I am now working with Stata and I found out I have a unit root in my regression. How can I correct for this, because I can read everywhere what the test does but if you have a unit root, what to do next?? Sorry, it is one of my first times working with Stata so it is all really confusing..

  • 1
    $\begingroup$ I can't see that this is a Stata question; conversely, if it is, it is off-topic here. If the doubt is what to do, then much of any time series text or course includes material on (e.g.) removing trends and/or building models that include trends. As formulated, this question is too broad to be answered, as it is too close to what should I do with nonstationary data? $\endgroup$ – Nick Cox Feb 26 '14 at 11:37
  • $\begingroup$ Sorry I am not really familiar with this. At the moment I am working on a thesis for my bachelor and I am working in stata. I was testing if my data had a unit root and this is now the case. I want to keep working with this data set but now I know it is Unit root I have to change something. I think it is something like this: y_t-y_(t-1) instead of y_t but what is the stata comment for that or do I have to change it in my excel format? $\endgroup$ – hannah Feb 26 '14 at 11:40
  • $\begingroup$ Sorry, but this clarification reveals that you want detailed advice on software use. (I take "comment" to be a typo for "command".) Please see stats.stackexchange.com/help/on-topic to learn that this kind of question is off-topic here. You may be better off on Statalist, but in all forums you are expected to do some basic learning for yourself first. $\endgroup$ – Nick Cox Feb 26 '14 at 11:44
  • $\begingroup$ Although this question is pretty broad, it does not appear to bet (just) about Stata. I think it might come down to "when a time series appears to have a unit root, what are the next steps I should consider taking in my analysis?" If I'm correct in this interpretation, Hannah, then please feel free to edit your question so that it clearly reflects that intent. To help focus it, consider including a line or two describing the purpose of your analysis. For instance, are you focusing on accurate forecasting or do you intend to interpret the coefficients of your model in some way? $\endgroup$ – whuber Feb 26 '14 at 14:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.