I just produced a 3 x 2 crosstab in SPSS 20.0.0 where my rows are risk grades I, II & III and the 2 columns: disease yes or no. I figured that I could use the Cramer's V and tahn Lambda in order to see if there is a trend towards more disease in the group that had a higher risk.
Is that correct?
Or would there be another consideration? I guess the Mantel-Haenszel linear by linear is not an option since the user's manual says "When both table variables are quantitative"
But this discussion confuses me again (https://groups.google.com/forum/#!topic ... hoD_kQ0H3c
Also how do I state the outcome correctly, in the end I'd like to come up with something like the higher risk, the more often disease.
Thank you very much, I appreciate your help.
Ok after a day of struggeling through books on statistics, I though that for a ordinal x nominal table like this
(3 x 2 crosstab in SPSS 20.0.0 where my rows are risk grades I, II & III and the 2 columns: disease yes or no)
I could use the Cramer V.
I also thought I could use the lambda but it seems like its always having some issues with a 3x3 table and needs at least a 3x3 table. I mean It'll alsways show me some
Value Asymp. Std. Errora Approx. T Approx. Sig.
0,000 0,000 .b .b
b. Cannot be computed because the asymptotic standard error equals zero.