| Re: Difference DTA and SAV [message #25915 is a reply to message #25907] |
Tue, 03 January 2023 08:34   |
Bridgette-DHS
Messages: 3199
Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS staff members, Tom Pullum and Trevor Croft:
In CSPro, the package in which the data files are originally constructed, we have two different codes:
1. Blank this is the not applicable code for use when a question is not applicable.
2. 9 (or 99, 999, etc.) this is the missing value code for use when a question is applicable, but a response was not given.
9 (etc.) is used very rarely nowadays with CAPI, but was needed for paper questionnaires, and was used by data entry staff when a particular question was not NA but the interviewer forgot to enter a response.
Usually in SPSS, blank is the system-missing value, and 9 is the user-missing value. In Stata, usually both cases are converted to missing.
We are surprised that the 9 shows up in the Stata file. Perhaps the DTA file for this survey was not created in our usual way, which would convert the CSPro 'missing' to a dot in Stata.
You are reading the Stata and SPSS files with R, and we believe the conversion to R is being handled in different ways. You will get agreement if, when you read the Stata file with R, you add a line to change the 9 to a dot.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
