Seeing the Truth in Teen Sex Surveys
MELISSA BLOCK, host:
When surveyors ask teens about their sex lives, how do they know they're getting honest answers? Researchers have a number of tools to validate their findings. NPR's Allison Aubrey reports that getting good information from teens starts with the right environment.
ALLISON AUBREY reporting:
Years of survey research show that teens are more likely to be forthcoming if they're convinced that no friends, parents or teachers will ever see their answers. This is particularly true when it comes to sex and drugs. So in this latest survey teens were given headphones and a laptop. They responded to a series of sensitive questions by typing in their answers.
Ms. MOLLYANN BRODIE (Kaiser Family Foundation): There is no other human being who is hearing what they say, and there's no chance that somebody is overhearing what they say. And it feels more confidential; it feels more anonymous.
AUBREY: Mollyann Brodie directs survey research for the Kaiser Family Foundation. Her group was not involved with this latest survey but conducts similar ones. She says it's important when talking to teens to clearly explain to them why they're being asked.
Ms. BRODIE: That they understand that their views and their experiences are important for a broader purpose.
AUBREY: In this case, to help public health officials understand teen behavior and perhaps help protect future teens. These techniques are no truth serum. Researchers know that some teens do overestimate their experiences; others underestimate.
Mr. LLOYD JOHNSTON (Director, Monitoring the Future Survey): There are a number of ways that we try to validate the data, partly because of our own skepticism initially.
AUBREY: Lloyd Johnston directs the Monitoring the Future Survey at the University of Michigan, which tracks teen drug use. He explains one way to check for truthfulness is to ask similar questions in a variety of ways, then look for inconsistencies.
Mr. JOHNSTON: We generally identify between 1 and 3 percent of our respondents whose data don't seem to pass various tests for consistency. And we usually remove them from the data before we generate the statistics.
AUBREY: There is no foolproof way to validate responses since scientists can't observe teens' sex and drug use. The only way to get a handle on prevalence is through these self-reporting surveys. Confidence in the findings ultimately depends on whether the results can be repeated.
Dr. FRED BRYANT (Psychologist, Loyola University Chicago): Anytime you find a result that's surprising or that's new, you can't necessarily know for sure that it's trustworthy.
AUBREY: Fred Bryant is a psychologist at Loyola University Chicago, who's trained in survey research. He says the teen sex survey released today is one of the first to ask about oral sex, so replicating results will be important.
Dr. BRYANT: They'll ask additional questions in different ways to see if that matters. They'll try different samples. They'll probably try asking peers of people, `What are your peers doing?' They're going to find out a lot more about this to see if all of the results converge on the same conclusion.
AUBREY: Kaiser's Mollyann Brodie says the new sex survey may not have precisely nailed the percentage of teens who've had oral sex; that 55 percent could easily be 49 or 51 percent.
Ms. BRODIE: What's really important is: How does that 51 differ from sort of the same question asked five years ago? How does that 51 differ from other behaviors or other attitudes that you're measuring at the same point in time? It's seeing that number in context, which really, I think, gives the number meaning.
AUBREY: The new numbers released today will help researchers establish meaning. They'll set a baseline from which results of future surveys can be measured. Allison Aubrey, NPR News, Washington.
ROBERT SIEGEL (Host): This is NPR, National Public Radio.