Cosmic Conservative

Rational Conservatism from the heart of America

How not to do a survey

By CosmicConservative
I apologize for linking to a story of a survey on sexual behavior, but I want to use it as an example of why you should be very skeptical of survey results when reported by the news media. After several paragraphs discussing the behavior of survey respondents, and many lurid lead sentences and headlines about how prevalent prostitution is among "men who already have partners" at literally the very last sentence of the article we learn this:
The authors used survey data from a standard health screening questionnaire completed by more than 2,500 men at one sexual health clinic between October 2002 and February 2004.
This is what is known as a "biased sample." In fact it is not merely biased by nature, it is biased by action. What sort of person typically goes to a "sexual health clinic"? You can be pretty sure it doesn't include many monogamous people, male or female. But does the story let you know that while you are reading about the apparent out of control sexual behavior of men? No, because if it did, you would probably stop reading immediately. Typical.

Comments (0)

No comments on this post.