Interpreting Interaction Effect in Proc Genmod (Categorical vs. Categorical with multiple levels

[PaigeMiller]

---

@varatt90 wrote:

When I specify "effectplot interaction (x=depression)" where Y = smoking status into my code, this is what I see. Is this what you meant?

 

---

Yes, these are the plots. I find looking at them helpful in some ways, girl seems to have opposite sign on the slope to boys and opposite sign on the slope to other. That's a useful piece of information that isn't immediately obvious looking at the tables of output. Another point is that the slope of other is flatter than the slope of boys and flatter than the slope of girls; again this is a piece of information that isn't immediately obvious looking at the tables of output. Finally, girls shows a bigger range from "No" to "Yes" than boys or other on both plots.

[varatt90]

Thank you for interpreting that graphs! It was nice to see what to look out for when reading the graphs.

[PaigeMiller]

---

@varatt90 wrote:
Thank you for interpreting that graphs! It was nice to see what to look out for when reading the graphs.

---

Some famous person once created the hashtag #APictureIsWorthAThousandNumbers

[varatt90]

Haha definitely a wise person...who also probably loved interaction plots.