Re: Chisquare and gamma coefficient

Posted by lucameyer on Jan 26, 2011; 8:12am
URL: http://spssx-discussion.165.s1.nabble.com/SPSS-16-0-0-error-Two-step-cluster-Options-tp3353123p3357519.html

If you sample size is 77 and you have 8 cells in your table you might get an unstable result. SPSS CTABLES should print the following warning:  "More than 20% of cells in this subtable have expected cell counts less than 5. Chi-square results may be invalid." similar warning should be printed by FREQUENCIES

As an alternative you should into NPAR TESTS.

HTH,
Luca

Luca Meyer
www.lucameyer.com