Dear Statalist members,
I am working on a project where the main analysis focuses on logit/probit with interaction terms between one binary covariate (x1) and other covariates that are continuous (x2, x3, x4).
I would like to add graphs which would enrich a lot the analysis and help me with the interpretation.
I would like to do something like the graphs in Greene (2010, Testing hypotheses about interaction terms in nonlinear models, Figures 1 and 4).
Array
Array
I found this document www.stata-press.com/books/mlmus3_ch10.pdf, which is not exactly like Greene, but which helps me making my point.
Array
The code they suggest is this one:
twoway (line predicet_probability x2 if x1=0, sort)(line predicet_probability x2 if x1=1, sort lpatt(dash)).
But if I do this, what I get is:
Array
Can someone offer me some guidance on how to do it properly?
Best,
Mari
I am working on a project where the main analysis focuses on logit/probit with interaction terms between one binary covariate (x1) and other covariates that are continuous (x2, x3, x4).
I would like to add graphs which would enrich a lot the analysis and help me with the interpretation.
I would like to do something like the graphs in Greene (2010, Testing hypotheses about interaction terms in nonlinear models, Figures 1 and 4).
Array
Array
I found this document www.stata-press.com/books/mlmus3_ch10.pdf, which is not exactly like Greene, but which helps me making my point.
Array
The code they suggest is this one:
twoway (line predicet_probability x2 if x1=0, sort)(line predicet_probability x2 if x1=1, sort lpatt(dash)).
But if I do this, what I get is:
Array
Can someone offer me some guidance on how to do it properly?
Best,
Mari