we can calculate p for trend for Qualitative data while we calculate chi square by SPSS! but what about quantitative data in anova? do i have to use STATA for it? 

Similar questions and discussions