Hot to make Autocorrelation Test in STATA for Panel Data

[Sikinti Yok]

Dear all,

I have been strugling with autocorrealtion test for my STATA project.

My study contains 20 countries and 21 years.

I have tried variaties of commands I have found on internet such as xtserial or sort id and t however I am not able to complete the test correctly.

Could you please advice how can I implement and autocorrelation test to my project?

Thanks in advance,
Sikinti Yok

[George Ford]

You'll need to provide sufficient data (search dataex) to reproduce the result (or lack thereof).

[Carlo Lazzaro]

Sikinti:
as George wisely noticed, you provide interested listers with too scant details.
You do not share any Stata code and the specification of your panel data regression (-fe- or -re-) is also obscure.
All that said, shamelessly elaborating on a worked out example reported in

https://www.stata.com/bookstore/microeconometrics-stata, page 252,

what follows may be of some help:

use "https://www.stata-press.com/data/r17/nlswork.dta"
quietly xtreg ln_wage c.age##c.age
predict epsilon, e
. forvalues i=1/5 {
  2. forvalues j=1(1)5 {
  3. quietly corr epsilon L`j'.epsilon if idcode==`i'
  4. gen delta`j'=r(rho)
  5. sum delta`j', meanonly
  6. bysort idcode: replace epsilon_corr=r(mean) if idcode==`i' & _n==`j'
  7. drop delta`j'
  8.  }
  9. sum epsilon_corr if idcode==`i'
 10.  }


    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
epsilon_corr |          5    .5729781    .1993669   .3150713   .8134871

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
epsilon_corr |          5    -.388444    .5590177  -.9798838    .275066


    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
epsilon_corr |          5    .5206494    .1727996   .3151894   .7879882


    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
epsilon_corr |          5    .7507667    .4060343   .0323193          1


    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
epsilon_corr |          5   -.0774237    .4821721  -.5287769   .6038273

.

The average correlation of -epsilon- residuals seems to deserve attention for -idcode- #4,1 and 3.