Question 580 A maintenance system emits a separate row whenever TaskID T8 changes from Open to Assigned to Resolved. Operations needs time-to-resolution and transition history. Which source design is appropriate? Retain the state-change events with TaskID and event time/version; deduplicate only true duplicate deliveries Remove duplicates on TaskID and keep one arbitrary state Replace all event times with a single timestamp Append each state change as a new task with an unrelated randomly generated TaskID Explanation Correct Answer A:Repeated task IDs across events reflect history, not automatically duplicate records. Incorrect Answers B:Dropping transitions prevents duration and status-sequence calculations. C:Constant times destroy the transition chronology. D:Random task IDs break linkage among one task’s states.
Question 581 A regional salesperson has Viewer workspace access and additional Build permission on a certified sales model. They connect from Excel to analyze sales and expect records outside their RLS region. What is the effective security result? They remain restricted to rows allowed by their RLS role even when querying with Build Build automatically upgrades the user to Contributor and removes RLS Excel connections are always exempt from Power BI semantic-model RLS Certification causes all model rows to be visible to any Builder Explanation Correct Answer A:Microsoft documents that Viewers with Build remain subject to RLS, including when they use Analyze in Excel. Incorrect Answers B:Build grants model reuse capabilities, not a workspace edit-role upgrade. C:RLS still applies to eligible semantic-model queries through client applications. D:Endorsement describes content quality and does not alter row-level permissions.
Question 582 A school enrollment fact contains PersonID values P1, P1, P2, and BLANK(). How many distinct nonblank people does DISTINCTCOUNTNOBLANK return? Two people: P1 and P2 Three by including BLANK as a third person Four by counting the duplicate P1 twice One by counting only the first ID Explanation Correct Answer A:The function ignores the blank member and deduplicates repeated P1. Incorrect Answers B:Three is the distinct count including a blank. C:Four is the row count, not unique people. D:P2 is also a valid distinct nonblank person.
Question 583 A factory group has a finished production records transformation to reuse as a frozen starting point for an experiment. The experiment requirements have changed: the temporary report now must pick up every future correction to the production query automatically. Which construction is best? Create a reference to the maintained production query Keep an independent duplicate with no dependencies Manually copy rows into an Excel file each week Transpose the original to avoid dependencies Explanation Correct Answer A:A reference consumes updated production results while retaining separate downstream steps. Incorrect Answers B:A duplicate does not automatically inherit future production-step edits. C:Manual copying is not a query dependency and risks data drift. D:Transpose changes shape, not maintenance relationship.
Question 584 A financial analyst creates a data alert on a dashboard KPI showing cash exposure. The department assumes the alert makes the underlying confidential model inaccessible to other users. What is accurate? Data alerts automatically replace every semantic-model security role A numeric alert threshold encrypts unauthorized rows Anyone receiving an alert automatically becomes a workspace Member Alerts notify on configured value conditions; item permissions and RLS still control data access Explanation Correct Answer D:Notification logic and data entitlement are distinct Power BI mechanisms. Incorrect Answers A:Alerts do not create or remove model RLS definitions. B:A threshold does not encrypt or limit source records. C:Receiving an email does not grant workspace authoring roles.
Question 585 A sports league has OrderDate and ShipDate in its match tickets fact and one Date dimension. Analysts now need independent date slicers for order month and ship month on the same page, without special measures. Which model alteration makes both paths active? Disable the ship-date relationship permanently Create separate Order Date and Ship Date role dimensions with active fact relationships Try to activate two conflicting paths from one Date dimension Use one date slicer for both independent roles Explanation Correct Answer B:Two date-role dimensions allow both independent slicers to filter the fact through active paths. Incorrect Answers A:Disabled ShipDate filtering defeats the second role. C:Competing paths from one Date table cannot both be active in the simple model. D:One slicer cannot independently select two different date-role ranges.
Question 586 An energy utility needs to view Consumption by country and city in a geographic report. The location column contains internal store codes such as CA and IN rather than verified geographic coordinates. Automatic mapping puts codes in unintended countries. What should be corrected first? Plot internal codes as numeric sales totals Provide unambiguous location fields or coordinates with suitable geographic categories Keep ambiguous two-letter codes and assume correct geocoding Convert every store to the same latitude and longitude Explanation Correct Answer B:Explicit coordinates or complete place metadata prevent erroneous geographic resolution. Incorrect Answers A:Store codes are identifiers, not monetary measures. C:Ambiguous codes may resolve to places outside the intended region. D:A single coordinate loses true store locations.
Question 587 A manufacturer groups plants by energy use, output and downtime using the Power BI scatter-plot clustering capability. A stakeholder wants to treat each cluster as a regulatory compliance classification. What is the responsible interpretation? Treat clusters as exploratory similarity groups and validate them against actual compliance rules before use Assume automatically formed clusters are legally authoritative compliance labels Delete all plants outside the largest cluster as incorrect records Change the cluster color to green to certify regulatory status Explanation Correct Answer A:Clustering groups observations using selected measures, not external legal criteria; compliance requires independently defined and tested rules. Incorrect Answers B:Unsupervised groups have no automatic legal interpretation. C:Outliers can represent genuine data needing investigation. D:Formatting a visual does not confer a governed classification or compliance status.
Question 588 An analyst adds an expensive nonfolding M function ahead of a selective Date filter on a DirectQuery-compatible source. Interactive reports stall. What should they evaluate? Reorder or redesign transformations so supported selective filters fold to the source before expensive local processing Disable all SQL-side filters Add a global page bookmark for each date Duplicate the expensive calculation across every visual Explanation Correct Answer A:Source-side filtering reduces the records and work sent into later evaluation when folding is supported. Incorrect Answers B:Removing filters increases source rows. C:Bookmarks do not change query folding. D:Duplicating calculations increases cost without restoring folding.
Question 589 An analyst has transaction-level Sales and Cost columns and needs the mean gross-profit *amount per transaction*, calculated for each row before averaging. Which DAX expression correctly computes it? AVERAGE(Fact[Sales] - Fact[Cost]) using an arithmetic expression as AVERAGE’s only column argument AVERAGEX(Fact, Fact[Sales] - Fact[Cost]) SUM(Fact[Sales]) - SUM(Fact[Cost]) without dividing by transaction count COUNTROWS(Fact) without using either monetary column Explanation Correct Answer B:AVERAGEX iterates the table, evaluates Sales-Cost for each transaction, then averages those per-row numeric amounts. Incorrect Answers A:AVERAGE expects a column reference rather than an arbitrary row expression. C:Total profit answers aggregate profit, not average profit amount per row. D:A row count does not calculate any gross-profit amounts.
Question 590 A warehouse wants to shorten a 100-column imported fact table. Six columns are currently hidden in the Fields pane but used in relationship definitions and security expressions. Which action is safe? Delete every hidden column automatically because hidden means unused Review all model dependencies before physically removing columns; hidden fields may still be essential Replace the hidden relationship keys with a visual bookmark Increase source rows to compensate for removed attributes Explanation Correct Answer B:Hidden is only a visibility setting; it does not prove that model columns are unused. Incorrect Answers A:Deleting essential hidden keys can break model joins or RLS. C:Bookmarks cannot substitute for semantic-model fields. D:Adding more rows increases memory and does not repair missing relationships.
