Generate variable with percentage of another variable

[MD Sorho]

Hello everyone,

I need to create a variable that measures neighborhood effects.

Let me explain: I have a dataset consisting of students from different classes / school and a variable that measures the retention rate. I want to create a variable that groups the retention rate percentage of the class without considering the individual student.

Any idea on how to do this would be very much appreciated.

Thank you
MD

[George Ford]

?

egen retrate_class = mean(retrate), by(class)

[Rich Goldstein]

please read the FAQ to learn how to ask questions that are capable of being answered and how to post data examples; at a guess, the following FAQ may help: https://www.stata.com/support/faqs/d...ng-properties/

[MD Sorho]

?

egen retrate_class = mean(retrate), by(class)

Hi,

I'd tried that before but it doesn't make a difference depending on the individual considered.

[Nick Cox]

https://www.stata.com/support/faqs/d...ng-properties/

[Benno Schoenberger]

I think this is quite easy to do, but without providing a useful data example, hardly anyone will be willing to help. Read again #3!

[MD Sorho]

I've finally found the solution. Thank to all

I put the code that I used for someone who looking for:


bysort classes: egen tot_students = count(classes)
bysort classes: egen total_retention = total(retention)
gen pc_retent = (total_retention - retention) / (tot_students - 1)

[Nick Cox]

The essence of the code in #7 was already given in the FAQ linked in #5, As said, this is a Frequently Asked Question.

Note that you are assuming thatretentionis never missing. If it is ever missing then your variable (which despite your name looks like a proportion to me, not a percentage) will be miscalculated for those individuals. Even if retention is missing, it's still possible to calculate the retention proportion for the other students.

So, more general code would be

gen pr_retent = (total_retention - cond(missing(retention), 0, retention)) / (tot_students - !missing(retention))