EGypt's DHS 2008 sample weighting [message #2421] |
Tue, 17 June 2014 08:37 |
dralsumry@gmail.com
Messages: 1 Registered: June 2014 Location: UK
|
Member |
|
|
HI,
im trying to use the EDHS2008 to estimate the prevalence of modern contraception use in the ever-married women questionnaire,and I realised I have to weight the sample for the prevalence to be representative to that of the population,Im using STATA 11.0 ,I know i need to set the syntax svyset but don't know what variables to go in the syntax?
Thank you
|
|
|
Re: EGypt's DHS 2008 sample weighting [message #2434 is a reply to message #2421] |
Tue, 17 June 2014 17:17 |
Liz-DHS
Messages: 1516 Registered: February 2013
|
Senior Member |
|
|
Dear User,
Please check our website's section on Using Datasets for Analysis http://dhsprogram.com/data/Using-Datasets-for-Analysis.cfm
This will step you through the analysis process and help you locate resources you may need. The map which was distributed with the data set will also help you identify the variables you need. The Guide to DHS Statistics http://dhsprogram.com/pubs/pdf/DHSG1/Guide_to_DHS_Statistics _29Oct2012_DHSG1.pdf can also be very useful along with the Standard Recode Manual http://dhsprogram.com/pubs/pdf/DHSG4/Recode6_DHS_22March2013 _DHSG4.pdf. The Standard Recode Manual has a description of the variables along with commentary on how and why they are used. Your map has the recoded variables which are specifically for the data set you downloaded.
Here's a little bit of CSPro code that might help identify some of the variables you may need:
box v632 => V632W;
1 => 1;
2 => 3;
3 => 2;
=> 4;
endbox;
modmet2 = notappl;
tradmet2 = notappl;
nousing = notappl;
meth = V312;
box meth => method;
0 => 0;
1-3 => meth; { 1-Pill, 2-IUD, 3-Injections }
11 => 4; { 4-Implants }
17 => 5; { 5-Diaphragm/foam jelly}
5-7 => METH+1;
16 => 9 ; { 12-Emergency contraception }
8,9 => meth+5; { 13-Periodic abstinence, 14-Withdrawal }
18 => 15;
10 => 16; { 15-Other }
=> default;
endbox;
if V313 = 0 then { not using }
nousing = 1
elseif V313 = 3 then { using modern }
modmet2 = method;
elseif V313 = 1 | V313 = 2 then { using traditional & others }
tradmet2 = method;
endif;
|
|
|
Re: EGypt's DHS 2008 sample weighting [message #14603 is a reply to message #2421] |
Sun, 22 April 2018 14:55 |
kingx025
Messages: 95 Registered: August 2016 Location: Minneapolis. Minnesota
|
Senior Member |
|
|
Another way to identify variables that you may be interested in for a particular topic--say, current use of contraception by method in Egypt--is to look on the IPUMS-DHS website.
Here are steps involved:
1) Go to idhsdata.org
3) Click on Select Data
3) Choose women as the unit of analysis
4) From the drop-down menu of topics, click on family planning---Recent Use of Family Planning
Here's what I pulled up when I restricted the samples to Egypt, using the Sample Selection tool. An X means the variable is available for the sample listed above it:
https://www.idhsdata.org/idhs-action/variables/group?id=fp_u serecent
5) If you want to learn more about a variable, click on its name, and you can read a description, see (unweighted) codes and frequencies, see who was asked the question (universe), and read the survey text.
6) If you want to make a customized dataset using any of these variables, log in with your DHS user email and password and add samples and variables of interest to your "data cart" (to get a free data file in your preferred format).
Miriam King
Dr. Miriam King
IPUMS-DHS Project Manager (www.idhsdata.org)
|
|
|