Home » Countries » Ethiopia » Calculating Age structure explanatory variables on Household Wealth
| Calculating Age structure explanatory variables on Household Wealth [message #11184] |
Wed, 16 November 2016 09:44  |
xrl1g11
Messages: 4
Registered: November 2016
Location: England
|
Member |
|
|
Dear DHS Staff,
I am using the 2011 Ethiopia DHS data set on SPSS. My thesis is about examining the role of age structure on household wealth. I am finding it difficult to construct the variables I need for my multinomial logistic regression analysis (household wealth is a multicategorical dependent variable).
The variables I need are:
youth dependency ratio (household members aged 0-14 and 65+ divided by household members aged 15-64)
number of young dependents aged under 15 per household
number of working age members (15-64) per household
(Other variables are included but I'm stuck with the demographic variables needed).
The information needed, for example, to create the new variable Number of Young Dependents per household, requires information on the age of household members and the household size (which includes information on the no. of households as well). I've recoded the variable age of household members into three distinct groups: young dependents (aged 0-14), working ages (15-64) and old dependents (65+). I've cross tabulated this new variable with the number of the de jure household members. The table produced shows me the information I need to now calculate the number of young dependents per household as it gives me information on how many household members aged 0-14, 15-64, 65+ live in a certain household size. However, I just can't seem to extract the information from the cross tab, and divide the number of household members under 15 by the number of households.
I attempted to export the cross tabulation to excel, and then manually do the calculation to calculate average number of children per household by household wealth. But I do not know if that's the appropriate way ( I was thinking to calculate it and then import the results back into spss, but I'm unsure if that will work).
I'm certain this should be relatively easy to perform on SPSS, but I just can't seem to figure out how to do it on SPSS. I hope it is possible on SPSS, as I don't think what I want to obtain is complex. I want to say for example, that households with a greater number of dependent children are more likely to be from poorer households than wealthier households. And that households with a higher youth dependency ratio are more likely to be poorer than richer.
I hope this made sense, I just need some solid guidance as to how to create the variables I need, using information from the household recode file, or the household member file. My unit of analysis is the household and ideally I want to use the household recode file, but I'm having even more trouble obtaining the information I need, as the age of household members has 22 response variables (HV105$01-$22) which makes things more tougher for me. I'd be happy to give any more information.
In short, I need to know how to create a youth dependency ratio, number of young dependents per household and number of working ages per household from the 2011 EDHS data set.
Regards,
Xavier
[Updated on: Wed, 16 November 2016 10:19] Report message to a moderator |
|
|
|
Current Time: Fri Apr 19 00:57:44 Coordinated Universal Time 2024
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
