Exam Bundle

Changing Product selection changes only the matching sales amount. That check would be meaningful for the separate case: A report shows the same total for every Product because the Products table is disconnected from the sales fact. It would not directly verify this scenario's selected implementation: Set measure format and model data categories appropriate to business currency.

A year-month view includes dates with no fact rows and time calculations use a continuous calendar. That check would be meaningful for the separate case: A sales report shows totals by day but has no complete calendar table and dates disappear on days with no sales. It would not directly verify this scenario's selected implementation: Set measure format and model data categories appropriate to business currency.

A product slicer filters the correct matching sales rows without duplicate-dimension-key errors. That check would be meaningful for the separate case: A sales fact has repeated ProductKey values while a product dimension contains one row per ProductKey. It would not directly verify this scenario's selected implementation: Set measure format and model data categories appropriate to business currency.

Using 'Resolve duplicate RegionCode values or introduce a truly unique dimension key' supports the result 'The relationship's one-side keys are unique and refresh does not fail on duplicates' for A geography lookup contains multiple rows for each RegionCode, preventing its use as the one-side of a relationship. It does not implement 'Create separate one-to-many relationships from a conformed Salesperson dimension to both facts' or prove 'Selecting one salesperson filters the sales and commission measures consistently' for this question.

'Build a unique Store lookup or use an appropriate location surrogate key before creating the relationship' is validated when 'Refresh can validate the unique one-side and stores filter their intended transactions' for A store dimension has duplicate StoreID values because source address rows were not normalized. It does not demonstrate 'Selecting one salesperson filters the sales and commission measures consistently', which requires 'Create separate one-to-many relationships from a conformed Salesperson dimension to both facts'.

A successful 'Keep each fact at its native documented grain rather than treating target rows as daily transactions' would show 'Monthly target comparisons aggregate correctly without multiplying one target by the number of transaction days' for A finance model imports monthly targets at Store-Month grain and daily transactions at Store-Day grain. This case needs 'Create separate one-to-many relationships from a conformed Salesperson dimension to both facts' and evidence that 'Selecting one salesperson filters the sales and commission measures consistently', not the other outcome.

Monthly target comparisons aggregate correctly without multiplying one target by the number of transaction days. That check would be meaningful for the separate case: A finance model imports monthly targets at Store-Month grain and daily transactions at Store-Day grain. It would not directly verify this scenario's selected implementation: Create separate one-to-many relationships from a conformed Salesperson dimension to both facts.

The relationship's one-side keys are unique and refresh does not fail on duplicates. That check would be meaningful for the separate case: A geography lookup contains multiple rows for each RegionCode, preventing its use as the one-side of a relationship. It would not directly verify this scenario's selected implementation: Create separate one-to-many relationships from a conformed Salesperson dimension to both facts.

Refresh can validate the unique one-side and stores filter their intended transactions. That check would be meaningful for the separate case: A store dimension has duplicate StoreID values because source address rows were not normalized. It would not directly verify this scenario's selected implementation: Create separate one-to-many relationships from a conformed Salesperson dimension to both facts.

The check for 'Introduce a customer dimension with unique keys and connect each fact using controlled one-to-many relationships' is 'Customer selections filter both facts through the conformed dimension without unintended fanout' for Two fact tables both repeat CustomerID and a developer proposes a direct many-to-many join for every filter. That does not establish 'The primary measure uses the default path while specialized measures use the alternative key deliberately'; the stated requirement calls for 'Keep a primary active relationship and use a specific inactive relationship in targeted DAX when appropriate'.

'Extend calendar boundaries to cover fact dates and necessary comparison periods' addresses the mechanism measured by 'Time-intelligence calculations have calendar dates for the required prior periods' for A generated Date table starts after the earliest recorded transaction date and misses prior-year comparisons. The current outcome is 'The primary measure uses the default path while specialized measures use the alternative key deliberately' after 'Keep a primary active relationship and use a specific inactive relationship in targeted DAX when appropriate', which the distractor leaves unverified.

