Notice: On March 31, it was **announced** that Statalist is moving from an email list to a **forum**. The old list will shut down on April 23, and its replacement, **statalist.org** is already up and running.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

From |
Ken Ning <kenscholar2008@gmail.com> |

To |
statalist@hsphsun2.harvard.edu |

Subject |
Re: st: Panel data by subgroups |

Date |
Fri, 14 Jun 2013 14:59:49 +0100 |

Hi Muhammad, thank you very much for taking the time to reply to me. The issue is that a company in the last five years might have more than one case for left_personal_reasons. The turnover itself is rather a cross sectional event. If I simply run the regression using dummy variables for the whole period, those periods that have other reasons for financial managers turnover events are also used and affect the estimation. A friend said i should take each group and pool the three year before the turnover and three year after the turnover together. e.g. left_personal_reasons (occured in 2002) so take 1999, 2000, 2001 ,(2002) , 2003 , 2004 and 2005. the next left_personal_reasons (occured in 2005) so take 2002 , 2003 , 2004 , (2005), 2006, 2007, 2008. I think this becomes pooled data? or is there a way to solve this? many thanks, Ken Hello Either you can create a categorical variable, says its name is category, which has a value of 1 for promotion, 2 for fired and 3 for left_personal_reasons, and can run the category specific regressions like Before running the regressions you can define another set of three variables like category1 (having value of 1 if promoted otherwise 0), category2 (having value of 1 if fired otherwise 0) and category3 (having value of 1 if left_personal_reasons otherwise 0) by using a -tab- command on the variable containing these categories, assuming. tab assuming_variable_cateogries, gen(category) this will create the above defined three variables. Now you can run the regression for the first case as xtreg dep_var indep_vars if category==1, (re or fe as required) xtreg dep_var indep_vars if category==2, (re or fe as required) xtreg dep_var indep_vars if category==3, (re or fe as required) or for the second case Now you can run the regression for the first case as xtreg dep_var indep_vars if category1==1, (re or fe as required) xtreg dep_var indep_vars if category2==1, (re or fe as required) xtreg dep_var indep_vars if category3==1, (re or fe as required) This is what I assumed you needed to do. There are other options too to think of I hope this helps to begin working. Best of Personal Regards --------------------------- Muhammad Anees Assistant Professor/Programme Coordinator COMSATS Institute of Information Technology Attock 43600, Pakistan Learn Econometrics & Stata: elearning.aneconomist.com http://www.aneconomist.com * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/

**References**:**st: Panel data by subgroups***From:*Ken Ning <kenscholar2008@gmail.com>

- Prev by Date:
**Re: st: Saving bootstrap samples** - Next by Date:
**Re: st: Interpretation of interaction term in log linear (non linear) model** - Previous by thread:
**st: Panel data by subgroups** - Next by thread:
**st: graph combine with (non-standard) one legend** - Index(es):