Hi everyone, I am conducting the Wooldridge test for serial autocorrelation for my mediation analysis. My data is unbalanced panel data.

I wanted to know if I should include the mediator and control variables to the xtserial command or not? Basically I have added lagged Dependent Variable (DV) as a control which is why I wanted to check for autocorrelation. But I am not sure if I need to include all control variables as well as the mediator variable in the test. Right now I am using the following equation:

1. xtserial DV IV MV CV1 CV2 CV3 CV4

Where DV= Dependent Variable, IV= Independent Variable, MV= Mediator Variable, CV1-CV4= Control Variables

Should I use this set up or should I just keep the following set up with just one control variable which is the lagged DV:
2. xtserial DV IV MV CV1(Lagged DV)

Thank you in advance!