Question 230 A report connected only to Import-mode data needs second-by-second visual updates from an operational SQL source. The Page refresh formatting card is unavailable. Why? Import-only reports always enable one-second page refresh regardless of connection mode Automatic page refresh normally requires a supported DirectQuery/proxy model rather than an ordinary Import-only model A Page navigator setting hides the refresh card The report has too few bookmarks to enable it Explanation Correct Answer B:Power BI exposes automatic page refresh for supported live-query scenarios, not standard Import-only refresh. Incorrect Answers A:Import mode uses scheduled model refresh rather than live requery of visuals. C:Page navigation is unrelated to connection storage mode. D:Bookmark count does not enable the feature.
Question 231 A retail returns API repeats RequestID R77 twice after a network retry. The two rows match on ID, Timestamp, RefundAmount, and Status; a third row has a different request ID. How should the ingestion validation classify them? Confirm identical payloads are retry duplicates and remove one copy without removing the independent request Delete all rows with Status=Refunded, including unique requests Generate random IDs so both retry rows count as new refunds Sum both identical R77 rows to make the refund twice as large Explanation Correct Answer A:Two identical rows carrying one request describe one event, whereas the independent request remains valid. Incorrect Answers B:Status alone is not an idempotency key. C:Random IDs hide the duplicate instead of identifying it. D:Double-counting retry events inflates refund totals.
Question 232 A finance analyst has Build permission on a shared semantic model and can create a thin report. They want to change a DAX measure inside the centrally governed model itself. Which additional authority is relevant? Model Write/edit permission through the applicable governed permission path Only a report bookmark Only a visualization theme change Only Build, which always allows replacing model measures Explanation Correct Answer A:Build supports reuse; changing model definitions requires the appropriate Write/owner rights. Incorrect Answers B:Bookmarks capture view state rather than model editing. C:Themes do not grant or change model metadata. D:Build is not the same as Write permission.
Question 233 A hospital has seven visit rows with PatientID values P11,P11,P11,P12,P13,P13,P14. The report must count distinct people seen, not total encounters. What does DISTINCTCOUNT(Fact[PatientID]) return? 7, the number of separate visit events 3, counting only patients who appear exactly once 2, the number of patient IDs that have repeated visits 4, one for each P11, P12, P13 and P14 Explanation Correct Answer D:Four distinct PatientID values exist in seven visit events; repeated encounters for P11 and P13 do not add unique people. Incorrect Answers A:Seven counts encounters, not unique patients. B:The count is not restricted to one-visit patients; P11 and P13 are still people. C:Repeated-visitor IDs form only a subset, not the full distinct-patient population.
Question 234 A governed Power Query stage standardizes customer identifiers. Two downstream output tables must always receive future changes to that normalization rule, while applying different filters afterward. Which operation should build the outputs? Create each output as a Reference of the shared cleaned staging query Duplicate the base twice and independently maintain all normalization steps Export screenshots of the staging preview for both outputs Append the stage to itself to create two filtered models Explanation Correct Answer A:Reference creates a dependent query whose downstream steps follow changes to the shared source preparation. Incorrect Answers B:Independent duplicates drift as centralized cleaning rules evolve. C:Screenshots are not reusable query dependencies. D:An append duplicates rows and does not create controlled branch outputs.
Question 235 An insurer has individual claim-payment amounts and needs the median payment for whichever claim type and region a viewer filters. The median must recalculate dynamically rather than being stored on each claim at refresh. Which model expression is appropriate? A calculated column containing the all-time portfolio median repeated for every claim A Power Query parameter representing a single fixed value chosen at refresh A report display folder renamed Median without any measure expression A reusable DAX measure using MEDIAN(Fact[ClaimPayment]) in current filter context Explanation Correct Answer D:MEDIAN evaluates the current filtered numeric column, giving a central payment amount resistant to outliers without fixing the result at refresh. Incorrect Answers A:One static value repeated on facts cannot respond to different report category selections. B:A query configuration parameter does not compute the median of interactive filtered rows. C:Folders improve navigation but contain no statistical aggregation logic.
Question 236 A facilities team compares service requests across 16 buildings with long descriptive names. A vertical column chart truncates the labels. What visual adjustment improves readability? Convert the labels into numeric timestamps Use a pie with 16 tiny slices and no legend Show one Card with only the all-building sum Use a sorted horizontal bar chart so long category labels have more room Explanation Correct Answer D:Horizontal bars provide label space and make category magnitude comparison easier. Incorrect Answers A:Timestamps cannot represent building names correctly. B:Many small pie slices are harder to compare and label. C:A Card hides per-building variation.
Question 237 After an account password and OAuth token reset, a telecom provider can still view cached monthly charges charts but the next refresh fails with an authentication error. The network path and gateway status are healthy. Which first change should be made? Remove the calendar table relationships Change Import to DirectLake without a lake source Reauthenticate the affected data-source credential in the Power BI service Disable row-level security on the model Explanation Correct Answer C:The credential needs an updated sign-in/token for the original monthly charges source to refresh successfully. Incorrect Answers A:Calendar relationships do not cause a credential error immediately after a token reset. B:DirectLake needs an appropriate OneLake source; switching modes is not a credential repair. D:RLS controls row visibility and does not resolve failed upstream authentication.
Question 238 A purchase report has three rows with (Revenue,Cost) pairs (100,60), (200,260), and (100,100). It evaluates AVERAGEX(Fact, DIVIDE(Fact[Revenue]-Fact[Cost], Fact[Revenue])). What is the result, to two decimal places? 3.33% 20.00%, by excluding the loss-making row 40.00%, by taking only the first row’s gain -30.00%, by taking only the loss-making row Explanation Correct Answer A:Per-row margin rates are +40%, -30%, and 0%. Their arithmetic mean is (40-30+0)/3 = 3.333…%. Incorrect Answers B:The negative row is valid and must not be silently removed from AVERAGEX. C:The first-row percentage alone omits two other valid rows. D:One loss rate is not the average of three outcomes.
Question 239 A fact table uses RangeStart and RangeEnd parameters but its Power Query transformation stops folding before the date predicate. Scheduled refresh rereads almost the whole source. What should be investigated? Add more report bookmarks without altering the source query Turn every datetime into untyped text Ensure the partition date filter can fold to the source and that RangeStart/RangeEnd are applied correctly Force manual CSV downloads for each partition Explanation Correct Answer C:Efficient incremental refresh relies on source-side partition filtering when supported. Incorrect Answers A:Bookmarks do not control the partition query. B:Text conversion harms reliable date partitioning. D:Manual exports do not fix the underlying folding boundary.
Question 240 A retailer uses Summary and Detail bookmarks to show or hide different visuals on one Power BI report page. Each bookmark currently restores the Region slicer value saved by the author, overriding the reader’s latest selection. The visuals must change, but selected Region must persist. What should the author change? Rebuild the Sales fact table separately for each bookmark. Grant every reader workspace Admin rights to preserve their personal slicers. Update the bookmarks to capture Display state but not Data state, and test switching with an active slicer selection. Remove relationships between Date and Sales so no slicer can affect the visuals. Explanation Correct Answer C:Display governs visibility, while Data restores filters and slicers; disabling Data preserves the reader’s current slicer choices when switching layouts. Incorrect Answers A:Duplicating fact tables does not control what report filter state a bookmark records. B:Role elevation is unrelated to bookmark data capture and violates least privilege. D:Breaking model relationships destroys analytical filtering rather than repairing bookmark behavior.
