View Single Post
  #14 (permalink)  
Old 08-May-2007, 05:48 PM
Eckelston Eckelston is offline
Established Member
 
Join Date: Dec 2005
Posts: 258
Default

Quote:
Originally Posted by Robert Tulip View Post
Thank you for the correction on my trend analysis, which as you have shown is not significant. It shows my statistics is a bit rusty. The underlying intention was that a bigger sample and longer time series would enable comparison of frequencies of specific aspects by decade, with potential to identify any trends.
Actually there is a way to test that by calculating correlations between the two subsamples. (or even just plotting the frequency in one subsample against frequency in the other) Unfortunately I don't know how you can do significance testing on that, as not all values above 0 are significant.
__________________
"The cure for boredom is curiosity. There is no cure for curiosity."

Dorothy Parker (?)