Here at TTAG, we like to think we get things right most of the time. The rest of the time our readers correct us, and we make every effort to fix the issue in a timely fashion. One article slipped through the cracks a few months ago, and as I’m sitting here sending it to that great big circular file in the sky I just wanted to apologize on the behalf of TTAG for being asleep at the switch that day. Here’s what happened…

The post, titled “George Mason University’s Freedom Index is Anti-Gun,” was an analysis by one of the writers of a study done by George Mason University ranking the “Freedom” of the 50 States.  Click here for the study [pdf]. This ranking was done by assessing a number of different factors and mathematically comparing the scores. One of these factors was “Paternalism,” which is essentially a “Nanny State Index.”

Among the criteria for this factor were a number of gun control issues, some which our writer identified as “anti-gun.” The writer saw that a coding system was being used where states were awarded “points” for each of these items that they had in place and cried foul, calling the survey “anti-gun.”

What actually happened is a common problem called “polarity confusion.” Our writer believed that these points were directly boosting the state’s “freedom” ranking, meaning that the more gun control measures that were in place the more “free” the state would be. In fact these points were being used in a mathematical calculation to form a “paternalism” factor, which ultimately was a detractor from the “freedom” score.

The points were positive for a factor with a negative impact, leading the writer (untrained in the finer arts of arithmetical kung-fu) to form the “anti-gun” opinion that was presented.

As a confirmation of the effects of this factor, a quick look at the final freedom ranking is all that is needed.

TABLE V: OVERALL FREEDOM RANKING
1. New Hampshire 0.432
3. South Dakota 0.392
4. Idaho 0.356
5. Texas 0.346
[…]
46. Maryland – 0.4 05
47. California -0.413
48. Rhode Island – 0.430
49. New Jersey – 0.457
50. New York – 0.784

There were a good number of comments (specifically Jason’s comments) on the article pointing out the error in math, but to me Bob H’s sticks out.

Isn’t one of the TTAG writers an analyst in real life? Could we get his opinion on what the study is doing?

I completely missed this one. I dropped the ball. To be fair I was a little busy that day — a diabetic patient tried to die on me. But I fixed him, and I’m fixing this now.

The study in question is not “anti-gun.” Well, at least not that we can tell, and not for the reason stated in the first article. I apologize to GMU on the behalf of TTAG for getting this one wrong, and thank all of those who left comments correcting us.

If you want to read the original article, click here for a ZIP file with the article in a PDF document within. Sorry, it was the only way to keep Google’s prying eyes out of it. We now return you to your normally scheduled semi-arrogant snark.

## 6 Responses to Retraction: George Mason University’s Freedom Index is Anti-Gun [Not Necessarily]

1. duucfho says:

Just a suggestion: I don’t think you should have disabled the link to the original post. I didn’t catch it the first time around, and kind of wanted to read the post and its comments.

I’m guessing you didn’t want the incorrect post to show up in search results or what not. But most sites usually just put an “UPDATED” or something in the title & body explaining the situation. You could even link back to this full redaction.

“that link doesn’t work anymore, BTW”

I always feel that it’s better to leave things working – make a note on the original post that it’s wrong and/or either link to the correction or state the correction before the text of the original post. One, this keeps any inbound links working (unless you’ve redirected that URL to the correction post, which you didn’t). Two, removing things feels sketchy.

• I don’t like deleting things either, but I wanted to try and fix any search engine results we may have skewed by linking to the study.