Marion Underwood is drowning in teenage texting data.
For the last four years, the University of Texas at Dallas professor has been collecting texts sent by and to 175 adolescent students at a large suburban Texas high school as part of a study dubbed the BlackBerry Project.
Participating students receive free BlackBerrys from Underwood’s team, as well as a data plan that includes unlimited texting and a limited number of voice minutes.
In return, the young participants allow researchers to record all the texts and emails they send and receive from the phone, which they agree to use as their primary communication device.
As you might guess, the number of texts that Underwood and her team have collected is mind-numbing.
Over homecoming weekend of 2009 -- which included a big football game and a dance - -her team took in 43,305 texts over a two-day period.
On average, Underwood said she adds 500,000 new texts to her database each month.
“It is a vast amount of data,” she said in an interview with the Los Angeles Times.
It is also a potential privacy nightmare.
Underwood has been careful to make sure that the teenage participants are fully aware of the extent of the monitoring her team is doing. She has participating students sign a release allowing the researchers to access their texts each year, and every time the students turn on their cellphones they see a message that indicates that the texts will be monitored.
You’d think the students would self-censor their texts - -since they know Underwood and her students are watching -- but that doesn’t seem to be the case.
In an early look at the data, Underwood found that 7% of the texts contain profane language, and that 6.6% of messages contained sexual language, which is similar to what other researchers found when analyzing conversations in teenage chat rooms.
In fact, when a friend texted one participant about selling drugs, the participant wrote back, “Hey, be careful, the BlackBerry people are watching, don’t worry, they won’t tell anyone.”
Underwood has promised the participants that their privacy will be protected, but she has a researcher monitoring the stream of texts they send and receive each day to look for worrisome words like “rape,” “kill myself” or “older man.”
She has also intervened a handful of times when a student has run away from home.
So far, Underwood has been busy mostly with collecting the texting data rather than analyzing it.
And in fact, she says she has tried to limit the amount of time she spends reading the texts as they come in.
“I have someone who works for me who checks it everyday, but I don’t look at it too often because I’m immediately absorbed by it,” she said. “It is so rich and they say so much to each other, and they use such sophisticated language, it just sucks me in.”
But some new information has come to light. For example, reports that teen boys text less than teen girls appears to be incorrect.
Underwood’s data suggests that boys and girls text about the same amount, but that boys may be less likely to tell researchers how much they text when asked. (She has hypothesized that communicating a lot may be seen as a sign of girliness).
She also found that both boys and girls were very bad reporters of their texting behaviors, which makes the real-life monitoring of texts that Underwood is undertaking seem all the more important to paint an accurate portrait of teen texting behavior.
The information that Underwood has collected in the last four years could be useful to psychologists, sociologists and anthropologists, but it is unclear whether Underwood will ever be able to share this data with others.
“Whether I share this entire archive with anybody depends on it being completely de-identified,” she said, and unfortunately, all the security experts she’s spoken with say it is not possible.