Dear IPUMS Team,
I was working on Mexico’s 2010 supplemental data and found that the ‘serial’ and ‘seqm’ does not uniquely identify themselves. Is there any reason for that? I tried finding some difference but could not.
If you look at the data with serial numbers 2902365000, 2894774000, 2894716000 and 2894295000, you will find the characteristics of the migrants are the same.
Thank you
The IPUMS International team is currently looking into this. I’ll post here again once we have an answer to your question.
The IPUMS International team has verified that this is an issue with the raw data they received. There appear to be two issues:
Duplicate records (two records that are identical in every way).
Duplicate seqm number for apparently different people. They suspect that this is mainly caused by one issue: there were only four places to list migrants on the questionnaire, and enumerators were instructed to use a second form if there were more than four migrants. They believe the seqm was then repeated for the additional attached questionnaires.
To fix these problems, the IPUMS International team plans to re-number duplicate values of seqm within a household. These will be posted on the website during the next data release. They do not plan to drop records, so there may still be a need to manually drop individuals with apparent duplicate records (issue #1).
Thank you so much!
Do you have any approx time when the next data will be released?
The IPUMS International team is aiming for the end of July for the next release.
Thank You!
Has this issue been already taken account of in the current supplemental data file?
No, this hasn’t yet been updated. Please check back again in mid-October. You can see all the data updates on the revisions page. If you need to use these data sooner, I suggest you implement the fix I described in my earlier post.
Okay! will look into it.
Thanks, Matt!