How to run an analysis on only certain observations in a dataset

[Aditi Rao]

Hello everyone!

I have a wide dataset with several rows per individual. How would you make sure to consider only one of the observations while calculating things like mean age. I only want to use the value of the first observation for each individual.

Let's say I want to calculate mean age and it varies per row because these are values recorded over several years. How would I consider only the first observation for each individual

Here is an example:

Beneficiary ID	Unique ID	Age
EEEAVH9E	1	72
EEEAVH9E	1	72
EEEAVH9E	1	73
EEEAVG7F	2	80
EEEAVG7F	2	81

Any help would be appreciated!

[Carlo Lazzaro]

Aditi:
do you mean something along the following lines?:

use "https://www.stata-press.com/data/r17/nlswork.dta"
. bysort idcode (year): gen wanted=age if _n==1


. tabstat age, stat(count mean sd p25 p50 p75 min max)

    Variable |         N      Mean        SD       p25       p50       p75       Min       Max
-------------+--------------------------------------------------------------------------------
         age |     28510  29.04511  6.700584        23        28        34        14        46
----------------------------------------------------------------------------------------------

.

[William Lisowski]

I think Carlo meant

. use "https://www.stata-press.com/data/r17/nlswork.dta"
(National Longitudinal Survey of Young Women, 14-24 years old in 1968)

. bysort idcode (year): gen wanted=age if _n==1
(23,825 missing values generated)

. tabstat wanted, stat(count mean sd p25 p50 p75 min max)

    Variable |         N      Mean        SD       p25       p50       p75       Min       Max
-------------+--------------------------------------------------------------------------------
      wanted |      4709  23.94479  5.940831        20        22        26        14        45
----------------------------------------------------------------------------------------------

or equivalently

. use "https://www.stata-press.com/data/r17/nlswork.dta"
(National Longitudinal Survey of Young Women, 14-24 years old in 1968)

. bysort idcode (year): gen wanted = _n==1

. tabstat age if wanted, stat(count mean sd p25 p50 p75 min max)

    Variable |         N      Mean        SD       p25       p50       p75       Min       Max
-------------+--------------------------------------------------------------------------------
         age |      4709  23.94479  5.940831        20        22        26        14        45
----------------------------------------------------------------------------------------------

[Carlo Lazzaro]

William is correct (thank you!) and I was sloppy in my previous reply.

[Aditi Rao]

Thank you so much for your responses. There are multiple observations within the same year in some cases, so I just want to be able to consider the first observation for the individual irrespective of year. Is this possible?

[William Lisowski]

generate seq = _n // keep the observations in the same order they are in currently within each UniqueID
bysort UniqueID (seq): generate wanted = _n==1
tabstat age if wanted