Build a useful investigation brief
Build a propagation fixture
Create rows that differ by the filter field and include a card measure with a known total. Record expected populations and values for no selection, one selection, an excluded selection, and a control that should not apply. Preserve page state and report revision.
Follow control to value
Capture filter values, serialized state, predicate or model input, included identifiers, aggregation, formatted result, and neighboring visual output. Compare the card with an independently scoped calculation. Mark where the selected value is dropped, broadened, or attached to the wrong measure.
Test one mapping change
Change only the filter relationship, exclusion rule, or card measure and rerun the fixture. Compare card, chart, table, and export when available. Treat an intentional card scope as a policy hypothesis until the report owner explains why the control should or should not affect it.
What to carry forward
The investigation is ready when the brief names the propagation fixture, first filter-to-card divergence, affected measure, neighboring evidence, and tested scope alternative. Recommend a scoped mapping or explanation change. Keep the result open when card independence is undocumented.
Technical background: Metabase documentation.
Keep the decision with the work.
Use a Work Item in Aglet to record the problem, the evidence you have, and the next decision. Add an owner and priority, then keep updates in the discussion so the next person can follow the reasoning.
Create an account See the product workflow