Dear stata list, I have a question regarding the Wooldridge test for autocorrelation (command: 'xtserial') (Drukker, 2003)**. After conducting the test, Stata gives an error of "no observations r(2000)". To install the Wooldridge test I have followed the following steps: net sj 3-2 st0039 net install st0039 After that: . xtset panel variable: Country_ID (unbalanced) time variable: Year, 2014 to 2016, but with gaps delta: 1 unit . xtserial DVEgovdevelopment IVLegal IVTechnical IVOrganisational IVCapacitybuidling IVCooperation CVGNI CVCorruption CVInnovation no observations r(2000); I note that other have said that there is a problem with xtserial and unbalanced time periods. Can this be the case, and if yes, is there another way to test for autocorrelation in an unbalanced panel dataset? Help would be appreciated! Regards, Joeri