Chi_square and post hocs tests

my post is somehow inspired by that post:

But what would be an interpretation of the results ? Which group is compared with which ?

Additionally I have tried to recreate these post-hoc test for chi square from here (at the bottom of that page there is an edu-marit.sav file for download):

So I have done a following crosstable and then a relevant matrix in R:


chisq_matrix <- as.table(rbind(c(18,	36,	21,	9, 6),
                                   c(12, 36, 45, 36,	21),
                                   c(6,	9,	9,	3,	3),
                                   c(3,	9,	9,	6,	3)))

dimnames(chisq_matrix) <- list(Marital_status = c("Never_married", "Married", "Divorced", "Widowed"),
                    Education_level = c("Middleschoolorlower",	"Highschool",	"Bachelors",	"Masters",	"PhDorhigher"))

getting this result:



chisq.theo.multcomp(chisq_matrix, p.method = "bonferroni")

getting that:

I would like to ask for help with interpretation which group was compared with which and to compare results with Z-test performed in SPSS that saying:

in Never married level Middle school or lower group (significantly) differs from Master's group, which is NOT clearly visible looking at results I have obtained from R. Any help would be much appreciated, thank you.

Is it possible that chi square is statistically significant but there is no differences detected between levels ?

This topic was automatically closed 42 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.