| variable definition [message #17984] |
Fri, 09 August 2019 15:47 |
Nyamwiza
Messages: 3
Registered: August 2019
|
Member |
|
|
Dear Admin and members, I need some help. I asked this question earlier but realized that I was in the India forum.
I am working on a project where my main outcome variables are;
Type of cooking fuel hv226
Exposure to tobacco smoke in the hone hv241
Exposure to biomass smoke (derived as Tab hv241 if solid==1) where solid is those using biomass fuels for cooking.
I have currently done this using the PR file, Should I use the HR file instead?
My challenge is with some of the variables that are my independent variables;
How is wealth measured in the PR file? Is every individual member measured or is it household wealth? I went through the do file but I didn't understand.
If the aim is to study the prevalence of the above outcomes and describe their variation with sex, age, wealth quintile, province, etc? should I use PR or HR file? I am confused.
I am also looking at smoking in men and women. Should I merge them? In my initial analysis, I had done it separately (Stata Do file attached).
I am looking at exposure to biomass in the home with the independent variables and I get no statistically significant difference with age, sex, and residence (urban/rural) which is not what we know from the literature. Should I have specified age groups instead of looking at all the 49,064 household members?
Regarding province, which of the 2 variables is the most accurate? In my analysis, province 6 was the poorest and had the highest exposures but an earlier analysis showed that province 2 was the poorest with the highest risk.Am I doing some wrong? Some guidance, please.
[Updated on: Fri, 09 August 2019 16:06] Report message to a moderator |
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
