Question 460 Several departments build thin reports on one approved financial semantic model. The governance team corrects its GrossMargin measure. What is the operational advantage of this architecture? Every dependent report must import independent fact copies for each new measure One corrected model definition can flow to authorized dependent reports, reducing inconsistent local versions Central model changes automatically bypass consumer RLS roles Thin report visuals cannot be edited after initial creation Explanation Correct Answer B:Reuse centralizes calculation definitions and improves cross-report consistency. Incorrect Answers A:Independent copies recreate the very drift the organization is avoiding. C:Shared measures do not remove row entitlements. D:Report authors can change visuals while retaining the shared model connection.
Question 461 A customer-retention CSV has CustomerID, Segment, Jan, Feb, and Mar. New month columns arrive every quarter, and the analyst needs CustomerID, Segment, Month, Revenue. Which operation is most robust to new month columns? Transpose the entire table, making identifiers into rows Pivot Jan, Feb, and Mar to introduce one more column per month Select CustomerID and Segment and choose Unpivot Other Columns Select CustomerID and unpivot only it, turning account IDs into measure values Explanation Correct Answer C:Unpivot Other Columns preserves the selected identifier fields and reshapes current and future month columns into month/revenue pairs. Incorrect Answers A:Transpose swaps dimensions and destroys the intended transactional shape. B:Pivot would retain or add separate month fields, the opposite of a normalized long table. D:Unpivoting CustomerID discards the stable entity identifier from the output grain.
Question 462 A marketing analyst can edit a thin report but must not change its centrally managed semantic model. Which governance design supports this? Grant appropriate report authoring rights while keeping model edit/ownership restricted and providing only required Build access Give the analyst model and workspace Admin permissions everywhere Copy the semantic model as an uncontrolled new source Hide the original model name and assume all users become editors Explanation Correct Answer A:Build enables authorized reuse without requiring full source-model management rights. Incorrect Answers B:Admin grants excess privileges. C:Copies reintroduce governance drift. D:Hiding an artifact name does not grant controlled authoring access.
Question 463 An operation report uses a one-second page refresh configuration, but its SQL query consistently takes four seconds to complete. What behavior should the analyst anticipate? Power BI fires unlimited overlapping copies each second regardless of completion Effective refresh cadence is limited by query completion; tune source queries and choose a realistic interval Switching colors makes the SQL query instantaneous Report bookmarks queue the missed queries for later delivery Explanation Correct Answer B:Automatic refresh waits for active queries; slow SQL results in updates slower than configured. Incorrect Answers A:Repeated overlapping queries are not the documented automatic page refresh behavior. C:Visual styling cannot reduce source execution time by itself. D:Bookmarks are not a query job scheduler.
Question 464 A fact column contains IDs K1, K1, K2, BLANK(), BLANK(). The analyst needs the number of distinct populated IDs without treating missing as an entity. Which DAX function/result is appropriate? DISTINCTCOUNT(Fact[ID]) returns 2 by automatically ignoring blanks DISTINCTCOUNTNOBLANK(Fact[ID]) returns 2 COUNTROWS(Fact) returns 2 COUNT(Fact[ID]) returns the distinct entity count Explanation Correct Answer B:K1 and K2 are the only distinct nonblank identifiers; DISTINCTCOUNTNOBLANK ignores the repeated blank. Incorrect Answers A:DISTINCTCOUNT includes BLANK and would return 3 for these entries. C:COUNTROWS counts five rows, not distinct identifiers. D:COUNT considers populated entries, not unique IDs, and does not provide this distinctness.
Question 465 A Customer table has two legitimate historical versions per CustomerID, each with EffectiveFrom and EffectiveTo. A fact table must link each sale to the version effective when it occurred. Which data-model step preserves correct history? Resolve each fact row to the effective dimension version and store its unique surrogate version key Use CustomerID as a one-side unique key even though it repeats across history Delete every previous customer record and lose historic attributes Set both cross-filter directions without resolving which version each sale belongs to Explanation Correct Answer A:A type-2 dimension uses unique version surrogate keys; date-aware matching attaches each sale to the intended historic member. Incorrect Answers B:Repeated business IDs are not unique in the versioned dimension. C:Deleting history would misstate past sales attributes. D:Relationship direction cannot choose a historically correct version.
Question 466 A retailer wants to group stores on profitability and footfall. A user draws one horizontal line and calls the points above it clusters. Which statement is technically accurate? A one-axis threshold does not perform two-variable clustering; use a suitable grouping analysis of both coordinates Any constant line is mathematically equivalent to multivariate clustering Only a title change is needed for clusters to emerge A source password controls which clusters are created Explanation Correct Answer A:A fixed threshold partitions by one feature, whereas clustering considers multidimensional similarity. Incorrect Answers B:A line cannot account for proximity across both x and y positions. C:Renaming a visual provides no clustering computation. D:Authentication governs data access, not analytic cluster assignments.
Question 467 A clinic has only a static threshold line at $200 on a monthly cost chart. The CFO requests a probabilistic forecast for next quarter with uncertainty bands. What should the analyst explain? Treat the threshold as a complete prediction for every future month A static target does not estimate future costs; assess a suitable historical time series and configure Forecast with a horizon and interval if eligible Hide actual historic data and call the remaining line a forecast Use a month-name pie chart to generate confidence intervals Explanation Correct Answer B:Forecasting needs time-series observations, not a reference threshold value. Incorrect Answers A:Targets and model predictions are different concepts. C:Removing observations discards evidence rather than generating future estimates. D:A pie has no supported time-series forecast analytics.
Question 468 A charitable donation import has 35 values: 30 valid, 4 nulls, and 1 actual conversion error. Which statement is accurate about Power Query Column quality? All four nulls must be treated as errors and added to the single failed conversion A null value always means the numeric donation was zero A field cannot have valid values and errors at the same time Empty and Error are separate categories, so the nulls do not count as type-conversion errors Explanation Correct Answer D:Column quality reports empty values and error values separately; four nulls and one error describe distinct defects. Incorrect Answers A:Missing cells do not inherently cause conversion errors. B:Null means absent or unknown, not a measured zero. C:A source column can contain a mix of valid, empty and invalid records.
Question 469 A partner organization receives a Power BI app invitation. The guest cannot open it although the email is correct. Which factors should be reviewed? Make the guest source database owner Rename the report title to External Allowed Convert the app’s numeric measures to strings Cross-tenant B2B configuration, app/model grants, and applicable licensing/capacity requirements Explanation Correct Answer D:External consumption depends on identity trust, tenant policies, entitlement and item access. Incorrect Answers A:Source database ownership is not required for ordinary report readers. B:A display name does not authorize external identities. C:Data type conversion cannot resolve B2B access.
Question 470 A cost file has Department, Month, and Amount rows. Several Department–Month pairs appear more than once because they represent separate valid expenses. The analyst must produce one column per month and sum all legitimate amounts. What is required when pivoting? Pivot and choose Do Not Aggregate even though there are repeated Department–Month pairs Remove duplicate Department–Month rows, losing valid expenses Unpivot Department identifiers into numeric Amount cells Pivot the Month column with Sum as the aggregation of Amount Explanation Correct Answer D:Sum makes each Department–Month output cell the total of its valid source rows; pivoting without aggregation cannot resolve duplicate intersections. Incorrect Answers A:Do Not Aggregate requires a unique pair and fails or loses information for repeated combinations. B:Deduplication would delete genuine expense transactions. C:Unpivot changes the orientation of columns and does not create monthly totals.
