Slightly-Off Granularity 1

[mattsturges]

I have an interesting problem with a fact table that I'm creating. In the source data, which relates to school enrollment, some sources provide separate enrollment data for male and female, and others only provide it for all students, regardless of gender. It seems bad to create an entirely separate fact table for just the gender business, but at the same time, each row would have a slightly different granularity if I kept them in the table, and slicing on gender lines wouldn't give an accurate representation of the data. The same holds true for demographic data, where some schools report enrollment broken down by ethnicity, and and some don't. What's the best way to proceed with this?

 

[blom0344]

I would create the fact with a gender key for all data and define the gender dimension as:

1. Male
2. Female
3. Unspecified

This should give you all reporting options. Abstain from using the gender dimension would yield gender independant data, but you could still add this for more detailed analysis..

Ties Blom

 

[johnherman]

I concur with Ties Blom (and give a star).

The trouble with doing something right the first time is that nobody appreciates how difficult it was - Steven Wright