle for half a year.

That might seem like science and data being marshaled to prove the obvious. But the researchers said their work, which also shows that people exhibit similar patterns whether they travel long distances or short ones, could open new frontiers in fields such as disease tracking and urban planning.

"Slices of our behavior are preserved in these electronic data sets," said Albert-Laszlo Barabasi, an author of the project and the director of the Center for Complex Network Research at Northeastern University in Boston. "This is creating huge opportunities for science."

The study found that nearly half of the people in the study largely kept to a circle little more than six miles wide and that 83 percent largely stayed within a 37-mile-wide circle.

Nearly 3 percent of the people regularly went beyond a 200-mile-wide circle. Less than 1 percent of people traveled often out of a 621-mile circle.

Movement patterns

The researchers said they used the potentially controversial data only after any information that could identify individuals had been scrambled. Even so, they wrote, people's wanderings are so subject to routine that by using the patterns of movement that emerged from the research, "we can obtain the likelihood of finding a user in any location."

That type of nonconsensual tracking would be illegal in the United States, said Rob Kenny, a spokesman for the Federal Communications Commission (FCC). Consensual tracking, however, is legal and marketed as a special feature by some U.S. cellphone-service providers.

The study, published today in the journal Nature, opens up the field of human tracking for science and calls attention to what experts said is an emerging issue of locational privacy.

The researchers were able to obtain the data from a European provider of cellphone service that was obligated to collect the information. By agreement with the company, the researchers did not identify the provider or disclose the country where the provider operates.

The researchers, including Barabasi's Northeastern colleagues Marta Gonzalez and Cesar Hidalgo, tracked 100,000 cellphone users selected at random from a population of 6 million for six months.

The location of the user was revealed whenever the user made or received a call or text message; the telephone company would record the nearest cell tower and time.

Because calls and messages tended to be sporadic, the researchers used a smaller data set that captured the location of 206 users every two hours. The results of the two data sets were similar, according to the report.

Privacy implications

The use of cellphones to track people, even anonymously, has implications for privacy that make this "a troubling study," said Marc Rotenberg, a founder of the Electronic Privacy Information Center in Washington, D.C.

The university said the U.S. Navy-funded experiment went to a Navy institutional review board, which determined the mobility study did not involve human research subjects. Bioethicist Arthur Caplan at the University of Pennsylvania disagreed with that board's determination.

"There is plenty going on here that sets off ethical alarm bells about privacy and trustworthiness," Caplan said.

Studies done on normal behavior at public places is "fair game for researchers" as long as no one can figure out identities, Caplan said in an e-mail.

"But my cellphone is not public. My cellphone is personal. Tracking it and thus its owner is an active intrusion into personal privacy."

Study co-author Hidalgo said there is a difference between being a statistic — such as how many people buy a certain brand of computer — and a specific example. The people tracked in the study are more statistics than examples, he said.

"In the wrong hands the data could be misused," Hidalgo said. "But in scientists' hands you're trying to look at broad patterns ... We're not trying to do evil things. We're trying to make the world a little better."

This story, published June 5, 2008 was corrected on June 21, 2008. A previous version of the story, The Associated Press, relying on information from a study co-author, erroneously reported that no ethics review of the study was required. The university says the U.S. Navy-funded experiment did go to a Navy institutional review board, which determined the mobility study did not involve human research subjects.