Calculate, count and use relationships giving incorrect results

[Bls-80]

Hi, hope someone can help point out where and why this doesn't work.  I have the following measure 

Count Findings = CALCULATE(

    COUNT(Findings[Finding ID]),

    USERELATIONSHIP('Date'[Date], Findings[Date created]),

    (

        Findings[Category of finding] = "Major NC" ||

        Findings[Category of finding] = "Minor NC" ||

        Findings[Category of finding] = "Opportunity for Improvement"

    )

)

I have a date table and relationship to the date created. I have multiple tables with multiple dates in each tables so multiple relationships across the whole model so need to use a date table. 

When I pull this into a pivot table with date[month] and use a slicer on date [year] I get completely wrong results.
results:

results I should get

My table filtered from 1st jan 2024 showing that the results are wrong (I have no January or Feb entries!).

I have checked all dates are dates and not date time, I have made all relationships from date table to other tables inactive, tried countx, counta, filter, creating separate measures or var for the or conditions. I just cannot find a solution to give me the correct result. I'm assuming I'm missing something in the filter context but I can't figure it out.

any Help and explanation appreciated.

[Bls-80 · Accepted solution]

Thanks, in posting it sent me to other posts where I solved it. The date had been changed on the table but not in transform data where it was still a date time, also needed to add keep filters. 

[Greg_Deckler]

@Bls-80 Sorry, having trouble following, can you post sample data as text and expected output?
Not really enough information to go on, please first check if your issue is a common issue listed here: https://community.powerbi.com/t5/Community-Blog/Before-You-Post-Read-This/ba-p/1116882

Also, please see this post regarding How to Get Your Question Answered Quickly: https://community.powerbi.com/t5/Community-Blog/How-to-Get-Your-Question-Answered-Quickly/ba-p/38490

The most important parts are:
1. Sample data as text, use the table tool in the editing bar
2. Expected output from sample data
3. Explanation in words of how to get from 1. to 2.