![]() ![]() ![]() See help matrix for additional methods of working with this matrix.Īlso, there are numerous posts on CrossValidated that may be more helpful in understanding pooled OLS. You can capture the variance-covariance matrix using the following immediately after reg matrix myVarCovar =e(V) As such, your command would look something like this: reg Y X1 X2 i.timeVar, robust It is typical to also include a time fixed effect to control for changes that affect the sample over time. This being said, the regress command is the correct tool for the job. The reverse is not true: If the sample is pooled, the researcher cannot use the standard fixed effects methods. If the researcher has strong reasons to believe that such heterogeneity does not exist, she may choose to use the pooled OLS on a longitudinal sample. In this case, the researcher typically uses a within (fixed effects) or similar estimator to sweep out unobserved heterogeneity in the individuals. ![]() (2) this relationship is additive (i.e., Y x1 + x2 + + xN) Technically, linear regression estimates how much Y changes when X changes one unit. This differentiates it from a panel (or longitudinal) sample where the same observational units are repeatedly observed. When running a regression, we are making two assumptions, (1) there is a linear relationship between two variables (i.e., X and Y) and. Under this sampling scheme, the observations form different time periods are pooled together and OLS is conducted on the pooled sample. My understanding of pooled OLS is that it is most appropriate when you have observational units observed in more than one time period, but individual units are not repeatedly observed across periods. ![]()
0 Comments
Leave a Reply. |