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 |
Nick Cox <n.j.cox@durham.ac.uk> |

To |
"'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu> |

Subject |
st: RE: rephrase of the question on backfilling missing data |

Date |
Tue, 24 Aug 2010 20:17:21 +0100 |

This is in -long-er form, but not in truly -long- form. The panel idea, or perhaps ideal, is that each distinct date for every panel member defines a distinct observation. Nick n.j.cox@durham.ac.uk David Torres So that I'm not admonished yet again on posing my question with regard to data in wide format, let me put it forward with an example of data in long format. My apologies to you stata experts who desire, perhaps, after reading my muddle-headed questions, to give me a good thrashing. Again, the data are in long format: pubid year interviewed? intdate stfin1_ stfin2_ 1 1997 1 13650 13546 13650 . . 1 1998 1 14200 13546 14200 . . 1 1999 0 . . . . . 1 2000 1 14840 13546 14840 . . 1 2001 1 15199 13546 15199 . . 2 1997 1 13611 13500 13611 13605 13611 2 1998 0 . . . . . 2 1999 0 . . . . . 2 2000 1 14869 14244 14500 14510 14869 2 2001 1 15300 14510 15222 . . 3 1997 1 13676 13482 13676 13599 13676 3 1998 1 14100 13599 14100 . . 3 1999 0 . . . . . 3 2000 0 . . . . . 3 2001 1 15300 14667 15300 . . I just need to copy dates from any round subsequent to a round missed that refers to that year as long as its begin or end date follows the date since last interview. In the example above, respondent 1 (pubid = 1) has data missing for year 1999. It is clear from the data from subsequent years that the job he first reported in 1997 is the same job at which he works. Since he worked right through the year that has missing data, I would copy the concatenated start and finish dates from the first subsequent round and paste it to 1999 (this concatenated variable will later be split, and the end date will be changed to an imputed interview date for the year). Respondent 2 (pubid = 2) missed two consecutive survey rounds, but reports during the 2000 survey that he begun or ended a job since the date of the 1997 interview. Actually the job start and finish dates listed in stfin1_, year 2000, include the latter end of 1998 (14244) and part of year 1999 (thru 14500). stfin2_ of includes years 1999 and 2000. Therefore it, too, can be backfilled to 1999. What is the best way to proceed with backfilling this information so that I can tie wages and compensation, as well as hours and weeks worked, to the year to which it ought to apply? * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

**References**:**st: rephrase of the question on backfilling missing data***From:*David Torres <torresd@umich.edu>

- Prev by Date:
**Re: st: Re: psmatch2 question** - Next by Date:
**Re: st: Op. sys. refuses to provide memory - a cautionary tale** - Previous by thread:
**st: rephrase of the question on backfilling missing data** - Index(es):