I have a survey dataset from two years of 2017 and 2018 and they are individual/respondent data. Each dataset has some common variables such as age, sex, zone. Each dataset has a variable called weight which is already created. This is not physical weight, but a weight variable which has been created given the survey population and size. Although they have some similar variables, they are not of the same individuals. But these individuals may have similar characteristics such as in terms of age, zone and sex. For example, lets say 'A' individual is 18 years old and male in 2017 and falls under zone 1. Another individual called 'B' is 18 years old and male in 2018 and falls under zone 1 as well. But they are not same individuals, rather different (A in 2017 and B in 2018). The weight for the individual is different across 2017 and 2018. The difference is that they are different individuals from different years and has different weight but they may have similar characteristics in terms of age, sex, zone.
Now my objective is to create a combined weight. This weight will be the same for both individuals A ans B since they have similar characteristics (both are 18 years old and male, falls under zone 1). This combined weight will be creater from weight of 2017 and 2018. The combined weight will be the same for both A and B. Lets say in 2017, A had 0.5 weight, B had 0.6 weight in 2018. Now I want to give the same weight to Both A and B because they have some similar characteristics (originally they had different weights). How to do that conceptually and in stata? I have thought of this:
Combined weight=weight of individual /mean (weight).
But I am not sure if this is correct. I have also comsidered to take mean of weight for all the individuals who has the similar characteristics and then do this:
by age sex zone: egen combined_weight, mean(weight)
But I am not sure if any of this is correct or has literature back up. If anybody can provide literature back up for any of these formula or maybe suggest a new method for calculating this combined weight, i would be grateful
Now my objective is to create a combined weight. This weight will be the same for both individuals A ans B since they have similar characteristics (both are 18 years old and male, falls under zone 1). This combined weight will be creater from weight of 2017 and 2018. The combined weight will be the same for both A and B. Lets say in 2017, A had 0.5 weight, B had 0.6 weight in 2018. Now I want to give the same weight to Both A and B because they have some similar characteristics (originally they had different weights). How to do that conceptually and in stata? I have thought of this:
Combined weight=weight of individual /mean (weight).
But I am not sure if this is correct. I have also comsidered to take mean of weight for all the individuals who has the similar characteristics and then do this:
by age sex zone: egen combined_weight, mean(weight)
But I am not sure if any of this is correct or has literature back up. If anybody can provide literature back up for any of these formula or maybe suggest a new method for calculating this combined weight, i would be grateful
