| Abortion question - Cote d'Ivoire 2012 [message #13939] |
Fri, 26 January 2018 15:37  |
cgreenba
Messages: 18
Registered: October 2017
|
Member |
|
|
Hello all,
I am interested in using the question on terminated pregnancies by abortion to get a rough estimate for rates of induced abortion in Cote d'Ivoire. I am wondering if anyone has an information on the validity and reliability of this data in the DHS survey. Other studies from 2002 and 2007 have estimated the prevalence of induced abortion at between 39 and 42.5%. In the DHS data, with question v228, 18.87% reported having ever had a terminated pregnancy, and only 8.12% report having had a terminated pregnancy due to an abortion (weight using the sample weights).
This seems unrealistically low and I wanted to know (1) if I am somehow looking at this wrong and (2) if others have used abortion data from the DHS or noticed the same thing.
Here is the code that I used:
gen weight=v005/1000000
gen abortion=0
replace abortion=1 if s230b==1
svyset v001 [pw=weight], strata(v022)
svy: mean abortion
Does this look like the correct way of looking at the abortion variable? Is there any reason that this would not give an accurate estimate of the prevalence of induced abortion (besides social desirability bias and respondents not wanting to admit that they had an abortion)? I am also wondering why the DHS does not include this in the final report. Is it a problem with the data or do the countries not want data on abortion published?
Any help would be greatly appreciated! Thank you so much.
Best regards,
Charlotte
|
|
|
|
| Re: Abortion question - Cote d'Ivoire 2012 [message #14615 is a reply to message #13939] |
Sun, 22 April 2018 17:37  |
kingx025
Messages: 95
Registered: August 2016
Location: Minneapolis. Minnesota
|
Senior Member |
|
|
One reason that reported rates for induced abortion may be low in many African countries is that abortion is not legal. In such a setting, there may be considerable pressure on women to reclassify an induced abortion in some other way--even in her own mind--as some other act, such as a miscarriage or a way of bringing on "blocked" menstruation. You might consider also reporting the numbers for "terminated pregnancies" rather than separately identifying induced abortions---even if the data allows you to do such separate calculations--as at least an alternative measure/upper bound estimate. I see you already acknowledge social desirability bias issues in your message, so what I'm saying here isn't new to you.
This topic is not my own area of specialization but I have heard presentations by researchers in this area who made the point about underreporting and misreporting in a setting where abortion is not legal.
Miriam King
Dr. Miriam King
IPUMS-DHS Project Manager (www.idhsdata.org)
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
