The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Domestic Violence » DRC DHS-2 sampling question for DV module
DRC DHS-2 sampling question for DV module [message #3549] Mon, 05 January 2015 14:06 Go to next message
carlson_monica is currently offline  carlson_monica
Messages: 9
Registered: November 2014
Location: United States
Member
I am just beginning to explore the domestic violence data for the recently released DHS-2 for the Democratic Republic of Congo. My question is to clarify the total number of women included in the DV sample and understand why the number of cases in the database are significantly lower than planned in the sampling methodology as I didn't see an explanation in the final report.

The total survey sample for the individual survey was 18,827 women. My understanding from the methodology laid out in the annex of the final report was that the sampling plan was to interview 50% of the women for the domestic violence module (so about 9,413 women). Taking a look in the database at women selected for the DV module (using the SELIGDV variable), I see that 9,588 values are "missing"; 6,857 are "yes" (i.e. administered the DV module); whereas 2,382 have a "no" value meaning they were not selected for the domestic violence module. I wanted to make sure this is accurate and to understand why the full 50% of the 18,827 were not selected for the DV module, if the total number of DV respondents were indeed 6,857 women only.

Many thanks for your support.



Re: DRC DHS-2 sampling question for DV module [message #3577 is a reply to message #3549] Tue, 13 January 2015 10:41 Go to previous messageGo to next message
Liz-DHS
Messages: 1516
Registered: February 2013
Senior Member
Dear User,
Here is a response from our experts, Dr. Hong, Dr. Ren and Mr. Amadou Sow:

1.	DV module was administered in 50% of the households interviewed 

                       | result of
   household selected  | household
for domestic violence  | interview
               module  | completed |     Total
-----------------------+-----------+----------
household not selected |     9,086 |     9,086 
   household selected  |     9,085 |     9,085 
-----------------------+-----------+----------
                Total  |    18,171 |    18,171

2.	In those households, if there is one or more eligible women, one woman was selected to be interviewed for DV module.
3.	In 2,121 of those 9,085 households, there is no eligible women (15-49). Therefore only 9,085-2,121=6,964 women were selected for DV interview.

number of  |  household selected
  eligible | for domestic violence
  women in |        module
household | household  household |     Total
-----------+----------------------+----------
         0 |     1,934      2,121 |     4,055 
         1 |     5,277      5,229 |    10,506 
         2 |     1,360      1,254 |     2,614 
         3 |       389        347 |       736 
         4 |        91         89 |       180 
         5 |        21         34 |        55 
         6 |        11          7 |        18 
         7 |         2          3 |         5 
         8 |         1          0 |         1 
         9 |         0          1 |         1 
-----------+----------------------+----------
     Total |     9,086      9,085 |    18,171 

display 9085-2121
6964

4.	Factor in the overall women response rate of 98.6% and response rate for DV, subtract the lack of privacy => 6,811

Re: DRC DHS-2 sampling question for DV module [message #3578 is a reply to message #3577] Tue, 13 January 2015 12:35 Go to previous messageGo to next message
carlson_monica is currently offline  carlson_monica
Messages: 9
Registered: November 2014
Location: United States
Member
Thank you -- very helpful. I am surprised by the large number of households with no eligible woman between 15 and 49.

I have just one follow up question on the DRC DHS dataset:

I am having problems recreating the physical and sexual violence prevalence from the 2014 DRC DHS (51.8% of women that have suffered physical violence since age of 15 and 27% of women have suffered sexual violence since the age of 15) using the dataset variables. I am interested in the combination of variables that tell experience or not of any physical or sexual violence whether by husband or other perpetrator, for married or never married respondents (therefore experience or not of physical and sexual violence for all women sampled for DV module).

I was planning to use variable D105 to reflect women who experienced physical or sexual violence by their husbands/partners. However, to capture experience of violence that is not IPV for married and non-married women, I was planning on using question #1316 from the DRC Survey instrument (that asks married women about physical violence by those other than husbands AND which asks women who have never been married or lived with a man if they have experienced physical violence since the age of 15).

I don't see the response to this question clearly reflected in the actual dataset -- is it reflected/imbedded in variable D115 which characterizes type of perpetrator of violence by non partners for both married and non-married women? I was a bit confused because in the Individual recode manual on page 106 the "base" for D115 is "women currently in union", yet on the survey the question is asked of both women in union and women not in any union/never married.

