topic Re: PROC SURVEYMEANS and calculating SD in Statistical Procedures

PROC SURVEYMEANS and calculating SD

Emma_at_SAS — Fri, 16 Apr 2021 21:44:11 GMT

Here:

https://communities.sas.com/t5/Ask-the-Expert/What-Are-Best-Practices-for-Using-SAS-Survey-Procedures-Q-amp-A/ta-p/706830

I found this:

Why does PROC SURVEYMEANS not report Standard deviations?

PROC SURVEYMEANS computes the NOBS, MEAN, STDERR, and CLM statistics by default, but you can request standard deviation by adding STD as an option. Please check the documentation for additional keywords.

However, PROC SURVEYMEANS only calculates STD when I add SUM and the output for STD is Std Error
of Sum. Do you have other suggestions? Thanks

Re: PROC SURVEYMEANS and calculating SD

Reeza — Fri, 16 Apr 2021 22:13:15 GMT

Show your code.

FYI - looking at the reference you've linked to it does not match the documentation. SAS is doing what is stated in the documentation. I don't see an option to calculate a standard deviation. But I suspect there's a conceptually good reason for this with weighted survey data but my brain is kinda done for the weekend now.

I would first look into if this is statistically valid.

Re: PROC SURVEYMEANS and calculating SD

gcjfernandez — Sat, 17 Apr 2021 06:50:45 GMT

This topic is covered in the following esource: https://support.sas.com/kb/45/701.html

Please review.

Thanks

Re: PROC SURVEYMEANS and calculating SD

Emma_at_SAS — Mon, 19 Apr 2021 17:57:44 GMT

Thank you, Reeza. Yes, the good reason is explained in the document that gcjfernandez shared. It seems a finite population adjustment task. Thanks

Re: PROC SURVEYMEANS and calculating SD

Emma_at_SAS — Mon, 19 Apr 2021 18:07:17 GMT

Thank you, gcjfernandez! That is the solution.