VitalSTATS

Facts Hazy on High School Hazing

by Howard Fienberg
September 2000

Is hazing a ritualized form of bonding or an invitation to abuse? An August 29 Alfred University study of practices in American high schools said hazing was an epidemic of harassment and assault, not isolated exercises in camaraderie. Headlines reflected the finding that more than 1.5 million students have been subjected to hazing: “Students are humiliated, beaten to join even church groups” (Associated Press) ; “Hazing a regular part of high school experience, survey says” (Boston Globe) ; “Hazing is thriving in high schools, too, study shows” (Los Angeles Times).

Hazing was defined as “any humiliating or dangerous activity expected of you to join a group, regardless of your willingness to participate.” Hazing includes humiliation and coerced or forced substance abuse or criminal activity. Some examples in the report, such as urine drinking and mass beatings, would make even the Animal House crowd cringe. More importantly, high levels of hazing were reported for just about every conceivable extracurricular group except the nerds at the school newspaper and yearbook.

Fortunately, the evidence for a national hazing epidemic was awfully hazy. While some reporters explained how the survey was conducted, few hinted at its serious methodological deficiencies.

The researchers mailed a two-page questionnaire to a random sample of 20,000 high school juniors and seniors and received 1,541 useable responses. If they discussed the response rate, most journalists marked it at 8.28 percent. Unfortunately, that rate was calculated after subtracting from the original sample the 1,092 questionnaires returned by the post office and marked ‘undeliverable.’ A more accurate response rate would be an even smaller 7.71 percent.

STATS commends reporter Jerry Zremski of the Buffalo News for actually addressing the minuscule response rate. He quoted Harvard University researcher Sid Groeneman, who said that “when you get much below a response rate of 25 percent, any serious researcher would be skeptical of anything at all that the survey found.”

Given the low response rate, the possibility of non-response error is high. University of Buffalo sociologist Simon Singer told Zremski that “the kids who are angry or are victims could be more likely to respond to the survey.”

The Alfred University researchers admit that “high school students have a consistently low rate of response on mail surveys. The low response rate begs for further studies to confirm or refute, and further refine these findings.”

Don Dillman of Washington State University, an expert on mail surveys and response rates, took a much harsher stance when speaking to STATS. “This study does not appear to be well enough done to have the results taken seriously. At best the study suggests an area in which further inquiry is needed. The study needs to be redone, using better methods, before trying to draw conclusions.”

Before printing the results of this study and treating them as representative and valid, journalists should have run the study through a hazing, spanking it all the way down the line. Science requires certain hazing-like rituals before research can be considered valid. The media could learn a lot from that process.


return to Howard Fienberg's page