If you agree to include some code in your EG project, the previous answer is the simplest you can do. If you want to stick to EG "point-and-click" way of life, I suggest the following...
Start a query on your data.
Compute 2 new variables :
1) one named "SEQUENCE" (for example) with the following formula : MONOTONIC()
2) one named "FIRST" with the following formula : MIN(CALCULATED sequence)
The SEQUENCE variable will consist in numbers given to each data line, in a sequential way (hence its name) : 1, 2, 3, ... The FIRST variable will compute the minimum sequence number for each group. Don't forget to add in the query interface that the FIRST variable is to be computed for each group !
One last thing : in the "Group Filtering" tab, add the condition that SEQUENCE = FIRST. And then you're done.
The SAS code produced should somehow look like this :
PROC SQL ;
SELECT *,
MONOTONIC() AS order,
MIN(CALCULATED order) AS first
FROM sashelp.class
GROUP BY age
HAVING order = first
;
QUIT ;
This gives you the first child for each group of Age, from the SASHELP.CLASS dataset.