Star Schema Design: Use 2 Dimensions or 1 Conformed Dimension when source systems relate many to one (N:1)?

Question

I am creating a star schema to model Terms and Classes at a school.

The Learning Management System (LMS) - where the classes take place, associates each class to a particular LMS Term.

The Student Information System (SIS) - where the students register for classes, models the Terms in a more granular way than the LMS. Thus, there are multiple SIS Terms for each LMS Term.

Each Fact record is at the granularity of a Student within a Class, and associates to 1 LMS Term.

It seems I could make 2 dimension tables: DimSisTerm and DimLmsTerm.

Or, I could make 1 conformed dimension table: DimTerm

In the case of a single conformed dimension, there would be 1 record for each SIS Term, however the LMS Term key and it's attributes would repeat for all related SIS Term records.

Can someone who has experienced this before provide their guidance as to what the trade-offs are between the 2 scenarios?

Answer 1

That's a pretty standard 1:n hierarchy that you'd create a single dimension for. Repetition of values is accepted as a worthwhile attribute of dimension tables in order to avoid joins.