'Visuals display SKU values without automatic sum or average aggregation' would demonstrate 'Set the column's data type and default summarization to suit an identifier' for A dimension has a column of SKU labels for grouping but report authors should not aggregate it numerically, but not 'The primary measure uses the default path while specialized measures use the alternative key deliberately'. The latter requires 'Keep a primary active relationship and use a specific inactive relationship in targeted DAX when appropriate' in this scenario.

Visuals display SKU values without automatic sum or average aggregation. That check would be meaningful for the separate case: A dimension has a column of SKU labels for grouping but report authors should not aggregate it numerically. It would not directly verify this scenario's selected implementation: Keep a primary active relationship and use a specific inactive relationship in targeted DAX when appropriate.

Customer selections filter both facts through the conformed dimension without unintended fanout. That check would be meaningful for the separate case: Two fact tables both repeat CustomerID and a developer proposes a direct many-to-many join for every filter. It would not directly verify this scenario's selected implementation: Keep a primary active relationship and use a specific inactive relationship in targeted DAX when appropriate.

Time-intelligence calculations have calendar dates for the required prior periods. That check would be meaningful for the separate case: A generated Date table starts after the earliest recorded transaction date and misses prior-year comparisons. It would not directly verify this scenario's selected implementation: Keep a primary active relationship and use a specific inactive relationship in targeted DAX when appropriate.

Using 'Create role-playing date dimensions for Invoice and Ship analysis' supports the result 'InvoiceDate and ShipDate slicers independently narrow the intended date roles in one visual' for Sales has InvoiceDate and ShipDate but the business requires separate simultaneous slicers for both roles. It does not implement 'Create and validate the appropriate ProductKey relationship between dimension and fact' or prove 'Changing Product selection changes only the matching sales amount' for this question.

'Remove or constrain redundant paths and use intentional single-direction star relationships' is validated when 'Filter propagation is unambiguous and regional sales reconcile to source totals' for The model contains two potentially ambiguous bidirectional paths from Region to Sales through different tables. It does not demonstrate 'Changing Product selection changes only the matching sales amount', which requires 'Create and validate the appropriate ProductKey relationship between dimension and fact'.

A successful 'Hide relationship-key fields while keeping them available to the model relationships' would show 'Report creators see friendly descriptors and relationships still propagate filters' for A model exposes technical surrogate keys in the report field list, confusing self-service analysts. This case needs 'Create and validate the appropriate ProductKey relationship between dimension and fact' and evidence that 'Changing Product selection changes only the matching sales amount', not the other outcome.

Filter propagation is unambiguous and regional sales reconcile to source totals. That check would be meaningful for the separate case: The model contains two potentially ambiguous bidirectional paths from Region to Sales through different tables. It would not directly verify this scenario's selected implementation: Create and validate the appropriate ProductKey relationship between dimension and fact.

InvoiceDate and ShipDate slicers independently narrow the intended date roles in one visual. That check would be meaningful for the separate case: Sales has InvoiceDate and ShipDate but the business requires separate simultaneous slicers for both roles. It would not directly verify this scenario's selected implementation: Create and validate the appropriate ProductKey relationship between dimension and fact.

Report creators see friendly descriptors and relationships still propagate filters. That check would be meaningful for the separate case: A model exposes technical surrogate keys in the report field list, confusing self-service analysts. It would not directly verify this scenario's selected implementation: Create and validate the appropriate ProductKey relationship between dimension and fact.

The check for 'Keep the inactive ShipDate relationship and activate it within the relevant DAX measure using USERELATIONSHIP' is 'The shipping measure follows ShipDate while ordinary sales measures continue following OrderDate' for An analyst only needs an occasional shipping-date measure while an active relationship already uses OrderDate. That does not establish 'Refresh can validate the unique one-side and stores filter their intended transactions'; the stated requirement calls for 'Build a unique Store lookup or use an appropriate location surrogate key before creating the relationship'.

