demo, Sql Server - rextester

Language: Layout:

create table s(user_id int, exam_id int, question_id int, submitted_option_id int);

create table q(question_id int, correct_option_id int);

insert q values (1,11), (1,12), (2,21), (3, 31), (3, 32), (3, 33), (4,41), (4,42), (4,43), (4,44), (5,51), (5,52), (6,61);
-- select count(*) from q;

insert s values (1, 0, 1, 11), (1,0,1, 12) -- correct
, (1,0,2,29) -- wrong
, (1,0,3,31), (1,0,3,32), (1,0,3,33), (1,0,3,39) -- 31-33 is the correct answer, but user has submitted also 39.
, (1,0,4,41), (1,0,4,42) -- two out of four correct. The rest were not submitted. 
, (1,0,5,51), (1,0,5,59) -- first is correct, second is wrong.
-- nothing for question 6
;

select question_id, case when cnt = sum_test then 1 else 0 end as mark
from (
select question_id, count(*) cnt, sum(test) sum_test
from (
select coalesce(q.question_id, s.question_id) as question_id, correct_option_id, submitted_option_id
, case when correct_option_id = submitted_option_id then 1 else 0 end as test
from q full outer join s on q.question_id = s.question_id and q.correct_option_id = s.submitted_option_id
-- order by coalesce(q.question_id, s.question_id), coalesce(q.correct_option_id, s.submitted_option_id)
) x
group by question_id
) y

View schema
Execution time: 0,02 sec, rows selected: 6, rows affected: 24, absolute service time: 0,19 sec