In article <[EMAIL PROTECTED]>, Anders <[EMAIL PROTECTED]> wrote:
>Thanks for your replys. The problem is, that I only have an idea that >the two variables are correlated, I still don't know. From my theory I >just expect correlation between the two variables, that's all. > >In fact a collected all the datas myself (based on all 271 Danish >munipalities) so that should be all right I guess (but I surely may be >wrong). It seems unlikely that you have a theory of these 271 Danish municipalities in particular. Your theory is almost certainly intended to provide insight into the process by which these Danish municipalities ended up the way they did, and perhaps to provide predictions for municipalities elsewhere in the world, or municipalities in Denmark in the future. So you DON'T have the whole population. You therefore need to see whether the correlation in your data set (which is almost certainly non-zero) is statistically significant (ie, unlikely to be non-zero just due to chance). If you're wonder how to do that in SPSS, I can't tell you (not having ever used SPSS), but SPSS will certainly be able to do a linear regression of one variable on another, and will tell you the p-value for testing the hypothesis that the regression coefficient is zero. That's the same as testing the hypothesis that the correlation is zero. Radford Neal ---------------------------------------------------------------------------- Radford M. Neal [EMAIL PROTECTED] Dept. of Statistics and Dept. of Computer Science [EMAIL PROTECTED] University of Toronto http://www.cs.utoronto.ca/~radford ---------------------------------------------------------------------------- . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
