pbPassingBI
/
10 questions

Data modelling & relationships — quiz

10 questions covering this module. Star schema, relationships, calendar tables and role-playing dimensions.

  1. Why is a dedicated Date table required?

    1. To make reports look tidier
    2. Time intelligence functions depend on it
    3. It reduces file size
    4. DirectQuery needs it

    Answer: Time intelligence functions depend on it — Functions like SAMEPERIODLASTYEAR require a contiguous, marked date table to work correctly.

  2. You inherit a model with Product → Subcategory → Category. Best fix?

    1. Leave it — it is normalised
    2. Flatten into one Product dimension in Power Query
    3. Convert to DirectQuery
    4. Add bidirectional filters

    Answer: Flatten into one Product dimension in Power Query — Flattening the snowflake removes hops that slow filter propagation and complicate DAX.

  3. Sales has OrderDate and ShipDate, both joinable to Date. How do you measure by ShipDate?

    1. Duplicate the Date table
    2. Use USERELATIONSHIP inside CALCULATE
    3. Set cross-filter to Both
    4. Use a many-to-many relationship

    Answer: Use USERELATIONSHIP inside CALCULATE — USERELATIONSHIP activates the inactive ShipDate relationship for that one measure.

  4. Why avoid setting cross-filter direction to Both globally?

    1. It is slower to refresh
    2. It creates ambiguous filter paths and possible circular dependencies
    3. It breaks DirectQuery
    4. It disables slicers

    Answer: It creates ambiguous filter paths and possible circular dependencies — Ambiguity means the engine may resolve paths unexpectedly; CROSSFILTER in a measure is safer.

  5. You need line-level revenue (Qty × Price) totalled in a visual. Best approach?

    1. A calculated column then SUM
    2. SUMX(Sales, Sales[Qty] * Sales[Price])
    3. A Power Query merge
    4. A calculated table

    Answer: SUMX(Sales, Sales[Qty] * Sales[Price]) — SUMX creates row context on the fly without storing an extra column in the model.

  6. Which is stored in the model and increases file size?

    1. A measure
    2. A calculated column
    3. A visual
    4. A slicer

    Answer: A calculated column — Calculated columns are materialised at refresh and persist in memory.

  7. USERELATIONSHIP must be used inside:

    1. SUMX
    2. CALCULATE
    3. FILTER
    4. A calculated column

    Answer: CALCULATE — It is a filter modifier.

  8. Why can only one relationship between two tables be active?

    1. Performance
    2. Power BI could not determine which to filter through
    3. A licensing limit
    4. DAX cannot handle more

    Answer: Power BI could not determine which to filter through — Others exist as inactive and are activated on demand.

  9. Why is a date column on the fact table insufficient?

    1. It is the wrong type
    2. Dates with no transactions are missing, breaking the timeline
    3. It cannot be related
    4. It is too large

    Answer: Dates with no transactions are missing, breaking the timeline — Time intelligence needs a contiguous set of dates.

  10. Month names appear as Apr, Aug, Dec. Fix:

    1. Change the format
    2. Sort by column, using a month-number column
    3. Use a hierarchy
    4. Mark as date table

    Answer: Sort by column, using a month-number column — They are text, so they sort alphabetically by default.