panel data estimation
Posted: Wed Jun 08, 2011 5:53 am
Hello,
I am running a fixed effects panel data regression with 168 observations. Before running this estimation I have tested for endogeneity; heteroskedasticity and serial correlation. After running the fixed effects regression and in order to correct the problems of endogeneity as well as the co-existence of heteroskedasticity and serial correlation, I use the TSLS and 3SLS estimation method. Although the signs that come up are consistent with the theoretical part of my study and are proved to be statistically significant in accordance with the p-value, the t-statistic values are too high whereas the std. errors values too low (e.g. std. error 0.017704 and t-stat 178.2298). As far as I know, this is an error signal or an I wrong. I would be grateful If anyone could help me or suggest me something.
Thank you
I am running a fixed effects panel data regression with 168 observations. Before running this estimation I have tested for endogeneity; heteroskedasticity and serial correlation. After running the fixed effects regression and in order to correct the problems of endogeneity as well as the co-existence of heteroskedasticity and serial correlation, I use the TSLS and 3SLS estimation method. Although the signs that come up are consistent with the theoretical part of my study and are proved to be statistically significant in accordance with the p-value, the t-statistic values are too high whereas the std. errors values too low (e.g. std. error 0.017704 and t-stat 178.2298). As far as I know, this is an error signal or an I wrong. I would be grateful If anyone could help me or suggest me something.
Thank you