[ Contents | Search | Post | Reply | Next | Previous | Up ]
From: Lucky or not prove it by stats- Che
Date: 10/16/2003
Time: 3:07:08 PM
Remote Name: 81.137.35.37
Statistical approach: 3 methods that could determine lucky or not.
1. Take mean of 2nd place over say 10 years. Then establish Standard Deviation and correct with student t. I think you will discover that the time recorded by the owls was at least 2 sigma the wrong side of the mean.
2. Plot a linear regression (trend line) line for the above and see if the recorded time fit the line (possible) obviously correlation important.
3. Chi-square could be used in some way. I can't think how
I may do this week-end if I'm bored.
Any other methods see Dr Vince