27 January 2014

Don’t Trust Statistics



The joke’s on a generation of human-sexuality researchers: Adolescent “pranksters” responding to the widely cited National Longitudinal Study of Adolescent Health in the mid-1990s may have faked “nonheterosexuality.”
Preliminary results from the landmark study – known as “Add Health” – stunned researchers, parents and educators alike, recalls Cornell’s Ritch C. Savin-Williams, professor of human development: “How could it be that 5 to 7 percent of our youth were homosexual or bisexual!” Previous estimates of homosexuality and bisexuality among high schoolers had been around 1 percent.
So imagine the surprise and confusion when subsequent revisits to the same research subjects found more than 70 percent of the self-reported adolescent nonheterosexuals had somehow gone “straight” as older teens and young adults.
“We should have known something was amiss,” says Savin-Williams.  “One clue was that most of the kids who first claimed to have artificial limbs (in the physical-health assessment) miraculously regrew arms and legs when researchers came back to interview them.”

Self-reporting on sex is clearly faulty.  Self-reporting on drugs probably isn’t much more reliable.  I’m speaking from experience on this one, as I recall being given a drug questionnaire to fill out when I was a junior in high school.  According to the survey I filled out, I was a daily habitual user of all narcotics except for marijuana, which I was allergic to.  Of course, I never did any drugs in high school (I was so much of a square that I didn’t even give tobacco a try until I was nearly twenty, and I didn’t start drinking until last year), but that didn’t stop me from skewing results in the most hilarious ways I could think of.  Most of the other guys in my class filled their forms out like I did.  The only guys who didn’t claim massive drug use were, ironically, the actual drug addicts, as they thought the government would track them down based on their results.

At any rate, any statistics based on self-reported surveys is probably suspect, if said survey is asking questions about sex, drugs, or any other subject ripe for humor.