The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » General » Child Marriage (How to estimate number of annual child marriages )
Child Marriage [message #29839] Mon, 12 August 2024 05:24 Go to next message
pkaburi is currently offline  pkaburi
Messages: 9
Registered: February 2014
Location: Nairobi
Member
How can one use the prevalence of child marriages - those married by 18 years from the most current DHS data. In my mind, I was planning to use the prevalence of child marriages for those 20-24 years and estimated population 20-24 years to estimate the number of girls who got married before 18 years and then divide by 4 (age 15,16,17 and 18). Assuming that marriages are uniformly distributed. The reason I am using 20-20 is because this more current cohort rather than use for 20-49 years. Is this a sensible approach.
Your suggestion and support is appreciated.
Re: Child Marriage [message #29853 is a reply to message #29839] Tue, 13 August 2024 14:44 Go to previous message
Janet-DHS is currently offline  Janet-DHS
Messages: 888
Registered: April 2022
Senior Member
Following is a response from DHS staff member, Tom Pullum:

The indicators you are referring to are descriptive and are used to compare different countries, different sub-populations, and changes over time. If you want more detail, for example on the numbers of marriages at ages 15, 16, 17, and 18, you would need to go to the data files. It would definitely not be safe to assume equal numbers of marriages at each of those ages, but also it is NOT NECESSARY to make such an assumption. The data files include single years of age at marriage.

"Before 18" means before the 18th birthday. No one who married before their 18th birthday was 18 years old when they got married.

The calculation for women age 20-24 at the time of the survey is included in the relevant table in the final report because it is more recent, as you say. The calculation for age 20-49 has many more cases and that's a statistical advantage.

An important issue is that the women in the subgroup or the denominator must have had full exposure to the risk of the outcome. For example, women age 20+ have fully experienced all ages below 20. If you want to estimate the percentage of women who marry before age 20, you have to restrict to women age 20+
Previous Topic: Clarification on v511 in IR
Next Topic: Weithted analysis in R
Goto Forum:
  


Current Time: Fri Nov 22 23:30:16 Coordinated Universal Time 2024