As a teacher in a public school, each spring our students were given an anonymous survey asking questions on illegal drug use, thoughts of suicide, exposure to violence, and so on.  But the category that always received the most attention from the students was the one with questions on sexual abuse.

These surveys were usually given in the spring at the end of the school day when student behavior can start to decline.   After the surveys were turned in, the conversation among the students was ALWAYS about the questions dealing with sexual abuse.

I would hear them say (particularly the girls) that they had given answers such as having been raped several times or having been sexually assaulted in some other way.  The discussions always came with giggles.  It appeared to me that the survey was one big joke to my students, who were junior high school age.  Similar surveys are given at the high school level.

My takeaway was always that I hoped no statistician took these surveys too seriously and understood what junior high students can be like. The one thing I knew for sure was that the surveys were NOT reliable.  You do not walk into a public middle school, junior high school, or high school and hand the average student such a survey and expect all of them to take it seriously, particularly at the end of the school day.

So, it was no surprise to me to read at Reason.com that the Centers for Disease Control and Prevention had inflated data about teen girls and sexual assaults.

The percentage of girls reporting risky behavior did increase in the most recent surveys, but not by the magnitude that the CDC had released.

According to Reason.com, the problem with the CDC’s methodology is that a growing number of “schools surveyed refused to ask students questions about sexual violence.”  Washington Post fact checker Glenn Kessler said, “That could have biased the sample by possibly removing jurisdictions with lower rates of reporting rape and sexual violence.”  It would definitely have skewed the results.

Another problem with the CDC’s data is with their rounding.  Rather than use percentages such as 11.4 and 13.5, they were rounding to the nearest whole number, making 11.4% become 11% and 13.5% become 14%.  The math teachers of these CDC statisticians would be proud of their rounding ability, but every time you round a number to a higher place value, you are getting further and further away from the exact answer.

The difference between 11.4% and 13.5% is 2.1%; the difference between the rounded 11% and 14% is 3%.

The CDC reported a 27% increase in the number of girls in 2021 who reported being forced to have sex as compared to 2019.  The actual increase ended up being 18.4%.

This inflated data is primarily caused by some schools allowing students to answer questions about reporting being raped while other schools did not, by too much rounding, and by the unreliable samples used when administering the survey.