The DHS Program User Forum: India

Home » Countries » India » NFHS4

Show: Today's Messages :: Show Polls :: Message Navigator

NFHS4 [message #14114]

Mon, 19 February 2018 12:14

dumanb
Messages: 4
Registered: February 2018
Location: Burnaby

Member

I am hoping that someone can help me with the following.

I am creating a dummy variable using NFHS4 which takes value 1 if a woman had her first sex before her union with her first husband. So I use the following code in Stata

**************************************
gen sex_age=.

replace sex_age=v531 if v531>0

replace sex_age=. if v531==97 | v531==98

**************************************
gen preunion_sex=.

replace preunion_sex=1 if sex_age<v511

replace preunion_sex=0 if sex_age>=v511 | v531==0 | v525==96

replace preunion_sex=. if sex_age=.

tab v501 preunion_sex

The tab's results show irregularly high preunion sex rates among widowed/divorced/separated/deserted. Am I doing something wrong? or is it just the way the data is?

Report message to a moderator

Re: NFHS4 [message #14292 is a reply to message #14114]

Mon, 19 March 2018 10:35

Bridgette-DHS
Messages: 3190
Registered: February 2013

Senior Member

Following is a response from Senior DHS Stata Specialist, Tom Pullum:

If you enter "tab v511 v501, m" you will see that age at first cohabitation is not asked for women who are widowed/divorced/separated.

Note that it is possible for age at marriage and age at first sex to be the same, but for first sex to have occurred before the marriage.

You have to be very careful with the interpretation of any difference between v511 and v535. There has been some imputation of v531 and there is a very strong norm against premarital sex.

index.php?t=getfile&id=835&private=0

Attachment: tab v511 v501.JPG
(Size: 27.76KB, Downloaded 855 times)

Report message to a moderator

Re: NFHS4 [message #14314 is a reply to message #14292]

Wed, 21 March 2018 19:01

dumanb
Messages: 4
Registered: February 2018
Location: Burnaby

Member

Thanks a lot!
You're very helpful. I can't believed I haven't noticed missing values for this group of women.

What I am interested is the variation over time in premarital sex, so overestimation or underestimation of the variable (for the reasons you've mentioned) won't be an issue.

Best.

[Updated on: Wed, 21 March 2018 19:01]

Report message to a moderator

Previous Topic:	Out of Pocket Expenditure of last Delivery
Next Topic:	District analysis

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Mon Nov 11 09:41:26 Coordinated Universal Time 2024