These 25 original scenario questions assess the decision-making behind the current Microsoft Power BI Data Analyst exam. The Microsoft PL-300 exam has an English skills guide effective April 20, 2026, covering preparation, modeling, visualization/analysis and management/security of Power BI. This is independent practice material, not recalled live exam content or a pass guarantee. DA-100 is a retired predecessor referenced historically in this original article’s URL; PL-300 is the current exam.

Choose an answer before reading its rationale. The cases intentionally distinguish data quality, relationships, filter context, user interactions, asset management and authorization. Pair this question set with the current PL-300 blueprint and safe synthetic Power BI projects. A correct choice is useful only when you can also explain why the alternatives fail the requirement.

Prepare the data

Question 1: A regional retailer must analyze several years of transactional sales with com

A regional retailer must analyze several years of transactional sales with complex DAX and does not require second-by-second changes. Which storage mode is the strongest initial option?

1. DirectQuery for every source regardless of query volume
2. Use a dashboard tile as the original data store
3. Import the curated data into the semantic model
4. Create calculated tables for all raw source records

Correct answer: C. Import allows a compressed in-memory model and generally supports responsive analytical queries when refresh latency is acceptable. The business has not required real-time source querying. Validate model size, refresh windows and data governance before committing.

Why the other options do not fit: DirectQuery may be needed for data freshness or size constraints but can add source query limitations and latency; a dashboard tile is not a data store; calculated tables cannot replace a governed ingestion strategy.

Question 2: A CSV contains dates in day/month/year order, but the gateway machine uses a m

A CSV contains dates in day/month/year order, but the gateway machine uses a month/day/year locale. The resulting values are intermittently parsed incorrectly. What should the analyst change first?

1. Use Change Type with Locale in Power Query
2. Turn every date into a text label in the finished report
3. Set an arbitrary sort-by column in the visual
4. Increase the semantic model’s automatic refresh frequency

Correct answer: A. Explicitly apply the correct locale when converting the source text into a date type, and inspect invalid or ambiguous records. This makes transformation behavior reproducible across authoring and refresh environments.

Why the other options do not fit: Displaying text avoids parsing but breaks useful date semantics; sort order affects the visual, not parsing; refresh frequency repeats an incorrect transformation more often.

Question 3: A quantity column sometimes contains blanks for items not yet counted

A quantity column sometimes contains blanks for items not yet counted. Another system uses a true zero when an inventory count has been completed and no items remain. How should the analyst clean this data?

1. Replace all nulls with zero without recording meaning
2. Delete every row containing a null quantity
3. Convert the quantity column into a percentage
4. Preserve missing-value meaning and derive an explicit status if needed

Correct answer: D. A null and a confirmed zero express different facts. Preserve that distinction during data profiling, and create an explicit business-status field if analysis must separate uncounted from counted-zero records. Document any approved imputation rule.

Why the other options do not fit: Replacing null with zero manufactures completed counts; deleting all rows loses pending inventory; converting to percentages does not resolve missing-versus-zero semantics.

Question 4: Monthly exports arrive as separate tables with the same columns

Monthly exports arrive as separate tables with the same columns. The business wants one fact table covering every month. Which Power Query operation best matches the task?

1. Merge on a customer key
2. Append queries
3. Pivot a sales amount column
4. Create a calculated relationship

Correct answer: B. Append stacks rows from matching schemas into one logical table. Verify the column names, types and month identification so records remain traceable. The transformation should preserve sales rows rather than attempt to join unrelated months side by side.

Why the other options do not fit: Merge combines columns by matching keys, which would be appropriate for lookup enrichment; pivot restructures rows into columns; a model relationship does not physically consolidate the exports.

Question 5: An Excel file contains Product, January, February and March columns

An Excel file contains Product, January, February and March columns. The model needs Month and SalesAmount rows so a time dimension can filter all months consistently. Which transformation is most suitable?

