I’m having trouble distinguishing between households who are “not in universe” for the FSSTATUS variable in the Food Security Supplement, versus those who have “no response.” I’ve been trying to figure out the difference between these two groups by digging through their responses to FSSCREEN and FSSUPINT, as well as visually combing through the data, to no avail. I will be analyzing household food insecurity status and want to be able to distinguish between these two groups when I justify why I’m removing them from my dataset. Many thanks!
The FSSTATUS variable has unique codes for “no response” and “NIU”, respectively. FSSTATUS==98 is no response and FSSTATUS==99 is NIU. Additional information about how the universe of the variable is defined can always be found on the Universe Tab.
Thank you for your quick response. My question was actually more of a definitional question – there’s no clear distinction within the dataset, or in any technical documentation that I’ve read, between what constitutes a household that is NIU versus a non-respondent for this particular variable. That’s why I was checking the screening variable (FSSCREEN) and such. I’ll keep digging.
Ah, okay. Yes, this is a bit confusing with the food security supplement. The difference between NIU and non-response is subtle because the universe for the supplement is technically all households. That being said, there are some households that for whatever reason are non-respondents for the entire food security supplement. These households are identified in FSSUPINT. All observations from households with FSSUPINT==“FSS non-respondent” are coded as NIU in the other food security supplement questions. This is because the CPS Food Security Supplement codebooks code these “supplement non-respondents” NIU. On the other hand, “non-respondents”, as coded in variables such as FSSTATUS and FSSCREEN, identify households that are respondents to the food security supplement in general (FSSUPINT==“FSS respondent”) but did not respond to all of the necessary questions for a given variable. I hope this helps.