Weighted average by two categories

[Maye Ehab]

Dear All,

I have data for the monthly wage by gender and categories of commuting time. I am trying to calculate the weighted average of the monthly wage by using the number of people in each commuting category by gender. An example of the data is shown below.

* Example generated by -dataex-. To install: ssc install dataex
clear
input float(commute female mnthwg)
1 0       500
1 1 1883.3334
1 0 198.33333
1 0       600
1 1       400
1 0      1500
1 1       345
1 0 1188.5714
1 0       364
1 0       550
end
label values commute commutecate
label def commutecate 1 "less than 16 min." ///
                          2 "from 16 to 32 min." ///
                          3 "from 33 to 60 min." ///
                          4 "from 61 to 90 min." ///
                          5 "from 91 to 720 min.", modify
label values female sex
label def sex 0 "male", modify
label def sex 1 "female", modify

I tried the following codes but it only gave me the average not the weighted average:

egen weightwage = wtmean(mnthwg), weight(commute)

Many thanks,
Maye

[Nick Cox]

That syntax (using _gwtmean from SSC, as you are asked to explain) should give the single weighted mean for all the data.

https://www.statalist.org/forums/for...and-collapsing gives suggestions for weighted mean calculations.

[Maye Ehab]

Thank you
I did the following with reference to the suggested link:

bysort commute female : egen denom = total(weight) 
by commute female : egen numer = total(weight * mnthwg) 
gen wtmean = numer/denom

[Abdul Adam]

PS: posted this before seeing Nick's link.
Is this what you want:

egen weightwage = wtmean(mnthwg), weight(commute) by(female)

[Maye Ehab]

Thank you for your reply.
Both methods yield the same result.