SPA data set related question [message #13817] |
Wed, 03 January 2018 02:01 |
Seba
Messages: 13 Registered: October 2017
|
Member |
|
|
Hello, I am writing to ask if there is a way I can identify facility location (that is Urban Vs. Rural) from SPA data set.
Thank you,
Sebawit
|
|
|
|
|
Re: SPA dataset: antanatal care dataset [message #17935 is a reply to message #17869] |
Fri, 19 July 2019 11:28 |
Bridgette-DHS
Messages: 3171 Registered: February 2013
|
Senior Member |
|
|
Following is a response from DHS Senior Analysis and Research Manager, Wenjuan Wang:
The DRC SPA data are raw data, and the recode data are not yet available. That is why you saw the letter code as shown in the table below or in your tabulation.
Variable oa115 comes from Q115 in the ANC observation questionnaire, see below: In the raw data file, the letters remain as categories without recoding. For example, in the tabulation below, the first row means A (asked the client where she will deliver) was done for 147 observed clients; the second row "AB" means Both A and B were done for 124 clients, the third row, "ABC" means A, B, and C were done for 86 clients, so on and so forth... But when you analyze data, you probably want to know how many or what percentage of clients got A regardless instead of different combinations listed in variable oa115. In that case, you need to recode variable oa115. If you use Stata, you can use this command "regexm" to generate a new variable that indicates if A was done at all, regardless it was the only thing done or with other services.
Hope this helps and let us know if you have other questions.
gen a=0
replace a=1 if regexm(of115,"A")
|
|
|
|