1. Merge the data with a duplicate of itself
2. Unpivot the month columns
3. Replace every month value with its sum
4. Keep the wide layout and create an individual measure for every source month

Correct answer: B. Unpivot converts the month headers into attribute values and their amounts into rows, producing a more analysis-friendly shape. Preserve the product identifiers and assign correct data types to the resulting month and amount fields.

Why the other options do not fit: Self-merging adds unnecessary complexity; replacing all cells with totals discards detail; separate monthly measures are brittle as new periods arrive.

Question 6: A large relational source can perform many filters before returning data

A large relational source can perform many filters before returning data. A Power Query step unexpectedly causes the refresh to transfer millions of unfiltered rows. What should be investigated?

1. Whether the report page has a sufficient number of cards
2. Whether a bookmark contains the right visuals
3. Whether the legend title uses the correct case
4. Whether query folding stopped at that transformation

Correct answer: D. Check which transformations can fold into the source query and where folding ends, using supported diagnostics. Moving compatible filters earlier or rewriting a transformation may reduce data transfer while maintaining the required result. Verify source and gateway constraints before changing.

Why the other options do not fit: Cards, bookmarks and legend formatting affect presentation rather than the query execution path responsible for transferring the large data volume.

Model the data

Question 7: A sales model has a fact table with repeated ProductKey values and a products

A sales model has a fact table with repeated ProductKey values and a products table with one row per ProductKey. What relationship is normally preferred for product filtering?

1. One-to-many from the Products dimension to the Sales fact
2. One-to-one from Sales to Products
3. Many-to-many between the tables solely because sales has repeats
4. No relationship and duplicated product names on every visual

Correct answer: A. Use a unique product dimension as the one side of a one-to-many relationship and a many-side sales fact. This supports a star-schema design with controlled filter propagation. Check key uniqueness and handling of unknown products before activating it.

Why the other options do not fit: The fact has repeated product keys, so one-to-one is not valid; a many-to-many relationship is unnecessary here; disconnected tables complicate consistent filtering.

Question 8: A model author enables bidirectional filtering on every relationship to make o

A model author enables bidirectional filtering on every relationship to make one slicer appear to work. Several report totals now change unpredictably through ambiguous paths. What is the best next design step?

1. Add more bidirectional relationships to restore symmetry
2. Disable all dimensions and keep a single wide table
3. Use single-direction filters where possible and model the legitimate paths explicitly
4. Hide the slicer so the ambiguity cannot be seen

Correct answer: C. Prefer clear dimensional paths, usually single-direction from lookup dimensions to facts, and apply alternate behavior only when the business calculation requires it. Inspect ambiguous or many-to-many relationships rather than spreading bidirectional filters throughout the model.

Why the other options do not fit: More bidirectional paths intensify ambiguity; a single table may harm reuse and maintenance; hiding a slicer conceals rather than resolves incorrect filter propagation.

Question 9: Sales has OrderDate and ShipDate and one shared Date table

Sales has OrderDate and ShipDate and one shared Date table. The active relationship uses OrderDate, but one measure must calculate totals by ShipDate. Which DAX feature can express the alternative relation?

1. FORMAT
2. CONCATENATE
3. A report page bookmark
4. USERELATIONSHIP inside an appropriate CALCULATE measure

Correct answer: D. An inactive Date-to-ShipDate relationship can be used for a specific measure through USERELATIONSHIP within CALCULATE. Keep the active relationship for ordinary order-date measures. Review constraints when row-level security interacts with relationships and DAX evaluation.

Why the other options do not fit: FORMAT changes representation; CONCATENATE creates text; bookmarks control report view, not the relationship used by a DAX calculation.

Question 10: The report must display sales totals that change when users filter products, r

The report must display sales totals that change when users filter products, regions or dates. Should the analyst normally create a measure or a calculated column for this result?

