| Re: Convert DHS (SPSS?) missing value codes to Stata codes in Stata dataset [message #12187 is a reply to message #12179] |
Thu, 06 April 2017 13:15   |
Bridgette-DHS
Messages: 3058
Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS Stata Specialist, Tom Pullum:
You are deleting the cases with missing values? This is not necessary and not desirable.
Your message refers to SPSS, Stata, and SAS. They have somewhat different ways of handling missing cases. In Stata, the only package I use, a blank or dot means not applicable. Those values are not exactly "missing". There's no value because the relevant question was not applicable. Stata will automatically exclude those cases from calculations. There will often be other codes such as 9, 99, etc., meaning that there should have been a response but there wasn't. You will want to exclude from calculations. You can do that by assigning those codes to a dot or doing some similar recode; just be sure that you never permanently over-write the original code. Excluding a case by dropping it from the file is not a good idea because it is quite possible that you will want those cases later for some other reason, and once they have been dropped you can only get them back by starting over.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
