treatment effects and marginal structural models

[William Dupont]

The documentation on the teffects command appears to describe a program to solve a marginal structural model (MSM). This documentation does not, however, use this term nor does the paper by Cattaneo (J Econometrics 2010:155:138-54) which is referenced by this documentation. Is it correct to refer to this command as using a MSM? In epidemiology, such models are usually referred to as MSMs rather than treatment effect models. However, I do not want to describe the teffects command incorrectly.

Any clarification on this terminology would be greatly appreciated.

William D. Dupont

[Steve Samuels]

Perhaps this is a test post, not labeled as such. If not, it should be asked in the General Forum.

[William Dupont]

It was a test post. It was my understanding that we were supposed to try out posts in the sandbox before posting them elsewhere. Since the sandbox post appeared to work, I made an identical post in the General Forum. Is this the way that I am expected to interact with the forum?

[Steve Samuels]

There's nothing wrong with what you did. I also often draft my posts here. (Most are under a topic I started call "Test Post". However many new Statalisters look at what appear to be "real posts" and do post here erroneously. I couldn't tell if your post was one of them. I won't make this mistake about you again, but to avoid giving the wrong impression to newcomers, I suggest adding "Test Post" as prefix to your real topic.