Monday, October 31, 2011

A note on confidence

Let me take a moment to talk with all you fine people about confidence.

Clearly, I has it. But that's not what this post is about. This post is about confidence in the statistical sense.

Now when I'm asked a question, I'm always in a situation where I don't have all the information. By which I mean: I have like this tiny fucking cross-section of information and I'm doing my best to guess what the rest of it is. This means that, by and large, I'm taking a few liberties and guessing at the blank spots. What the hell else can I do?

So here's what I have decided to do about that guessing bit, which makes me pretty damn uncomfortable given that my goal is to reflect reality and all. I'm going to institue a scale of 1-10 that reflects the confidence i have in my answer where 1 is a W.A.G. (Wild Ass Guess) and 10 is not only am I certain of this answer in a factual way I have tested and verified my result in a scientifically valid double blind experiment and have found the result to be utterly conclusive.

Now I'm thinking it's pretty unlikely we'll see a 10, given that I'm not a research scientist. And, let's be real here, I've got no choice but to fill in some blanks given that I don't know any of you dudes first hand. As a result I'm going to cap any results at a confidence level of around 5 for the purposes of this blog.

So in the future expect to see a number attached to each answer that will reflect my overall confidence in the data provided and my analysis of that data. If it goes over 2, I'd pay some fucking attention.






No comments:

Post a Comment