Why doesn't my filter work properly in a dimensional model but it works fine as one big table?

[tbennett93]

I have this table (the data has been mocked up for demonstration but is repeatable in PBI):

Name    Score                   Question ID

John	10	1
Jane	20	1
John	30	2
Jane	40	2
Jim	50	2

If I want to show Jane's score for each question, I add the measure:

Jane's Score = CALCULATE(
                        SUM('Table'[Score]),
                    FILTER(
                        ALLEXCEPT(
                            'Table',
                            'Table'[Question ID]
                        ),
                    'Table'[Name] = "Jane"
                    )
)

This returns, as expected, Jane's score for each question:

Now if I transform this into a dimensional model, where my Name becomes a dimension. I have:

So if I want to display a table with everybody's score for each question and then a column for Jane's score too, I add the measure:

Janes Score v2 = CALCULATE(
sum(Merge1[Score]),
FILTER(
ALLEXCEPT(
Merge1,
Merge1[Question ID]
)
,RELATED(Names[Name])="Jane")
)

If I put this into a table with the Score and click 'Don't Summarize' on the score in the fields for the visual, my table turns into this:

Why are so many redundant rows being returned and how do I stop it?

It is related to 'don't summarize', but I don't know why this option causes this to happen and i'd like to understand why this is happening.

Thanks

[lbendlin]

You are applying the filter to the "wrong"  table (the fact table).  You are supposed to filter by dimension columns.

Look at the arrow in your data model.  The filter on the fact cannot travel back up to the dimension.

ALLEXCEPT is a very blunt instrument. It messes with your filter context - remove ALL filters EXCEPT that one.  Try to use it sparingly,  Use RemoveFilters, or AllSelected etc.