Question 190 A training-provider report has a single page with two layouts, Monthly and Quarterly, saved as bookmarks. The audience must switch layouts by clicking tabs without leaving the page. Which control fits? A drillthrough action that requires a selected transaction row A Bookmark navigator associated with the saved Monthly and Quarterly bookmarks A Page navigator targeting separate report pages An ordinary slicer whose categories have no model relationship Explanation Correct Answer B:Bookmark navigators select saved report states, including visibility, on the current page without needing new physical pages. Incorrect Answers A:Drillthrough is for context-filtered navigation to detail pages, not saved-layout tabs. C:A Page navigator switches actual report pages rather than states on one page. D:A disconnected slicer alone cannot toggle saved visual visibility states.
Question 191 An invoice report uses Import mode, but the scheduled refresh now fails with HTTP 401 after the service account password was rotated. The source and schema remain valid. What should the analyst check first? Increase the number of Power BI bookmarks to reset the password Rewrite the DAX measure definitions to include a login token Update and test the semantic model’s data-source credentials or applicable gateway connection identity Change the storage mode from Import to Direct Lake without creating a Fabric source Explanation Correct Answer C:The 401 after rotation points to an expired identity in connection/gateway settings, which should be renewed and tested. Incorrect Answers A:Bookmarks store report view state, not credentials for connector authentication. B:DAX measures cannot replace source connection credentials or resolve a 401 login response. D:Switching storage modes does not authenticate the original source and may be impossible with the given architecture.
Question 192 Orders have (Region, Amount) rows (East,90), (East,120), (West,160), (East,150), (West,80). How many satisfy FILTER(Orders, Orders[Region]="East" && Orders[Amount]>=100)? 3, counting the 160 West order too 1, excluding one of the East rows above the threshold 2, the East orders of 120 and 150 5, the count of all orders before filtering Explanation Correct Answer C:Both Region=East and Amount>=100 must hold; only the two qualifying East orders pass. Incorrect Answers A:The 160 row fails the Region=East criterion even though its amount is large. B:Both 120 and 150 meet the inclusive amount test and the East condition. D:COUNTROWS evaluates the result of FILTER rather than the original unfiltered table.
Question 193 After a report measure change, a call center visual takes nine seconds, mostly in DAX query time rather than display time. The team wants to reproduce the exact query executed by that visual for targeted tuning. What should the author retrieve? Copy the DAX query from the slow visual’s Performance Analyzer recording for further analysis Export only the report page thumbnail Change the visual’s theme and assume the DAX engine no longer runs Sort the supporting date dimension alphabetically Explanation Correct Answer A:Performance Analyzer can expose/copy the visual query, enabling investigation of the slow formula and model path. Incorrect Answers B:A thumbnail lacks executable query text or timing evidence. C:Theme changes affect appearance, not the DAX query execution plan. D:Chronological presentation sorting does not identify the expensive query behind this visual.
Question 194 A telecom billing feed uses day/month/year text. During conversion the analyst encounters a value “31/02/2026”, which is not a valid date in any source locale. The downstream model must not invent a billing date. Which cleanup is justified? Replace every conversion error with 1 January 2026 without review Parse using the correct source locale, then retain and investigate the invalid rows through a separate errors workflow Rename Date to TextDate so the invalid day becomes valid Sort invalid strings to the top and include them as successful Date conversions Explanation Correct Answer B:Using the source locale addresses valid regional formats, while deliberately isolating invalid calendar values prevents silent data corruption. Incorrect Answers A:A fabricated January date changes transaction-period membership and destroys the audit trail. C:Renaming a text column has no effect on whether 31 February is a possible calendar date. D:Sorting cannot make an impossible day/date combination valid; such rows need handling rather than aggregation as real dates.
Question 195 A distributor publishes a semantic model with a WestRegion RLS role, but an authorized West viewer receives an error that the user has no role. The rule itself is correct. What should the admin check? Rename the report bookmark to WestRegion Assign the individual or appropriate security group to the published model’s WestRegion role Disable every model relationship globally Convert the Region business key to a free-text visual title Explanation Correct Answer B:Creating a role definition is not enough; the consumer’s membership must be assigned in the service. Incorrect Answers A:Bookmark names do not set RLS role membership. C:Relationship deletion would damage filtering without assigning access. D:A title cannot establish the Viewer’s role.
Question 196 A software vendor has Revenue by Country, Segment, and ProductFamily. The executive needs to choose which dimension to split next while investigating a poor total. What visual behavior fits? One KPI Card only showing the grand total Interactive Decomposition Tree splits a selected measure through explanatory dimensions A tooltip without interactive splits A static report title concatenating all field names Explanation Correct Answer B:Decomposition Tree lets readers iteratively partition the metric by selected explanatory fields. Incorrect Answers A:A Card cannot drill into contributors. C:Tooltips do not offer tree-based multistage decomposition. D:Titles contain text but do not calculate dimension-specific branches.
Question 197 A software-development reporter must create and modify workspace reports but cannot remove users, grant new roles, or change app audiences. Which least-privileged role fits their authoring needs? Contributor Admin because only Admins can create reports Viewer because its Build permission allows editing shared model definitions Member because it is required for every report modification Explanation Correct Answer A:Contributor may edit workspace content while membership and app governance remain outside their normal scope. Incorrect Answers B:Admins can manage access and the workspace, which the task forbids. C:Viewer with Build can author elsewhere but not edit shared workspace items as a Contributor. D:Member gives additional app publication and invitation authority unnecessary here.
Question 198 A monthly revenue source has one CustomerID per row and Revenue_Jan, Revenue_Feb, Revenue_Mar columns. What resulting row structure should an unpivot of the three month fields produce? CustomerID, Month, and Revenue columns with one row per customer-month value A new separate CustomerID column for each month One total Revenue field without any month identifier The original three Revenue columns plus duplicates of every customer Explanation Correct Answer A:Unpivot creates Attribute/Value pairs for months and retains the customer grain in long form. Incorrect Answers B:Separate customer IDs make the schema wider, not normalized. C:Collapsing months into total loses period identification. D:A duplicate wide form is not the desired tidy long structure.
Question 199 A model connects Store and Product to Sales facts. An analyst enables bidirectional relationships on all edges and discovers ambiguous filtering between the dimensions. What should be evaluated? Add another bidirectional relationship between every fact and dimension Make every dimension a calculated measure Remove the Sales fact to make the model circularity disappear Use single-direction dimension-to-fact relationships unless a controlled exception is justified Explanation Correct Answer D:The star-schema default limits filter directions and reduces ambiguous filter paths. Incorrect Answers A:Adding more bidirectional paths can worsen ambiguity. B:Measures cannot replace grouping dimensions. C:Removing the fact destroys the analytic dataset.
Question 200 An AI-assisted Power BI page draft shows a chart titled Freight Revenue but plots CargoTons, even though the approved revenue measure is available. What is the author’s obligation? Assume the correct chart title guarantees its numeric axis Turn off RLS to encourage Copilot to select a different value Publish without review because the model has a Certified badge Correct the selected measure and reconcile displayed revenue with approved sample calculations before publishing Explanation Correct Answer D:AI-assisted report drafting can choose semantically wrong fields and requires author validation. Incorrect Answers A:Text labels do not prove which measure is bound. B:RLS controls data access, not visual measure selection. C:Model endorsement does not validate every generated visual.
