Designated persons in ATUS and CPS

Hi, I’ve extracted data on 201,051 ATUS respondents from all available years. Of these, more than 72,000 are identified as observations for which the ATUS designated person is NOT the same as the CPS designated person. However, there is valid data for all observations regardless of whether designated persons match, as shown below. So would I exclude the “atusdp=2” respondents only when I’m using matched CPS data? Or are there other circumstances in which I’d exclude them, even though there’s time use data for those observations? thanks much, Sanjiv

. tab atusdp, su(bls_hhact_hwork)

person same
as CPS Summary of BLS: Household
designated activities: Housework
person Mean Std. Dev. Freq.

Yes, same 45.134186 84.833199 128,173
No, diffe 33.72541 75.505407 72,978

Total 40.995058 81.75648 201,151

There is no restriction that requires the CPS designated person (in household’s 8th and final CPS interview) to be the same person as the ATUS designated person. Some information about the household, collected in the final CPS interview, is attached to the ATUS data and some information is asked again in the ATUS interview. Therefore, whether or not you exclude the individuals with ATUSDP==2 is ultimately an analytic decision that we at IPUMS User Support are not well-equipped to answer. Personally, unless your analytical method suggests excluding these individuals for a specific reason, I’d suggest running your analysis both with and without those with ATUSDP==2 and see if the results change significantly.

1 Like