Exploring Differences in Power BI Slicer Behavior: Composite Model vs. Simple Model

Sample Data Setup

Refer to the attached screenshot below. 

Notice that the D13 data element is not actually part of the dimension.

Dimension-Fact Relationship

Simple Model: See the attached diagram below.

Composite Model: Also attached below. 

The key difference is that the fact table originates from another model.

Slicer Behavior Variations

Case 1:Refer to the screenshot below. 

In the simple model, dimension elements not present in the Fact table are still considered, whereas in the composite model, only dimension elements present are shown.

Case 2:Refer to the screenshot below. 

In the simple model, selecting D11 and D12 effectively filters out these elements. Conversely, all entries in the fact table are displayed in the composite model, even when only D11 and D12 are selected. (Please note that if you only select D11 and D12 but leave Select all not selected, the D13 will also be filtered out for the composite model. ) 

Model Consideration

The composite model offers some advantages, such as the ability to add new slicers without impacting the results, as long as the "Select All" option is used. This essentially ignores slicers not presented. However, to align the slicer behavior of the simple model with that of the composite model, new elements must be added to the dimension, or data integrity must be enforced to provide users with a precise data depiction.