Continuing a Rant

[chris_gerrib]

In yesterday's post, I mentioned that Americans at least are horrible at assessing risk. I think a major contributing factor is what I will refer to as "statistical illiteracy" or the inability to understand data presented statistically.

Don't get me wrong - until I went to graduate school, I was ignorant of statistics. One can get a degree in history (or many other liberal arts degrees) without taking one statistics class. Even though the Navy (as part of ROTC) required additional math and science classes, statistics wasn't part of the requirement.

I got exposed to statistics at SIU Carbondale for my MBA program. It was an eye-opener. Now, the class was "statistics for dummies" and designed for Education grad students, but I learned a ton. Such things as:
* Correlation is not causation.
* There are degrees of correlation (weak to strong).
* A mean and a standard deviation tells you much more about a study's results then an average.

I could go on. My point is this - in modern society, risk is expressed statistically. For example, "you have X% chance of catching skin cancer," or "Y% of the men that eat lots of tomatoes won't get prostate cancer." In short, it's rather more complicated then "look out for that tiger!"

But since we don't require statistics for college degrees, people don't understand stats. Including people like journalists, who are reading the press releases from the latest study. It's like a tone-deaf person trying to be a music critic. One ends up focusing on things other then the music.

The educational establishment is partially to blame. Just like you can be a music critic (or at least appreciate music) without being able to perform music, you can understand statistics without being able to crank out massive equations. But "statistics for dummies" classes are at best well-hidden secrets, at worse non-existent.

From a public policy viewpoint, an inability to assess risk leads to "policy by hyperbole." Whomever can make the loudest or most exaggerated claims wins the policy debate. I think we need to fix statistical illiteracy.

Comments

[j-cheney.livejournal.com]

I really wish more high schools would offer statistics. Most offer only Calculus in upper level math, but statistics is honestly far more useful. (Having taught both. I love Calculus, but it's much easier to bludgeon others with Statictics.)