 |
 |
In addittion, I also thought these responses were informative Cathy smith =================== If you are running linear correlations of each set, you will get 2 beta values. for one slope, a t test will work: t = (r*sqrt(n-2))/(sqrt(1-r^2)) this test [Ho: the slope, beta, = 0], may be used as a 2-tail test. If you wish to compare 2 slopes, I believe you can do it with the same sort of equation and r(2) - r(1), testing if that is 0. However, there are some statistical question underneath, such as the distribution of r, given sample size of n -- is it OK for smaller n? Nope. Is it OK for r much different than 0? Dubious. How far is 'different'? Etc..... Cheers,, and please post the best responses. ================================================= had to find a way of comparing regression slopes when I was a postgrad (not so long ago) and was unable to find a way of doing it in SPSS back then. I'm not sure whether this capacity has been included in the latest edition (I had a brief look just now and couldn't see anything). I did find .pdf on the web with instructions for carrying out this analysis in SAS: www.biology.eku.edu/FREDERIC/bio710/linecomparison.pdf Your other option is to hand crank it, which is what I ended up doing. I found the necessary instructions in Jerrold Zar's Biostatistical Analysis, I forget which edition. The equations scared me slightly at first but are really quite straightforward to put together in Excel - if my relatively-statistically-ignorant, postgrad ecologist self could figure them out then they shouldn't present many problems to anyone else. The analysis is in two parts - first you test to see whether the elevations of the slopes differ. If you find that they do, your analysis stops there and you conclude that the relationship between x and y differs between your two sets. If there is no difference in elevation, you proceed to test for differences in the intercepts of the two lines, which would indicate that although the relationship between x and y is similar in each set, the two sets are significantly different with respect to y. Hope this helps. If you do find a way of doing this in SPSS I'd be grateful if you could let me know - it might come in handy in future... Regards Liz Hensor Dr Elizabeth M A Hensor PhD Data Analyst Academic Unit of Musculoskeletal and Rehabilitation Medicine 36 Clarendon Road Leeds _________________________________________________________________ Get ready for school! Find articles, homework help and more in the Back to School Guide! http://special.msn.com/network/04backtoschool.armx