Hello, I would like to find / or calculate estimates at the state level for a few metrics within the CPS Civic Engagement Supplement. I am wondering if producing state level estimates with these data would be valid and reliable estimates? Can you please direct me to some sample codes in R of others who have used the Civic Engagement Supplement weights to generate estimates at the state level? Any guidance or suggestions are greatly appreciated!
Chance
Hi Chance,
Yes, you should be able to produce state-level estimates using data from the Civic Engagement Supplement of the CPS. Estimates for smaller geographic units or specific subpopulations for which there are few respondents may have high standard errors. However, I doubt you’ll encounter this issue for state-level estimates. IPUMS doesn’t provide any R code specific to this supplement, but I recommend using the survey and srvyr packages. In fact for this case, the weighted.mean() function should suffice for generating your estimates. You may also find the ipumsr package useful for other types of analyses with IPUMS data in R.
Thanks so much for your helpful response!