I have N = 13 and T = 19. Is there another approach other than GMM ? Thanks

[Badiah Eljahimi]

I would be grateful for some guidance on which statistical model should be used for small samples. I have N = 13 and T = 19. Is there another approach other than GMM ? Thanks

[Dimitriy V. Masterov]

What's your research question?

[Carlo Lazzaro]

Badiah:
as Dimitriy highlighted, your question omits relevant details to provide (more) helpful replies.
That said, as you are dealing with a T>N panel dataset, you may want to consider -xtregar-.

[Jeff Wooldridge]

Typically when people refer to "GMM" in this context, it refers to a model with lagged y -- but we need to know this before we can help.

With such small dimensions, there's often little that can be done convincingly beyond two-way fixed effects without lagged y. I have some thoughts with lagged y, if that's what you mean.