?

Log in

No account? Create an account

March 23rd, 2005

HOTT

Kate and I were watching "Mythbusters" the other night, and they drew some conclusion that some stimulus-effect was real because 29% of their experimental group demonstrated the effect while only 25% of the control group did. This didn't seem like much of a difference given their n which was a modest 50. I mentioned this to Kate as we watched, but she already had gotten out a piece of paper and was doing a chi-squared test.

Sure enough, the chi-squared value was way, way, way below the value necessary for 95% confidence level. So there was no difference, statistically, between their experimental and control groups and thus we concluded their "myth confirmed" conclusion was incorrect.

Kate decided to post this on the show's bulletin board, and we figured we'd get only geek responses. But just now I looked and she got a response from Kari Byron, one of the show's interns. The hot one, too!

I love my wife. She's smart and persuasive and stuff. And a tip of the hat to Kari for actually responding. She weaseled a bit, but basically said that the stat they showed was not the one that caused them to conclude that the myth was confirmed. (So, what, bad editing?) And that stats is boring so a lot of their reasoning in confirming/busting a myth winds up on the editing room floor. Unfortunately, I'm sure she's right about this.

The BB entry is here.

Profile

14L
szasz
Charley

Latest Month

July 2013
S M T W T F S
 123456
78910111213
14151617181920
21222324252627
28293031