1. A measure evaluated under filter context
2. A calculated column storing the report-wide grand total for every row
3. A Power Query index column
4. A bookmark containing the current amount

Correct answer: A. A DAX measure is calculated as needed within the visual’s filter context and responds to slicers. Calculated columns are materialized at refresh and are suited to persistent row-level attributes. Decide according to whether the result must change with user interactions.

Why the other options do not fit: A repeated grand-total column consumes memory and cannot respond in the intended way; an index identifies rows; a bookmark does not perform a filter-sensitive aggregation.

Question 11: A measure must show revenue only for completed transactions even when the visu

A measure must show revenue only for completed transactions even when the visual contains other order statuses. Which DAX pattern is appropriate?

1. Change the fact table’s display name
2. Use RELATED without an expression to sum
3. Use CALCULATE with a completed-status filter
4. Set a visual title to Completed Revenue

Correct answer: C. CALCULATE can evaluate a revenue expression under a modified filter context that enforces the intended completed status. Specify whether external status filters should be overridden or intersected with the fixed condition, and test totals against known records.

Why the other options do not fit: Renaming a table or title changes presentation only; RELATED retrieves a value through a relationship but does not independently define the required conditional aggregation.

Question 12: An analyst calculates return rate as ReturnedUnits divided by ShippedUnits

An analyst calculates return rate as ReturnedUnits divided by ShippedUnits. Some product groups have no shipments, and reports currently show division errors. Which DAX function best expresses a safe ratio?

1. LEFT over a formatted string
2. DIVIDE with a suitable alternate result
3. COUNTROWS over the calendar table
4. SELECTEDVALUE over ProductName

Correct answer: B. DIVIDE handles a zero or blank denominator using a blank or specified alternate result, making the measure safe for no-shipment groups. The team should choose how to display undefined ratios rather than automatically treat them as a real zero.

Why the other options do not fit: LEFT extracts text, COUNTROWS counts rows unrelated to the ratio, and SELECTEDVALUE retrieves a scalar selection without performing safe division.

Question 13: A fiscal analysis needs cumulative revenue from the beginning of the year to e

A fiscal analysis needs cumulative revenue from the beginning of the year to each displayed date. The semantic model already has a proper marked Date table. What should the analyst build?

1. A year-to-date DAX measure using an appropriate time-intelligence calculation
2. A tooltip that repeats the monthly revenue
3. A calculated column with one fixed total for all dates
4. A separate report page for every week

Correct answer: A. Use an appropriate year-to-date calculation, such as TOTALYTD or DATESYTD inside CALCULATE, with the Date table and fiscal-calendar assumptions correctly defined. Verify boundary dates and compare cumulative totals with source records.

Why the other options do not fit: Tooltips and extra pages do not calculate cumulative totals; a static column does not naturally respond to report filters and the selected period.

Question 14: An invoice fact table has quantity and unit price per row, and the report need

An invoice fact table has quantity and unit price per row, and the report needs the sum of each line’s quantity times price. Which DAX operation fits the grain of the data?

1. SUM of the Quantity column multiplied by SUM of UnitPrice
2. DISTINCTCOUNT of Price values
3. CONCATENATEX over the invoices
4. SUMX iterating invoices and multiplying row Quantity by UnitPrice

Correct answer: D. SUMX evaluates the expression at each fact row and sums those row-level products. This preserves line-item granularity. Multiplying independently aggregated totals can produce a different and incorrect result when quantities and prices vary across rows.

Why the other options do not fit: A product of totals is not a sum of line amounts; DISTINCTCOUNT counts unique prices; CONCATENATEX produces text instead of a monetary aggregation.

Visualize and analyze

Question 15: Executives need to see revenue trends by month over two years and identify sea

Executives need to see revenue trends by month over two years and identify seasonal changes. Which visual is usually the strongest initial choice?

1. A pie chart with every month as a separate slice
2. A line chart with a continuous or properly ordered date axis
3. A single KPI card without historical context
4. A table displaying only product identifiers

Correct answer: B. A line chart preserves the time sequence and makes trend, seasonality and inflection points visible. Verify that month labels are sorted by real date rather than alphabetically, and consider appropriate reference lines or tooltips.