'Create a one-to-many relationship from unique Product to repeating Sales keys' addresses the mechanism measured by 'A product slicer filters the correct matching sales rows without duplicate-dimension-key errors' for A sales fact has repeated ProductKey values while a product dimension contains one row per ProductKey. The current outcome is 'Refresh can validate the unique one-side and stores filter their intended transactions' after 'Build a unique Store lookup or use an appropriate location surrogate key before creating the relationship', which the distractor leaves unverified.

'Year-level headcount reflects the selected endpoint rather than the sum of twelve monthly snapshots' would demonstrate 'Model the snapshot grain explicitly and calculate an appropriate period-end measure', but not 'Refresh can validate the unique one-side and stores filter their intended transactions'. The latter requires 'Build a unique Store lookup or use an appropriate location surrogate key before creating the relationship' in this scenario.

Year-level headcount reflects the selected endpoint rather than the sum of twelve monthly snapshots. That check would be meaningful for the separate case: A quarterly headcount metric is stored as one row per department and month, and monthly figures should not be summed across time. It would not directly verify this scenario's selected implementation: Build a unique Store lookup or use an appropriate location surrogate key before creating the relationship.

The shipping measure follows ShipDate while ordinary sales measures continue following OrderDate. That check would be meaningful for the separate case: An analyst only needs an occasional shipping-date measure while an active relationship already uses OrderDate. It would not directly verify this scenario's selected implementation: Build a unique Store lookup or use an appropriate location surrogate key before creating the relationship.

A product slicer filters the correct matching sales rows without duplicate-dimension-key errors. That check would be meaningful for the separate case: A sales fact has repeated ProductKey values while a product dimension contains one row per ProductKey. It would not directly verify this scenario's selected implementation: Build a unique Store lookup or use an appropriate location surrogate key before creating the relationship.

Using 'Create a continuous common Date table covering the full required period' supports the result 'A year-month view includes dates with no fact rows and time calculations use a continuous calendar' for A sales report shows totals by day but has no complete calendar table and dates disappear on days with no sales. It does not implement 'Extend calendar boundaries to cover fact dates and necessary comparison periods' or prove 'Time-intelligence calculations have calendar dates for the required prior periods' for this question.

'Use single-direction dimension-to-fact relationships unless a supported special case requires otherwise' is validated when 'Customer and date slicers filter deterministically without ambiguous propagation'. It does not demonstrate 'Time-intelligence calculations have calendar dates for the required prior periods', which requires 'Extend calendar boundaries to cover fact dates and necessary comparison periods'.

A successful 'Resolve duplicate RegionCode values or introduce a truly unique dimension key' would show 'The relationship's one-side keys are unique and refresh does not fail on duplicates' for A geography lookup contains multiple rows for each RegionCode, preventing its use as the one-side of a relationship. This case needs 'Extend calendar boundaries to cover fact dates and necessary comparison periods' and evidence that 'Time-intelligence calculations have calendar dates for the required prior periods', not the other outcome.

Customer and date slicers filter deterministically without ambiguous propagation. That check would be meaningful for the separate case: A table joins customer orders to two date dimensions and all relationships are configured Both, creating ambiguous filtering. It would not directly verify this scenario's selected implementation: Extend calendar boundaries to cover fact dates and necessary comparison periods.

A year-month view includes dates with no fact rows and time calculations use a continuous calendar. That check would be meaningful for the separate case: A sales report shows totals by day but has no complete calendar table and dates disappear on days with no sales. It would not directly verify this scenario's selected implementation: Extend calendar boundaries to cover fact dates and necessary comparison periods.

The relationship's one-side keys are unique and refresh does not fail on duplicates. That check would be meaningful for the separate case: A geography lookup contains multiple rows for each RegionCode, preventing its use as the one-side of a relationship. It would not directly verify this scenario's selected implementation: Extend calendar boundaries to cover fact dates and necessary comparison periods.