I've been reading up on endogeneity and I see many articles where authors use 2SLS or instrumental variables without checking whether or not endogeneity exists.
My data set is a balanced panel data set and I'm using a change HLM model: DeltaY_ij=〖β0〗_j+〖β1〗_j X_ij+Ɛ_ij.
Can't I just save the residuals from my model and run a correlation test each of my predictors? If they are correlated then endogeneity exists and then I would use an instrument, if not then endogeneity is not an issue.
i had asked the same question from few researchers around me, but all commented that u cant check endogeneity before estimation. post estimation test would confirm about presence of endogeneity,,,,,,you can only decide on the basis of theory... or theoritical relation of the variables. .........
while .....still m not satisfied on this reply......or might be ....we are now in practice of pre-testing of models, thats whay ....want pre-testing for endogeneity.......
Mehru Nisa thank you for your input. My question is about post estimation. So I run my model and I save the residuals and then correlate the residuals with the main IVs in my model. Will this correlation confirm the presence of endogeneity?
I have done Sargen test , after estimation through GMM.....as post estimation,,,,let me check then will discuss with u.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA