Question

SELECT student_id, section, count( * ) as total
FROM raw_data r
WHERE response = 1
GROUP BY student_id, section

There are 4 sections on the test, each with a different number of questions. I want to know, for each student, and each section, how many questions they answered correctly (response=1).

However, with this query, if a student gets no questions right in a given section, that row will be completely missing from my result set. How can I make sure that for every student, 4 rows are ALWAYS returned, even if the “total” for a row is 0?

Here’s what my result set looks like:

student_id  section     total
1           DAP--29     3
1           MEA--16     2
1           NNR--13     1  --> missing the 4th section for student #1
2           DAP--29     1
2           MEA--16     4
2           NNR--13     2  --> missing the 4th section for student #2
3           DAP--29     2
3           MEA--16     3
3           NNR--13     3 --> missing the 4th section for student #3
4           DAP--29     5
4           DAP--30     1
4           MEA--16     1
4           NNR--13     2 --> here, all 4 sections show up because student 4 got at least one question right in each section

Thanks for any insight!

UPDATE: I tried

 SELECT student_id, section, if(count( * ) is null, 0, count( * ))  as total

and that didn’t change the results at all. Other ideas?

UPDATE 2: I got it working thanks to the response below:

 SELECT student_id, section, SUM(CASE WHEN response = '1' THEN 1 ELSE 0 END ) AS total
 FROM raw_data r
 WHERE response = 1
 GROUP BY student_id, section

Answer 1

SELECT student_id, section, sum(case when response=1 then 1 else 0 end) as total
FROM raw_data_r GROUP BY student_id, section

Note that there’s no WHERE condition.