Login  Register

Re: Significance test between countries

Posted by Richard Ristow on Apr 20, 2013; 6:41pm
URL: http://spssx-discussion.165.s1.nabble.com/Significance-test-between-countries-tp5719541p5719601.html

At 01:59 PM 4/18/2013, mils wrote:

>I have the following data [below, with statistics added by hand]:
              EU4     France  Germany Italy   Spain
              104     23      35      34      12
              266     54      67      67      78
              207     67      54      32      54

*  Mean:             48.0    52.0    44.3    48.0     .
*  Min:              23      35      32      12       .
*  Max:              67      54      67      78       .

>I was wondering if is it correct to sig test each country "against"
>EU4, since this is the sum of the four countries.

Among other things, what kind of differences among the countries are
you looking for. Paul Swank said, correctly, that if you are looking
for differences in mean value, that's an ANOVA problem. But
differences in mean don't capture the real differences in your data.
For example, France and Spain have the same mean value; but that
misses that Spain has by far the widest range, with both the lowest
and highest values in the data.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD