Split variable


I don’t really understand how the variables SPLIT, SPLITHID and SPLITNU help to study resisdents of large prisons. How do we know whether those institutions are actually prisons?

The description of the variable SPLIT is as follows:

Researchers needing to analyze intact group quarters units containing more than 60 people (such as the residents of a large prison or poorhouse) should use SPLITHID and SPLITNUM, instead of SERIAL and PERNUM.

Thanks a lot


The SPLIT variables alone are not able to identify prisons. The GQ/GQTYPE variables identify institutions and if your analysis requires you to reassemble these large institutions you can use SPLITHID to group the individuals as they were originally enumerated. However, if you do not need the individuals to be grouped under a single household record you can simply use GQ/GQTYPE to identify individuals that were institutional inmates.

I hope this helps.