The DHS Program User Forum: Fertility » Teenage Pregnancy in BDHS 2017-18

Home » Topics » Fertility » Teenage Pregnancy in BDHS 2017-18 (Replicate Teenage Pregnancy table 5.11)

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

Teenage Pregnancy in BDHS 2017-18 [message #29995]

Sun, 08 September 2024 05:43

toukir86
Messages: 10
Registered: November 2023

Member

Dear Experts,

I want to determine the factors associated with teenage pregnancy and motherhood. From a previous topic discussion I found the following code.

**
gen teenmom=0 if v013==1
replace teenmom=1 if (v201>=1 | v213==1) & v013==1

gen wt = v005/1000000
svyset [pw=wt], psu(v021) strata(v023) singleunit(centered)
gen awfact = awfactt/100
svy: ratio teenmom/awfact, over(v012)

svy: ratio teenmom/awfact
**

It provides me the value of proportion, which aligns with the report. However, I want to generate n=4782 of Table 5.11 Teenage pregnancy and motherhood. Please let me know how I can do this.

Thank you in advance.

[Updated on: Sun, 08 September 2024 06:38]

Report message to a moderator

Re: Teenage Pregnancy in BDHS 2017-18 [message #30013 is a reply to message #29995]

Tue, 10 September 2024 12:32

Janet-DHS
Messages: 893
Registered: April 2022

Senior Member

Following is a response from DHS staff member, Tom Pullum:

Unfortunately, to match this table you have to use "all-women factors", as described in the footnote to the table. The choice of factor will depend on the panel. To get the 4782, you can use this line "tab v013 [iweight=(v005/1000000)*(awfactt/100)]". The weighted frequency for age 15-19 rounds to 4782. To get the frequencies for single years of age, use "tab v012 if v013==1 [iweight=(v005/1000000)*(awfactt/100)]". For wealth quintiles, use "tab v190 if v013==1 [iweight=(v005/1000000)*(awfactw/100)]". Note that for wealth, you use awfactw (w for wealth) rather than awfactt (t for total). The other panels require other versions of awfact (enter "describe awfact*").

You can get more information about all-women factors from the Guide to DHS Statistics. They are only required for surveys that only include ever-married women in the women's interview. They are a huge nuisance.

Report message to a moderator

Re: Teenage Pregnancy in BDHS 2017-18 [message #30015 is a reply to message #30013]

Tue, 10 September 2024 14:25

toukir86
Messages: 10
Registered: November 2023

Member

Thanks for your prompt reply; it worked out to generate 4782. However, when I use the DHS-provided code to generate the estimates of teenagers who have had a live birth, pregnancy with the first child, and have begun childbearing, the estimates do not match the report estimates. Where am I doing wrong?

//Teens (women age 15-19) who have had a live birth
*gen fe_teen_birth = v201 if v013==1
recode v201 (0=0 "No") (1/max=1 "Yes") if v013==1, gen(fe_teen_birth)
label var fe_teen_birth "Teens who have had a live birth"
//Teens (women age 15-19) pregnant with first child
gen fe_teen_preg = 0 if v013==1
replace fe_teen_preg = 1 if v201==0 & v213==1 & v013==1
label val fe_teen_preg yesnolabel
label var fe_teen_preg "Teens pregnant with first child"

//Teens (women age 15-19) who have begun childbearing
gen fe_teen_beg = 0 if v013==1
replace fe_teen_beg = 1 if (v201>0 | v213==1) & v013==1
label val fe_teen_beg yesnolabel
label var fe_teen_beg "Teens who have begun childbearing"

Also, can you let me know if I can use it to model the logistic regression model using the weight in the same way?

Attachment: Screenshot 2024-09-11 002203.png
(Size: 16.03KB, Downloaded 28 times)

[Updated on: Tue, 10 September 2024 14:25]

Report message to a moderator

Re: Teenage Pregnancy in BDHS 2017-18 [message #30049 is a reply to message #30015]

Mon, 16 September 2024 16:12

Janet-DHS
Messages: 893
Registered: April 2022

Senior Member

Following is a response from DHS staff member, Tom Pullum:

This is not an easy table. Besides having to change the weights with all-women factors, which are covariate-specific, you have to make some adjustments for ages below age 15, because the survey is limited to women age 15+. We prepared a Methodological Report on the topic of under-15 fertility.

I see you are using Stata code from GitHub. I don't have time to review and test it, but earlier this year I revised the GitHub code for DHS-8 (this has not yet been posted on GitHub). BD 2017-18 is not a DHS-8 survey but I think it may help. Look for generic table 5.12, which is numbered 5.11 in this BD report. Here is a link to the program, as a text tile: HERE.

Because the awfactors are different for different covariates. You cannot do multivariate analysis with logit regression.

If you search the forum you will find a discussion about constructing a revised or pseudo-IR file that incorporates the AW adjustment into the weight variable v005. If you do that, you can do logit regression but you will not get an exact match to DHS tables.

I am traveling and cannot provide much more help on this. Hope you can figure it out. As you may know, a new BD survey was just released.

Attachment: FE_FERT_do.txt
(Size: 21.42KB, Downloaded 27 times)

Report message to a moderator

Re: Teenage Pregnancy in BDHS 2017-18 [message #30054 is a reply to message #30049]

Mon, 16 September 2024 22:48

toukir86
Messages: 10
Registered: November 2023

Member

Thank you so much. I will try the attached code in the newly updated dataset. Your guidance has given me a clearer direction, and I will do my best to figure things out from here. Thanks again for your time and support.

Report message to a moderator

Previous Topic:	Wanted Fertility Rate
Next Topic:	Birth interval

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Thu Nov 28 04:33:33 Coordinated Universal Time 2024