1
$\begingroup$

i am currently working with a special version of the SCF that is hosted by uc berkeley and i am experiencing an issue with it.

https://sda.berkeley.edu/sdaweb/analysis/?dataset=scfcomb2019

the special version of this dataset also has this characteristic even though it's hosted externally:

As a result of multiple imputation, the dataset you are downloading contains five times the number of actual observations. Failure to account for the imputations and the complex sample design will result in incorrect estimation of standard errors.

https://www.federalreserve.gov/econres/scfindex.htm

this is the problem: my dataset doesn't appear to contain the replicate weights. i am only given the sample weights. the berkeley website does not seem to contain the replicate weights either.

i am working in r with srvyr and i can't use as_survey_rep() because i have no replicate weight variable that i know of. i can use as_survey(), but i don't think this corrects the sample size issue.

anybody have any advice? thank you in advance.

$\endgroup$
1
$\begingroup$

according to the people who maintain this dataset at berkeley, it can't be used for inferential work and the replicate weights for this data that is stored on the fed website is not compatible with this dataset.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.