Get 32.15% off this month when you join the Statistics Police today!

Dichotomania Defined


Dichotomania is a term coined in the 1970s by those studying brain asymmetry, and used by modern statisticians to reform science and save the world and stigmatize frequentism, for the human habit of reducing continuous measurements to categories, and then using those categories to commit atrocities like make scientific decisions. 

Even though science classifies, and we make discrete categories (shoe size) out of continuous measurements (foot length), and Fisher said to repeat experiments, this should be no grounds for putting up with extreme pseudoscience, like using distances between hypothesized models and test statistics that came from data from well-designed experiments to make decisions like hypothesis testing that has been used in sciences and other areas for 70+ years successfully all over the world, and writing (or even thinking of writing) things like "p-value < alpha" and "statistical significance".

What is the solution? Well, we can't provide alternatives that don't have their own set of problems, but, we do know we definitely need to ban things, obviously, starting with select words and random variables. And we need to form groups and write. Lots of writing.

According to Lindley, "We will all be Bayesians in 2020, and then we can be a united profession". Bayesian statistics has been around for several hundred years already, so you know he is correct, and nothing unites people like telling them what they should be. Don't be left behind.

an exciting  dictionary or collection of  signatures

an exciting dictionary or collection of signatures