Thank you in advance for any guidance on this.
Monica
Re: DRC DHS-2 sampling question for DV module [message #3607 is a reply to message #3578] Thu, 15 January 2015 12:22 Go to previous messageGo to next message
Liz-DHS
Messages: 1516
Registered: February 2013
Senior Member
Dear Monica,
Your query has been forwarded to one of our experts. We should have a response for you in a day or two. Thank you!
Re: DRC DHS-2 sampling question for DV module [message #3644 is a reply to message #3607] Wed, 21 January 2015 11:26 Go to previous messageGo to next message
carlson_monica is currently offline  carlson_monica
Messages: 9
Registered: November 2014
Location: United States
Member
Thank you so much, appreciate it!

Monica Carlson
Tulane University
Re: DRC DHS-2 sampling question for DV module [message #3691 is a reply to message #3644] Wed, 28 January 2015 11:40 Go to previous messageGo to next message
Liz-DHS
Messages: 1516
Registered: February 2013
Senior Member
Dear Monica,
We have not forgotten about you. Our experts are aware of your query and will respond as soon as possible. Thank you!
Re: DRC DHS-2 sampling question for DV module [message #3697 is a reply to message #3691] Wed, 28 January 2015 21:10 Go to previous messageGo to next message
carlson_monica is currently offline  carlson_monica
Messages: 9
Registered: November 2014
Location: United States
Member
Thanks so much -- looking forward to the response! Many thanks for your great support.

Monica Carlson
Tulane University
Re: DRC DHS-2 sampling question for DV module [message #3728 is a reply to message #3697] Mon, 02 February 2015 17:44 Go to previous messageGo to next message
Trevor-DHS is currently offline  Trevor-DHS
Messages: 803
Registered: January 2013
Senior Member
Yes, the response to question 1316 is embedded in the d115 series of variables. The easiest variable in this series to check is actually d115y. Below is a mapping of the codes in q1316 to d115y:
q1316 => d115y
1 => 0
2 => 1
3 => 6
The data here are available for both all women as can be seen if you tabulate d115y by v501, so the recode manual is not correct in this case.
Re: DRC DHS-2 sampling question for DV module [message #4241 is a reply to message #3577] Fri, 24 April 2015 11:03 Go to previous messageGo to next message
carlson_monica is currently offline  carlson_monica
Messages: 9
Registered: November 2014
Location: United States
Member
Just a quick questions about the DV sample for DRC. When I run the Province breakdown of the 6,811 women in the DV sample, and run descriptives with the DV weight on, I was a bit surprised by the sample breakdown by province -- mostly with Bandundu having the highest N and % of the sample? Just wanted to double check this with you all. Is this correct?

V101 Province
Frequency/Percent
Valid 1 Kinshasa 823/12.1
2 Bandundu 1089/16.0
3 Bas-Congo 312/4.6
4 Equateur 913/13.4
5 Kasai-Occidental 426/6.3
6 Kasai-Oriental 642/9.4
7 Katanga 648/9.5
8 Maniema 189/2.8
9 Nord-Kivu 561/8.2
10 Orientale 661/9.7
11 Sud-Kivu 546/8.0
Total 6811/100.0



Re: DRC DHS-2 sampling question for DV module [message #4284 is a reply to message #4241] Thu, 30 April 2015 18:12 Go to previous messageGo to next message
Trevor-DHS is currently offline  Trevor-DHS
Messages: 803
Registered: January 2013
Senior Member
That looks in line with the distribution of the population of women given in table 3.1 and matches the distribution given in table 18.1, so I'm not seeing a problem. Please also read the information given in Chapter 1 and Appendix A for the sample.
Re: DRC DHS-2 sampling question for DV module [message #6896 is a reply to message #3577] Fri, 31 July 2015 00:03 Go to previous message
seiha is currently offline  seiha
Messages: 1
Registered: July 2015
Location: poipet
Member
I agree with the post.



คาสิโนออนไลน์ ในรูปแ บใหม่ล่าสุดจาก ปอยเปต คาสิโน รีสอร์ท ได้ที่นี่ royal1688
Previous Topic: Logit regressions using two different weights
Next Topic: Multilevel weights
Goto Forum:
  


Current Time: Thu Nov 21 14:34:56 Coordinated Universal Time 2024