The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Nutrition and Anthropometry » definition of ha5_ and hb5_
definition of ha5_ and hb5_ [message #1712] Tue, 01 April 2014 15:38 Go to next message
abarenbe is currently offline  abarenbe
Messages: 1
Registered: April 2014
Location: US
I'm wondering where I can find a clear definition of the ha5 and hb5 variables? These are height for age standard deviations, but it not specified in the data from what population the data is taking SD from. Looking at the Ethiopian data it seems that gender was not controlled for causing women's values to be generally lower then men's (about 1.8 SD lower on average.)

Thank you
Re: definition of ha5_ and hb5_ [message #1713 is a reply to message #1712] Tue, 01 April 2014 18:27 Go to previous message
Messages: 292
Registered: March 2013
Senior Member

Hi Andy,

First, are you referring to ha5 and hw5/hc5? I see an ha5 variable, but not hb5. HA5 relates only to respondent women, and hc5/hw5 only to children.

Since HA5 is a measure of women's height and refers to respondent women, there aren't any men. Is it possible you are tabulating it wrong in some way (or just assuming that mean score for men is 0)? A mean of -1.8 would be not unexpected in a population of poor women - these height standards are referenced to a healthy, well-nourished population by gender and age. From the recode manual:

"The NCHS/FELS/CDC reference standard only contains data for children up to the age of 18 years. For all women aged 18 and over, the value of 215 months (17 years, 11 months) is used for their age, on the assumption that women are fully grown by the age of 18."

The standards are both age- and gender-specific and come from the late 1970's - I'm pretty sure it was based off a group of kids from Iowa. HW70 uses the new and improved WHO growth standards from 2007, which use a more diverse and larger base of healthy children.

So - yes, the standards are gender-specific but for the population of Ethiopian women they are not mean 0 (or variance 1, for that matter). HA5 and HC5 are mean 0 and variance 1 for the reference population of healthy, well-nourished people.
Previous Topic: Replicating table counts for children under 5 in Bangladesh DHS
Next Topic: child feces disposal
Goto Forum:

Current Time: Tue May 21 19:32:02 Coordinated Universal Time 2024