Hello,
I am trying to figure out the best way to see if different categories within a variable are statistically different to on another.
For example. I have my binary outcome variable Y (0=no, 1=yes) and a predictor categorical variable X (0=small, 1=medium, 2=large)
I used this command:
logit Y i.X
margins X, pwcompare(effects)
which gives me a table set up:
Pairwise comparisons of adjusted predictions
small vs medium p-value 0.005
small vs large p-value 0.234
medium vs large p-value 0.001
Is this the correct interpretation with p < 0.05?
small is significantly different to medium, and medium is significantly different to large across Y using the Pairwise comparisons of adjusted predictions?
or is there a different way I should be going about this?
Thank you !
I am trying to figure out the best way to see if different categories within a variable are statistically different to on another.
For example. I have my binary outcome variable Y (0=no, 1=yes) and a predictor categorical variable X (0=small, 1=medium, 2=large)
I used this command:
logit Y i.X
margins X, pwcompare(effects)
which gives me a table set up:
Pairwise comparisons of adjusted predictions
small vs medium p-value 0.005
small vs large p-value 0.234
medium vs large p-value 0.001
Is this the correct interpretation with p < 0.05?
small is significantly different to medium, and medium is significantly different to large across Y using the Pairwise comparisons of adjusted predictions?
or is there a different way I should be going about this?
Thank you !
