Skip to main content

UNITSF

HUD.GOV HUDUser.gov
eList
AHS
Hi.

Does anyone have experience with UNITSF (square footage of unit)?

I am using this variable in the 2001 National AHS.  I find ...

1, it is missing in 14.5% of the cases. That's among occupied regular
interviews (status=1).  According to documentation, there should be no
missing values in these cases.

2, the maximal value is 9911.  Documentation states that it ought to be
99998.  Also, there is a cluster of cases at that value of 9911 (251
cases), suggesting that it is a topcoding.  But also, that value is far
from the rest of the values; the next lower is 4300.  I.e., there is a
reasonable-looking distribution from 99 to 4300, then a big gap, then 251
cases at 9911.

3, minimal value is 99, as advertised.

Does anyone else find these same problems?

(I use Stat/Transfer to extract the data and convert to Stata.  It comes
out as an int variable.  That, as contrasted with long, would kill some
high values, but if that were the problem, I would expect a possibly-higher
maximal value, and no cluster at 9911.)

Thanks to anyone who might be able to help.
-- David

David Kantor
Institute for Policy Studies
Johns Hopkins University
dkantor@jhu.edu
410-516-5404