Question

Status:Closed    Asked:Jan 11, 2018 - 04:17 PM

Is HHINTYPE variable corrupted?

I downloaded a series of variables, but want to subset the dataset by only those who were interviewed or were Type A non-interviews (see codebook for description). If I go into "case count view" under the HHINTYPE variable, it gives me the cell breakdowns for this variable (e.g. for Oct 2017 there were 52,638 interviews and 8,402 Type A non-interviews). However, when I download the data with this variable (in .csv format), there is no variability, with everyone being listed as being interviewed. Any help would be appreciated.

 
Do you have the same question? Follow this Question
 

Staff Answer

avatar

Jeff Bloem

Staff

Actually, in IPUMS CPS samples downloaded in rectangular format (e.g. one record for every person) all non-interviewed households are excluded from the sample. Basically, this is because all non-interview records have no information in which to form a person-level record. If you want to have a count of non-interview households, you can always download IPUMS CPS samples in hierarchical format (e.g. with person records nested under each household record). This can be done by changing the data structure on the "Extract Request" page. Finally, note that hierarchical formatted data extracts are only able to be downloaded via a fixed-width .dat file. So, you'll need to use a statistical software (e.g. STATA, SPSS, or SAS) to see the data.

 

Jan 12, 2018 - 09:49 AM

0
0
Report it

OTHER QUESTIONS NEEDING ANSWERS

CPS - How to know whether a Type-A noninterview (HHINTYPE) is a refusal or a ...
Login   |   Register

Recently Active Members

View More »

Share |