[awptimus]

This article demonstrates the danger of thinking statistics is logic - you can only use statistics with the analytical framework firmly set up...not as a replacement.

The United States is just that - United. States. 38 mass shootings, however defined, means there is /less than one per State/. Guess what? That makes Finland's two /much/ higher. That's...statistically significant!

Explaining how completely incapable frequentist statistics are at describing anything at a low sample size doesn't mean that something succumbs to "random fluctuation." There is nothing random about a person deciding to take arms and shoot lots of people. Not only that, this data is based on a full population sample - it's not a survey of 5,000 people extrapolating to the whole population - it IS the full population.

And he removed the idea that the original chart was designed show - lack of data that shows compelling evidence that more restrictive gun laws prevent either mass shooting incidents or mass shooting deaths.

He should step away from the statistics - it's a loaded gun and he's not trained to use the weapon properly.