Allowing multiple majors in an Enrollment Cube

Question

I am working on an Enrollment Cube for my university. One of its dimensions has a hierarchy of College -> Department -> Major, with each student having one major per term. However, there are also students who are dual-majoring and therefore have two Majors. Now I know that as a rule an OLAP cube can only have a 1-to-1 relationship between the base levels of the dimensions, so the easiest way I can think of replicating this relation is to have two separate Dimensions, one for a Primary Major and one for a Secondary Major. However, this is less than ideal, as an analysis that breaks down a specific major will only give a total of those who have that major as either their Primary or Secondary, rather than including both.
One possible work-around I theorized about was having two Major Key fields on the fact table, both of them being Foreign Keys to the Primary Key of the Majors dimension. The Administration Tool threw on error at this though, and the results did not behave correctly on our test box (though it did seem promising at some levels). As a stretch, I then tried having a Major Group table that sat between the Majors dimension table and the fact table, with a bridge table giving me a many-to-many relation between Major Group and Majors (so that Major 1 would be in Major Group 1, and Major 2 will be in Major Group 2, and Major 1 and Major 2 would also be together in Major Group 3 for students who are dual majoring in Majors 1 and 2). Of course, this created a disconnect between the dimension and the fact table in Business Intelligence, and wouldn't break the numbers down at any level other than the Grand Total.
Might any of you have any ideas, or do you think the only option is to go with two Dimensions?

Answer 1

If you include a placeholder for NULL or NA in the Major and Minor dimensions, you can always have 2 dimensions and play with the stored measures with additional derived measures to reach across the Minor dimension and get you the results you want.
E.g:
If you have 4 subjects: Maths (Ma), Physics (Ph), Chemistry (Ch), Social Sciences (SS) then define a Not Applicable (NA) Subject with superficial key -1 or -99 or whatever.
Define Total level for both Major and Minor with aggregation = Sum.
Student with Maths major (no minor) goes to Ma-NA cell or combination (loaded as 1).
Student with Maths Major and Physics Minor goes to Ma-Ph cell/combination (loaded as 1)
How many Students take Maths = Count Measure (Major = Ma, Minor = Total) + Count Measure (Major = Total, Minor = Ma)
How many Students dont take Chemistry = Total - Students who take Chemistry = Count Measure (Major = Total, Minor = Total) - ( Count Measure (Major = Ch, Minor = Total) + Count Measure (Major = Total, Minor = Ch) )
If you have a Student Dimension then you need to retain the Student Id to prevent double/multiple counting. The aggregation would change based on each dimension (usually max for other dimensions and sum for Student Dimension).
-Shankar
NOTE: Terminology used ... Major = Primary Major, Minor = Secondary Major