[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: Merging using the best match from a set of variables
Lots of people have struggled with a similar problem trying to match
individuals across waves of the US Current Population Survey. You
might want to check out http://www.nber.org/data/cps_match.html and
related references on how well their solutions worked.
On 4/27/06, MA V <firstname.lastname@example.org> wrote:
> Dear statalist users,
> I have a couple of datasets that I am trying to merge. The problem is that
> there is no unique identifier that can easily help me merge the observations
> from the two datasets. Is there a way of merging the datasets such that if
> two observations match in "most" of the variables then it's a match?
> For example, if observation3 (dataset1) matches with observation2 (dataset2)
> in var1, var2 and var3 but not var 4 nor var5, and if observation1
> (dataset1) matches with obs2 only in var3 and var5 then, obs3 (from
> dataset1) - and not observation1- should be merged with obs2(from dataset2).
> I hope my explanation is not very confusing...
> Thanks for your help!
* For searches and help try: