Question 530 A national bookstore has an Orders-like book sales query with ReceiptID, a Shops key, and NetSales. The analyst is required to construct a quality-control subset containing only transactions with a matching approved lookup entry; missing keys must be excluded. Which merge join kind does this? Full outer join retaining even unmatched records Append the two differently shaped queries Inner join on the approved lookup key Left anti join of unmatched transaction keys Explanation Correct Answer C:Inner join returns only key matches on both sides, the approved subset requested. Incorrect Answers A:Full outer includes unapproved and orphaned lookup rows as well as matches. B:Appending is a vertical union, not the required key-validation join. D:Left anti returns the opposite set: rows with no matching lookup.
Question 531 A supplier dimension contains SupplierKey and SupplierName. Self-service authors keep placing SupplierKey (an integer surrogate) in their charts, while the key must remain present for filtering fact rows. What should the semantic-model designer do? Remove SupplierKey from the model and disconnect the fact relationship Convert SupplierKey into a date to discourage use Expose multiple copies of the key next to SupplierName Hide SupplierKey from the reporting field list while retaining its relationship Explanation Correct Answer D:Hidden technical fields still participate in relationships while authors see the descriptive attribute intended for reports. Incorrect Answers A:Deleting the key breaks a working lookup path. B:A date conversion introduces an invalid data meaning and likely mismatched joins. C:More surrogate-key fields increase confusion without fixing report usability.
Question 532 Copilot creates a quarterly sales chart labeled “Net Sales” but binds the ShippingCost measure to the y-axis. An analyst has approved NetSales values for the same period. What must happen before publishing? Add the ShippingCost field to the tooltip but leave the incorrect value mapping in place. Assume the semantic model’s endorsement guarantees every generated chart uses the intended field. Publish as-is because a correct chart title proves the measure binding is correct. Inspect and correct the field binding, then reconcile the rendered results with approved NetSales values. Explanation Correct Answer D:The analyst must validate Copilot suggestions against intended model semantics and known results before publishing. Incorrect Answers A:A tooltip addition cannot fix the misleading underlying numeric series. B:Model endorsement does not certify each field choice made by an AI-assisted chart draft. C:The heading is display text and does not validate the y-axis measure or calculations.
Question 533 A risk analyst needs to inspect a selected account’s history, scroll a large transaction table, and change subsequent filters. The design team proposes a hover tooltip page instead of drillthrough. Why is that unsuitable? Report tooltips are not interactive scrolling pages; use a drillthrough destination for the continuing investigation Hover tooltips always allow users to modify slicers and scroll freely A tooltip changes the account’s RLS role automatically A tooltip is a source connector with a refresh schedule Explanation Correct Answer A:A drillthrough page offers persistent navigation and supported interaction; a hover tooltip cannot provide the required exploration. Incorrect Answers B:Microsoft limits tooltip interactivity and scrolling. C:Tooltip content cannot elevate a viewer’s data access. D:Tooltips are visual UI elements, not source connectors.
Question 534 An enterprise model uses a RegionEast row-security role. A published report appears correctly filtered to the author, who is workspace Admin. What is the proper validation before customer rollout? Test as role and validate the same results with an eligible Viewer, checking row and total visibility Assume the Admin’s unrestricted view proves the security role works Hide the Region field and conclude rows are protected Apply a sensitivity label and skip user-context tests Explanation Correct Answer A:RLS applies to Viewer consumers, not workspace Admins. Testing must validate effective role membership and actual consumer results. Incorrect Answers B:An Admin bypasses RLS and therefore cannot validate row restrictions through their normal access. C:Hiding a field is presentation, not row filtering. D:A label does not test or implement the RLS predicate.
Question 535 A report selects July 10–20 on a complete classic Date column. The analyst uses PARALLELPERIOD(Date[Date], -1, MONTH) for a monthly benchmark. What dates form the result? June 10–20 only July 1–31 because it does not shift All of June, rather than only June 10–20 The prior calendar year Explanation Correct Answer C:PARALLELPERIOD returns the full shifted month even when the current selection covers part of July. Incorrect Answers A:DATEADD is the closer partial-period shift behavior; PARALLELPERIOD completes the month. B:The function shifts the chosen month one interval backward. D:The interval argument specifies MONTH, not YEAR.