Why the other options do not fit: A pie chart becomes difficult across many periods; a card removes the historical pattern; product identifiers alone do not show temporal change.

Question 16: A manager selects a customer from a summary page and needs a detailed transact

A manager selects a customer from a summary page and needs a detailed transactions page filtered to that customer without manually reselecting it. Which report feature is designed for this workflow?

1. A new calculated column on every fact row
2. Disabling all visual interactions
3. Drillthrough with a customer field and suitable navigation
4. Converting the report into a dashboard image

Correct answer: C. Configure a drillthrough target page using the customer field and provide appropriate navigation. The target carries the selected context so the manager can inspect underlying records. Verify how additional filters are retained and how the Back navigation behaves.

Why the other options do not fit: Calculated columns do not provide page navigation, disabling interactions prevents context propagation, and a static image has no filtered transaction view.

Question 17: A report’s Region slicer filters the table but should not change the national

A report’s Region slicer filters the table but should not change the national benchmark card. What should the report designer configure?

1. Disable all slicers globally
2. Place the benchmark in a calculated table with no documentation
3. Edit interactions so the slicer does not filter the benchmark card
4. Delete the card whenever a region is selected

Correct answer: C. Use Edit interactions to exclude the benchmark visual from the slicer’s filtering if that matches its intended semantics. Also verify the benchmark measure itself uses the right filter context, and label the card so users understand what it represents.

Why the other options do not fit: Removing every slicer reduces the report’s usefulness; a disconnected calculation can introduce hidden inconsistency; deleting the benchmark defeats the comparison requirement.

Question 18: Users want the same ProductCategory slicer selection to apply on several repor

Users want the same ProductCategory slicer selection to apply on several report pages. Which feature should the designer use?

1. Create a distinct unrelated category column on each page
2. Sync slicers across the chosen pages
3. Use one static image with the selected products
4. Recreate the semantic model for each page

Correct answer: B. Synchronize slicer selections between relevant pages and control whether the slicer is visible on each page. Test that users retain the selection as they navigate while pages that require independent filtering remain excluded.

Why the other options do not fit: Unrelated columns and separate models risk inconsistent semantics; a static image does not maintain an interactive filter state.

Question 19: A report needs to toggle between a summary view and a detailed layout without

A report needs to toggle between a summary view and a detailed layout without creating multiple copies of its entire semantic model. Which Power BI feature suits the presentation requirement?

1. Bookmarks combined with the Selection pane where appropriate
2. Increasing refresh concurrency
3. A new gateway for each layout
4. Changing every numeric value to text

Correct answer: A. Bookmarks can capture and restore selected report states and visibility, with the Selection pane managing which visual layers appear. Define whether each bookmark captures data, display or current-page state to avoid unexpected slicer changes.

Why the other options do not fit: Refresh concurrency and gateways affect data movement; converting numbers to text damages analysis and does not provide controlled presentation states.

Question 20: A report includes meaningful icons indicating an increase or decline, but some

A report includes meaningful icons indicating an increase or decline, but some users rely on screen readers. Which change best supports accessibility?

1. Use only red and green with no labels
2. Replace all numeric detail with an unlabeled icon
3. Disable report navigation
4. Provide meaningful text alternatives and check contrast and keyboard navigation

Correct answer: D. Give nondecorative visuals descriptive alternative text and avoid relying solely on color for meaning. Test reading order, contrast and keyboard access in the actual report. Accessibility is part of reliable communication, not merely a formatting preference.

Why the other options do not fit: Color alone may exclude users with visual differences; unlabeled icons hide semantics; disabling navigation makes the report less usable.

Question 21: A report page has slow interactive filtering

A report page has slow interactive filtering. The team needs to know whether DAX query time, visual rendering or another component is responsible. What tool should be used first?

