The easiest way to lie with statistics is simply to lie

Ezra Klein (of Journolist fame) is running a list of “facts” about guns in America. Most of them are nonsense, but for most of those showing so requires some analysis that I will leave to other commentators. I want to focus on just one, this chart:

assault-deaths-some-oecd

The chart shows the US with a significantly higher rate of assault deaths than all “other OECD”. Klein adds: “We are a clear outlier.”

Now, you can’t tell with them drawn atop each other, but there are 24 graphs on that chart. (They are shown separately here.) But there are 34 OECD members. If you include those (which the graph’s author did later, after being questioned on the point) you get a very different picture:

assault-deaths-all-oecd

Not so much of an outlier with eight more countries put in — particularly if you look at 2010 and not the late 1970s — are we? The original graph left out all the countries with a higher or similar rate.

But wait, eight more countries? Weren’t we missing ten? If you peruse the list, it is still missing Iceland and Turkey. I wonder what’s going on in those two. The author (who does not study violence, by the way) says he couldn’t get data for Turkey. He doesn’t mention Iceland at all.

Finally, I should mention that other measures of violence paint a very different picture (see number 6, here), but that gets into the realm of analysis, so I’ll stop here.

Advertisements

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s