Question 536 A retail analyst finds one branch with exceptionally high revenue and low customer count. The chart correctly identifies it as an outlier. What should happen before removing the row from the model? Delete the branch record immediately because extreme coordinates must be invalid Investigate the branch data, grain, and valid business conditions; an outlier alone is not a data error Switch the axis to a random category and declare the variance solved Use only the global average to hide the branch Explanation Correct Answer B:An unusual observation can be genuine; exclusion needs a documented data-quality or business justification. Incorrect Answers A:Automated deletion risks suppressing a real high-value branch. C:Relabeling does not validate the observation. D:An average conceals the potential problem instead of evaluating it.
Question 537 An enterprise data source is accessible only inside a secured Azure virtual network, and teams do not want to maintain an on-premises gateway host. Which gateway option should they evaluate? A personal gateway installed on every employee phone A report bookmark configured as a private endpoint Anonymous public export of the protected database Microsoft-managed VNet data gateway subject to supported connector and tenant requirements Explanation Correct Answer D:The VNet data gateway supplies managed connectivity into supported secured virtual-network sources. Incorrect Answers A:Personal laptop/phone gateways do not provide an enterprise managed VNet service. B:Bookmarks are not connectivity components. C:Publishing restricted data publicly defeats its network isolation.
Question 538 A hotel imports January and February extracts with matching column names BookingID, StayDate, and RoomRevenue. January RoomRevenue is Decimal, while February RoomRevenue arrives as text "129.50". What should be done before appending and aggregating the combined months? Append both without validating types and depend on text concatenation to act as a total Use a merge on BookingID to replace February transactions with January records Rename February RoomRevenue to a different unrelated text field and exclude it from the totals Apply an explicit Decimal conversion to February RoomRevenue, validate exceptions, then append the compatible tables Explanation Correct Answer D:Consistent types across the same RoomRevenue column permit reliable append and later sums; invalid text values should be surfaced rather than silently converted to zero. Incorrect Answers A:Mixed text/numeric storage can prevent arithmetic and obscure import errors. B:Merge joins matching records; it does not vertically combine month-to-month transactions. C:Excluding February revenue leaves legitimate transactions out of the combined KPI.
Question 539 A product dimension has nine rows with blank ProductID, each from a different raw feed. A relationship expects the dimension key to be unique. Transactions with no valid product still need to be reported as Unknown. What is a defensible data-model repair? Normalize unmatched products to a single governed Unknown member with one surrogate key Keep all nine blank ProductID dimension rows on the one side of a one-to-many relationship Create nine duplicate Unknown entries and enable both-direction filtering Delete all transactions that have no mapped product without checking completeness Explanation Correct Answer A:A single unknown member preserves referential mapping and one-side uniqueness while allowing unmatched transactions to be counted. Incorrect Answers B:A one-side relationship cannot use multiple indistinguishable blank keys as a unique dimension key. C:Bidirectional filtering does not make duplicate dimension keys unique. D:Dropping unmatched transactions changes business totals and masks data quality problems.
Question 540 A cashflow table contains profit and loss measures ranging from -5000 to +6000. The author wants a diverging gradient with zero as neutral. What is important when configuring conditional formatting? Choose a diverging color scale anchored at the meaningful zero midpoint, using the signed numeric measure Use absolute values so losses and gains both appear positive Assign the maximum positive color to all negative values Convert the numeric measure into a date serial before formatting Explanation Correct Answer A:A zero-centered gradient preserves direction and magnitude of losses and gains. Incorrect Answers B:Absolute value destroys the sign and makes losses indistinguishable from profits. C:Using positive colors for negatives misrepresents business meaning. D:Date serial values do not express signed cashflow.