1. A sensitivity label
2. Power BI Performance Analyzer
3. A workspace description field
4. A bookmark that stores the current view

Correct answer: B. Performance Analyzer can break down time spent loading individual visuals, including DAX query and visual display contributions. Use a reproducible interaction and analyze the most expensive component before redesigning the model or visual.

Why the other options do not fit: Sensitivity labels classify information, descriptions document assets, and bookmarks preserve presentation state; none diagnoses the observed refresh latency.

Manage and secure Power BI

Question 22: Two regional sales managers may view only their own regions in a shared publis

Two regional sales managers may view only their own regions in a shared published semantic model. Which control should enforce data restrictions?

1. Hide the Region slicer from one page
2. Place a warning in the report title
3. Publish the same unrestricted model to everyone
4. Define RLS roles with suitable model filters and assign members after publishing

Correct answer: D. Configure row-level security in the semantic model, validate filter propagation through appropriate active relationships, and map the authorized users or groups to the roles in the service. Test with View As and actual distribution permissions.

Why the other options do not fit: Hiding a slicer is not security, a title has no enforcement effect, and unrestricted publication exposes all rows. Note that higher workspace roles can have different RLS behavior from consumers.

Question 23: A user should consume a report but not edit its semantic model or workspace as

A user should consume a report but not edit its semantic model or workspace assets. Which workspace responsibility is usually appropriate?

1. Workspace Admin so all items remain visible
2. Member with full edit rights for every user
3. Viewer or app-audience access suited to the distribution design
4. Contributor with deployment permissions

Correct answer: C. Provide report consumption through app audiences or an appropriate Viewer role, subject to the organization’s permission design. Test the actual user experience and related semantic model permissions. Authoring and administration privileges should remain limited to people responsible for maintaining content.

Why the other options do not fit: Admin, Member and Contributor provide broader authoring or management capabilities than simple report consumption and can affect how data restrictions are applied.

Question 24: A published Import semantic model reads an on-premises SQL Server database

A published Import semantic model reads an on-premises SQL Server database. Refresh succeeds on the analyst’s desktop but fails in the Power BI service. What should be checked?

1. An appropriate on-premises data gateway, data source mapping and supported credentials
2. The line chart color palette
3. The report bookmark selection
4. The visual’s subtitle capitalization

Correct answer: A. The service needs a supported gateway path and valid data-source credentials for the on-premises system, along with scheduled refresh configuration. Verify which dataset and gateway mapping the service uses and inspect the refresh error rather than assuming a local desktop connection transfers automatically.

Why the other options do not fit: Visual styling and bookmark state do not establish the network or credential path needed for scheduled data refresh.

Question 25: A trusted finance report should be recognized as reviewed by its governing ana

A trusted finance report should be recognized as reviewed by its governing analytics team, and also contain confidential fields requiring classification. Which governance interpretation is best?

1. Certification automatically grants all employees permission to view the report
2. A sensitivity label guarantees that model calculations are factually correct
3. Endorsement and sensitivity labels address different purposes and may both be appropriate
4. Endorsement replaces row-level security

Correct answer: C. Promotion or certification helps users identify governed or trusted content according to the organization’s process. A sensitivity label classifies handling requirements for the information. Neither inherently grants viewing rights or replaces data-model authorization and correctness testing.

Why the other options do not fit: Endorsement does not open restricted resources; sensitivity labels do not validate formulas; and row-level security must remain enforced independently where required.

How to use the answer key

For every missed item, identify the underlying misconception rather than simply recording an answer letter. If the error involves data typing or transformation, inspect the Power Query steps. If it concerns relationships or DAX, verify the current filter context and relevant table grain. If it concerns RLS or governance, test both an intended consumer and a prohibited access path. The DAX reference and Power BI data modeling guidance provide useful background when a single calculation or relationship error reappears in several scenarios.

Change the case after reviewing the rationale: use a different source, an alternate role, another relationship or a changed business requirement. A learner who can defend the revised case has stronger evidence of readiness than one who remembers the same 25 option letters.