Statistics for Smartasses

Shoot Down Stupid Ideas with the Skill of a Japanese Archer

Are you tired of sitting through endless meetings where some dope tries to bury you (the audience) in data and then convince you that his pile of numbers holds meaning?

I'll tell you, I've seen people try to build the Eiffel tower out of numbers that have the strength of wet paper towels.

Unfortunately, I've never really had the skill or tools to challenge those folks, until now...

I'm just completing the first statistics class of my life, and you know what? It's kind of fun!

My eyes have been opened to the flimsiness of the conclusions that I, and others, have drawn from databases, spreadsheets, and surveys.

I now have three questions that I'm going to ask about every conclusion I see drawn from a big pile of data: 

1. Where's the histogram and correlation coefficient that supports your conclusion?
2. Did you run a t-test, chi-square, or some other test to validate your inferences?
3. Does you test pass the .05 level of significance?

With these three questions, I'm guessing that you will become a corporate martial artist - able to shoot down the worst ideas without removing an arrow from your quiver.

If it turns out they know what you're talking about and have the answers to your questions? Well then, everyone wins, don't they?