Hi
I was going to say that you can use the program (.exe) here to compare
correlations (both related/unrelated) as well as other details:
(see week 18)
http://homepages.gold.ac.uk/aphome/misc.html
also there is a useful spreadsheet available from stat-help.com
http://www.stat-help.com/spreadsheets.html
The only thing with comparing correlations is that I am still unsure of
whether if makes sense to compare correlations of different sign. As I am
lead to believe correlation comparisons are concerned with the strength of
any association (i.e., not the direction) - so should two correlations of
say -.4 and +.4 be considered as different?
The Fisher transformation may account for the sign issue, but I have heard
arguments on both sides as to whether this makes sense or not and am
undecided.
sorry, I digressed there...
Anyhow I hope the above are useful.
cheers
Ian
|