Modeling: Pros and Cons connected scenario dimension

[ThiasW]

Hi community, 

I am struggling with a rather simple question. I want to set up a financial tabular model. As I saw it is quite common to integrate a "scenario dimension" connected to the financial fact table containing attributes like "actuals", "plan" and "forcasts". 

But somehow I have the opinion, that this dimension should not be connected to the fact table. Instead the actual, plan values and so on should be placed in separated columns to follow best practise star schema approach. I am struggling with mixing two different business processes like acutal values and planning cooperate values in one column. Especially for huge amounts of data I think the compression and performance would be much better, if I have three column (AC, PL, FC) instead of one, which forces me to filter via a connected dimension. But maybe I am wrong.

Is there any source for best practice approaches one can refer to, gathering some arguments for this topic? 

[Syndicate_Admin]

Hello. Were you able to solve your doubt in the end? I have the same question right now. Thank you!

[PaulDBrown]

The single dimension approach has the benefit of providing the correct sorting order for the final visuals

[ThiasW]

Hi thanks for your answer, 

I am not sure which option one would call "single dimension approach", connected or unconnected "scenario dimension". 

I think in both cases I can force the right sort order by sorting scenario names through a sort column.