Checking the student grades in different semesters (long format)

[Hatim Tareq]

Dear All,

I have a data in a long format (students data with their grades in each semester). I would like to be able to check if the student has the same grades throughout the semesters, or does s/he moves up/down in terms of his grades (for example from 50 in semester 1 to 60 in semester2, .. )

I have tried to use the following code, to check if the grades are same in all the semesters

by id semester, sort: gen scores = 1 if semester[_n] = semester[_n-1]

But I am sure its not correct, and not sure how work around it,

- I was also thinking of reshaping my data into wide format, is the way to go?

Here is how my data looks:

id semester grade
1 1 90
1 2 90
1 3 90
1 4 90
2 1 85
2 2 .
2 3 80
3 1 50
3 2 60
3 3 70
3 4 70
4 1 95
4 2 90
4 3 85
4 4 80

Thank in advance for your help.

[Justin Niakamal]

Are you looking for something like this?

bys id (semester): gen scores = (grade == grade[_n-1])

Or this?

egen avg = mean(grade), by(id)
gen  scores = (grade == avg)
drop avg

[Hatim Tareq]

Thanks Justin.

It almost what I was looking for. But I just noticed that I wouldn't be able to report the changes in the grades if it captured in my data as percentage. So, I have updated my data as follow:

id semester grade
1 1 Excellent
1 2 Excellent
1 3 Excellent
1 4 Excellent
2 1 Very good
2 2 .
2 3 Very good
3 1 Pass
3 2 Good
3 3 Very good
3 4 Very good
4 1 Excellent
4 2 Excellent
4 3 Very good
4 4 Very good

So, what I am looking for is to able to count the number of the students: (1) Their grades never changed (first student with id 1). (2) Students who their grades have changed (from Excellent to Pass, or from Pass to Excellent, ...)

Hope its clear now.

[Hatim Tareq]

I have reshaped my data into wide format, then I used the "concat" function.

I think its better, in this case, to handle my question in wide format data.

Any suggestions?

[Andrew Musau]

So, what I am looking for is to able to count the number of the students: (1) Their grades never changed (first student with id 1). (2) Students who their grades have changed (from Excellent to Pass, or from Pass to Excellent, ...)

* Example generated by -dataex-. To install: ssc install dataex
clear
input byte(id semester) str9 grade
1 1 "Excellent"
1 2 "Excellent"
1 3 "Excellent"
1 4 "Excellent"
2 1 "Very good"
2 2 "."        
2 3 "Very good"
3 1 "Pass"     
3 2 "Good"     
3 3 "Very good"
3 4 "Very good"
4 1 "Excellent"
4 2 "Excellent"
4 3 "Very good"
4 4 "Very good"
end

bys id (grade): gen wanted1=grade[1]==grade[_N]
bys id (sem): gen wanted2=grade=="Pass" & grade[_n-1]=="Excellent"|grade=="Excellent" & grade[_n-1]=="Pass"
bys id (wanted2): replace wanted2=wanted2[_N]
egen tag= tag(id)
count if wanted1 & tag
count if wanted2 & tag

Res.:

. count if wanted1 & tag
  1

. count if wanted2 & tag
  0