Data Modeling help with relationships

[Jtischler]

Hello PBI Geniuses!

I have 4 tables:

1 - Course Completions - all course data as well as user demographic data only for users that are enrolled in a course

2 - User info - Table of all users and their demographic data

3 - Track - Created as a common demographic to filter both 1 & 2

4 - Region - Created as a common demographic to filter both 1 & 2

I have many visuals that want to take unique user counts of each table based on completion dates, subscription dates, Region, Track etc. and find % of enrolments and or completions based on Table 1 compared to user counts in Table 2

Table 1 is main data table with no column having unique value

Table 2 -> 1 is 1 to many with USER ID as unique column bidirectional filtering

Table 3  has 1 to many relationship with 1 & 2 based on Track  and 1 way filtering

Table 4 has many to many relationship with 1 and 2 using country and region with 1 way filtering

Table 1&2 have several common fields, I have connected 3 with many to many 2 way filtering

I have introduced common slicers using Table 3 and 4 in each to slice to years, or track, or region

When I do this though, the results end up wrong, I should never have a %age over 100% for users enrolled vs total count, because the  course data unique user count can never be higher than the User data Unique user count.  Equal to maybe, but never higher.

How should the relationships be set so that the slicers filter each table respectively and I get the right counts back?

I am not sure I understand the Inactive vs active relationships to know if I need to link more common fields?