Showing posts with label Statistically Significant. Show all posts
Showing posts with label Statistically Significant. Show all posts

Wednesday, February 4, 2009

Lies, Damn Lies, and Statistics

Here are my overall results and a nice picture of Uncle Sam for you.


I love Excel. Well except for the fact that there doesn't seem to be any good way of displaying an excel table in a blog. Unfortunately this means i'm going to have to PrintScreen and copy-paste everything. I didn't realize but I don't think anybody who has come actually was able to see the results i'm posting....I'll fix the old posts when I get a chance. Ok, the results.

Basically I accurately predicted the scores of 9 of 29 movies. Statistically I had a 10.84% chance of doing this by pure luck (specifically that is the chance of predicting 9 or more correctly, not just 9 which would be something like 5%); that means the result does not yet fall into that much coveted category of "statistically significant". Did you know that every time you say "statistically significant" a social scientist somewhere orgasms? It doesn't matter if the correlation is meaningless, it's statistically significant! I digress....

So Uncle Sam is giving you attitude up there because we want to be at least 95% sure that all this is not just a big, 5-sided die flipping fluke. We'll see if the score improves as the sample size gets larger.