Question 440 A report author fixes a chart title in the workspace. The app’s consumers still see the old title even though the semantic model refreshed successfully. What is the missing publication step? Run an unrelated SQL schema migration Grant a consumer database admin Have the authorized app publisher update the app to distribute the report change Change the model’s calendar format only Explanation Correct Answer C:Model refresh updates data, but distributing changed app report content requires an app update. Incorrect Answers A:Schema migration is irrelevant to a report title change. B:Database admin cannot update app content. D:Date formatting does not republish changed report visual titles.
Question 441 The invoice ID field contains I7, I7, I8, and BLANK() after filtering to one branch. Which statement about DISTINCTCOUNT is correct? It returns 4 because every row is unique by occurrence It returns 2 because it always discards blanks It returns 3 by counting I7, I8 and one BLANK value It returns 1 because only one branch is filtered Explanation Correct Answer C:The measure counts distinct identifier values inside the filter context, including one blank. Incorrect Answers A:Four counts the duplicated I7 twice. B:Blank exclusion needs an explicit different expression/filter. D:The number of branches does not equal the number of invoice values.
Question 442 A retail report contains a Region column with actual nulls and valid label `Unknown` meaning the organization has genuinely classified a store that way. Analysts are tempted to replace every null with `Unknown`. The model must distinguish missing source data from a deliberate Unknown classification. What should they do? Convert the Region column to Decimal Number to distinguish null from text Keep or explicitly flag null as missing, separately from the valid `Unknown` category, and document the two meanings Replace every null with `Unknown` and report the combined count as fully classified Replace every `Unknown` with null and discard the legitimate business classification Explanation Correct Answer B:Keeping missingness distinct from a legitimate category preserves the business meaning of both records. Incorrect Answers A:The values are categorical; a numeric type would not encode a distinction between missing and known label. C:Collapsing null and a valid category misstates data quality and classified membership. D:Replacing the legitimate label with null erases a real business classification and inflates missingness.
Question 443 An order export contains ProductKey values 1,2,99; the approved Product table contains only 1 and 2. The analyst must isolate invalid fact rows before applying a relationship. Which join orientation finds the orphan? Left anti join with OrderLines on the left and Product on the right Left anti join with Product on the left and OrderLines on the right Inner join from OrderLines to Product Full outer join with no filtering of matched records Explanation Correct Answer A:With facts on the left, left anti produces order rows whose ProductKey has no dimension match, including 99. Incorrect Answers B:Reversing the sides finds unused product members, not invalid fact keys. C:Inner join discards the unmatched fact record being investigated. D:Full outer contains unmatched rows but also matched rows, so it does not directly isolate the target set.
Question 444 A risk table marks overdue payments only in red and current payments in green. Some readers cannot distinguish the hues. What should the report author provide? More saturated red and green while keeping the status color-only Remove every row label and rely only on the background Replace status definitions with an arbitrary tooltip title A clear status text or icon in addition to suitably contrasted color cues Explanation Correct Answer D:Redundant labels or icons make the condition understandable without relying on color discrimination. Incorrect Answers A:Greater saturation still relies entirely on the same inaccessible channel. B:Removing labels reduces context rather than improving it. C:A tooltip title alone is not a robust visible status encoding.
Question 445 An analyst added an RLS role to a loan dataset. A colleague with only Viewer access reports seeing extra branches. Before distributing the app, the author must confirm which branch rows the role actually filters in the report. What should the author do? Add Build permission and assume it automatically corrects a faulty RLS filter Use View as (or test role as an appropriate identity) to examine effective visible rows under the role Grant the user workspace Contributor, which bypasses RLS and proves the viewer experience Hide unauthorized branch names in the Fields pane rather than restricting rows Explanation Correct Answer B:Testing the role evaluates the effective filter behavior; an inappropriate role expression or membership becomes visible during validation. Incorrect Answers A:Build does not repair filter expressions and does not exempt Viewer users from RLS. C:Contributor bypass behavior would no longer model the intended Viewer security boundary. D:Hiding columns from field lists does not prevent unauthorized fact rows from being queried.
Question 446 A private clinic group loads appointments and Clinics keys where the dimension has accidentally duplicated keys after a merge. Multiple blank or null members are being represented as separate dimension rows and the key is no longer unique. What should the analyst inspect? Switch table colors in the report Add a distinct transaction ID to every duplicate dimension row Normalize missing business keys and determine a consistent unknown-member mapping Enable many-to-many automatically Explanation Correct Answer C:Cleaning missing dimension keys and using an appropriate unknown mapping restores one-side uniqueness. Incorrect Answers A:Visual formatting does not change duplicate keys. B:Fact IDs are not replacement dimension member identifiers. D:Many-to-many conceals an inconsistent dimension representation.
Question 447 A factory monitors monthly throughput forecasts. The data team has measures [Forecast], [LowEstimate], and [HighEstimate] for every month; executives need to see the central trajectory and its uncertainty band. What should the report designer do? Hide the interval measures and show only a single total Card Use an unrelated pie chart grouping months by name without any bound values Use an appropriate time-series line visual and configure supported error bars or confidence bounds from lower and upper measures Convert throughput to text to prevent the upper and lower series from being aggregated Explanation Correct Answer C:A time-series line with explicitly supplied range measures visualizes both predicted movement and the uncertainty around each period. Incorrect Answers A:A single value cannot communicate an uncertainty range across the monthly timeline. B:Pie slices discard the time axis and do not convey lower/upper uncertainty per month. D:Changing numeric measures to text would prevent valid quantitative charting.
Question 448 A dataset’s YearToLoad parameter is currently text, and a later query compares it numerically to a Year column. Refresh fails for nonnumeric values. Which parameter improvement is appropriate? Declare an appropriate Whole Number parameter and validate allowed year values Change the Year column to a report image Replace every numeric comparison with a bookmark Store an arbitrary string and rely on implicit conversion everywhere Explanation Correct Answer A:A typed year parameter prevents invalid text from silently entering numeric source filters. Incorrect Answers B:An image cannot act as a date filter. C:Bookmarks do not type-check source extraction parameters. D:Implicit coercion leaves avoidable refresh errors and inconsistent comparisons.
Question 449 A retail model has two stores with [MarginPct] measures of 10% and 90%. One store has far more orders than the other, but a KPI explicitly asks for the simple mean of the two store-level percentages. The DAX expression is AVERAGEX(VALUES(Store[StoreID]), [MarginPct]). What is returned? 90%, because the maximum store rate becomes the mean A transaction-weighted margin that depends on each store's order mix 100%, by summing both store rates without dividing 50% Explanation Correct Answer D:VALUES creates one row per store and AVERAGEX averages the two evaluated store measures equally: (10%+90%)/2=50%. Incorrect Answers A:The 90% higher store does not replace the 10% store in the two-row iterator. B:Order-count weighting is a different aggregation from AVERAGEX over the distinct store table. C:Adding two percentages without dividing by two is not an arithmetic mean.
Question 450 A sales team adds a monthly-by-region aggregate table, but users frequently slice on individual product SKU. What determines whether visuals can safely use that aggregate instead of the detail fact? A monthly region total always answers any SKU-level question exactly Removing Product from the model will automatically create correct SKU results The report theme selects the fact table needed for each filter The aggregation must contain or correctly relate the dimensions needed by the query; SKU-level filters may require detail access Explanation Correct Answer D:An aggregation missing SKU detail cannot independently resolve product-level filtering without a valid detail path. Incorrect Answers A:An overall region total collapses distinctions among products. B:Removing a dimension worsens the mismatch instead of recovering lost grain. C:Themes affect styling, not source-grain query accuracy.
