Kenya 1998 Dataset Content Questions [message #28598] |
Fri, 02 February 2024 18:40 ![Go to next message Go to next message](/theme/default/images/down.png) |
BevB
Messages: 3 Registered: February 2024
|
Member |
|
|
Hi,
I'd like to ask if the Kenya datasets after 1998 do not include the excessive bleeding (DELPROBLE) variable or something similar. I'm wondering if it was renamed or if it is just not included after 1998.
Also, in Kenya 1998, the only delivery method variable option is cesarean delivery. Should the assumption be that if a birth is not cesarean delivery, that it is vaginal?
Is there a stillbirth delineation within IPUMS for Kenya DHS 1998?
Thank you in advance for the support.
|
|
|
|
|
Re: Kenya 1998 Dataset Content Questions [message #28643 is a reply to message #28639] |
Tue, 13 February 2024 11:05 ![Go to previous message Go to previous message](/theme/default/images/up.png) ![Go to next message Go to next message](/theme/default/images/down.png) |
Bridgette-DHS
Messages: 3152 Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS staff member, Tom Pullum:
What the variable refers to is described here: https://www.acog.org/womens-health/faqs/bleeding-during-preg nancy. It is discussed in Chapter 9 of the final report. I see that the Kenya 2022 report does not include the questionnaire in an appendix. However, other DHS-8 final reports do include the questionnaire. You can search for "bleeding" in the Nepal 2022 women's questionnaire (English version), for example. The questions would have been the same in the Kenya 2022 survey.
The PSUs, or sample clusters, do not affect the weights. They do not affect tabulations or estimates of means, rates, proportions, etc. They can be included in an adjustment that affects the standard errors of the estimates. In Stata those adjustments are made with "svyset" and "svy". I don't know how the adjustment is made in R but you could probably get some help from our GitHub site: https://github.com/DHSProgram/DHS-Indicators-R,
|
|
|
Re: Kenya 1998 Dataset Content Questions [message #28828 is a reply to message #28643] |
Fri, 15 March 2024 13:20 ![Go to previous message Go to previous message](/theme/default/images/up.png) ![Go to next message Go to next message](/theme/default/images/down.png) |
BevB
Messages: 3 Registered: February 2024
|
Member |
|
|
Hello,
Thank you for the response. I'd like to ask if you have any recommendations for this error message that appears in R when I try to run a chi square test between the 'excessive bleeding' variable and the 'urban', as I am trying to see if there is a significant association with excessive bleeding by urban-rural status.
The following error message is indicating that there's a problem with the survey design object 'mysurveydesign'. Specifically, it's indicating that one of the strata defined in the survey design has only one Primary Sampling Unit (PSU) at stage 1. However, my DHS dataset has multiple strata and PSUs.
> # chi square results for excessive bleeding (last birth)
> chisquareresult1 <- svychisq(~ excessbleed1 + urban, design = mysurveydesign)
Error in onestrat(`attr<-`(x[index, , drop = FALSE], "recentering", recentering), :
Stratum (2) has only one PSU at stage 1
|
|
|
Re: Kenya 1998 Dataset Content Questions [message #28845 is a reply to message #28828] |
Mon, 18 March 2024 15:38 ![Go to previous message Go to previous message](/theme/default/images/up.png) |
Bridgette-DHS
Messages: 3152 Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS staff member, Tom Pullum:
As I said, I do not use R, but in Stata, the sample design adjustments are specified with this statement (there are variations on it):
"svyset cluster_ID [pweight=v005], strata(stratum_ID) singleunit(centered)"
The problem you have is handled with the singleunit option. You just need to find what is the R equivalent of that Stata option.
|
|
|