Difference-in-Difference-in-Differences Method (DDD) | Estimation Methods | Stata Tutorials Topic 43

[@BobWenEcon]

10:01 there should be a level term for the “state” variable in the model.

[@bianfang7978]

thank you for your sharing very informative

[@waititstiiina]

Is it possible to conduct did for two treatments? Like one treatment occurred and then another one occurred?

[@MichaelAdu-Gyamfi-m2k]

Hello Dr. Bob, could you please explain the situation where the Parallel trends assumption can be identified to have been violated in the model?

[@BobWenEcon]

Suppose the policy took effect in time t. If you can find different trends during t-2 and t-1 for the two groups, it is evidence that the parallel trends assumption is not valid.

[@MichaelAdu-Gyamfi-m2k]

That’s fine I get that. Thank you so much ❤