Creation of new variable from tertile mean values

[Ginie Nathan]

Dear people,

Could you please help me on the following issue ?

I have a continuous variable describing the mental health score, I created tertile from this continuous variable, and tabstat to have the mean in each tertile. Now I would like to create a new variable of these tertile mean. Is that possible ? I tried to do all combination, without success.

Here is my data look like :

* Example generated by -dataex-. To install: ssc install dataex
clear
input float fatherK2
 0
 2
 1
25
 0
 1
 0
 2
 4
 3
 1
 3
 3
 0
 2
 0
 6
 1
 2
 3
 0
end
label values fatherK2 ght
label def ght 25 "No father at home", modify

xtile fatherK2_T3=fatherK2, nq(3)

tabstat fatherK2, stat (mean sd n) by (fatherK2_T3)

What I want is to extract the means from tabstat and create a new variable with it, such as the following example table. Because afterwards I need to do a linear regression.

tertile mean
1 xx
2 xx
3 xx

Many thanks for your help
Ginie

[Andrew Musau]

collapse fatherK2, by(fatherK2_T3)

See

help collapse

[Clyde Schechter]

by fatherK2_T3, sort: egen wanted = mean(fatherK2)

Added: Crossed with #2. The approach there will reduce the data set to a single observation for each tertile. The approach here maintains the original number of observations and creates a variable equal to the mean of the father_K2 in that tertile.

[Ginie Nathan]

Thank you so much that is exactly what I needed.