Select Two (Multi-Select) Questions

Enable encryption at rest for the dataset.

Why wrong: Encryption at rest protects data from unauthorized physical access but does not control user permissions.

Use Power Query to mask PII columns by replacing values with '***'.

Why wrong: Masking in Power Query is static and applies to all users; it does not provide granular access control.

Apply sensitivity labels to the dataset.

Why wrong: Sensitivity labels classify data but do not enforce access restrictions at the row level.

VALUES

Why wrong: VALUES returns distinct values but does not filter.

SUM

Why wrong: SUM is an aggregation, not a filter function.

ALL

Why wrong: ALL removes filters, it does not apply them.

Merge tables as early as possible in the query to combine data.

Why wrong: Merging early may cause large data volumes to be processed; delay merging if possible.

Keep all columns in the table to avoid missing data.

Why wrong: Removing unnecessary columns reduces memory and processing time.

Split columns by delimiter to normalize data.

Why wrong: Splitting columns increases the number of columns and processing steps, potentially reducing performance.

Split a large fact table into multiple smaller fact tables.

Why wrong: This can complicate the model and may not improve performance.

Use DirectQuery mode instead of Import mode to reduce data load time.

Why wrong: DirectQuery does not load data, but it is not always a performance best practice.

RANKX

Why wrong: RANKX is a ranking function.

CONCATENATEX

Why wrong: CONCATENATEX concatenates strings.

MINX

Why wrong: MINX is an iterator function.

Show external users in lists of suggested people.

Why wrong: 'Show external users in lists of suggested people' controls UI display, not app access.

Allow external users to edit and manage content in the organization.

Why wrong: 'Allow external users to edit and manage content' grants editing permissions; not required for viewing.

Excel workbook

Why wrong: Excel only supports Import mode.

SharePoint Online List

Why wrong: SharePoint Online List only supports Import mode.

JSON file

Why wrong: JSON files only support Import mode.

It is required for any relationship involving a date column

Why wrong: Not required for relationships.

The date column can contain duplicate dates

Why wrong: Must be unique.

It automatically creates all date hierarchy columns (Year, Quarter, Month, Day)

Why wrong: It does not automatically create columns; you must add them.

Configure Microsoft Defender for Cloud Apps to forward logs to Microsoft Sentinel.

Why wrong: Defender for Cloud Apps can ingest logs, but it is not the source of audit logs.

Use the 'Export' feature in Power BI to export audit logs to a CSV file.

Why wrong: Power BI does not have an export feature for audit logs.

Set up a diagnostic setting in Azure Monitor to collect Power BI logs.

Why wrong: Azure Monitor does not collect Power BI audit logs directly.

Disable 'Create app workspaces' for non-admins.

Why wrong: Prevents workspace creation, not sharing.

Disable 'Export to Excel' for all reports.

Why wrong: Controls export, not sharing.

Remove rows that contain null values using the 'Remove Rows' > 'Remove Blank Rows' option.

Why wrong: Remove Blank Rows is an incorrect method for handling rows with null values because it only removes rows where all columns are null or empty. If a row has a null in one column but contains data in others, it is considered a partial row and is retained. To remove rows with any null values, you would need to filter each relevant column individually or use a custom row-level condition, not simply Remove Blank Rows.

Merge the table with another table that has no nulls.

Why wrong: Merging the table with another table that has no nulls does not handle existing null values. A merge operation combines columns based on keys, but the original nulls in the primary table remain unchanged unless the new columns are explicitly used to replace or coalesce them. Without adding a conditional column or expanding and applying logic, merging alone does not clean nulls, and it can introduce additional duplicates or nulls if key matches are incomplete.

Change the data type of the column to a non-nullable type.

Why wrong: Changing a column's data type to a non-nullable type is not a valid null-handling method because Power Query data types do not enforce nullability. Nulls remain valid values for any data type, and forcing a non-nullable type in the data model would cause import errors rather than transforming the data. This approach fails to address existing nulls and simply shifts the problem to load-time validation.

CALCULATE(SUM(Sales[Amount]), ALL(Sales))

Why wrong: INVALID: CALCULATE returns a scalar value, not a table, so it cannot be used to define a calculated table.

Use CALENDARAUTO in a measure

Why wrong: Using CALENDARAUTO in a measure is incorrect because the requirement is for a calculated table, not a measure.

Use the GROUPBY function

Why wrong: Using GROUPBY is possible but not listed as a step in the options; it could be used instead of SUMMARIZECOLUMNS, but the question asks which steps to take, and GROUPBY is not the most direct for this scenario.

Create a date table using CALENDAR or CALENDARAUTO

Why wrong: Creating a date table using CALENDAR or CALENDARAUTO is not required because you can directly group by the month of ShipDate from the Orders table. While a date table can be useful for other analyses, it is not a necessary step for this specific calculated table.

Use the VALUES function to get unique months

Why wrong: Using the VALUES function to get unique months is not sufficient because it only returns distinct month values without aggregation.

Publish to public web (Publish to web)

Why wrong: Makes report public, not secure.

Send them an email subscription with the report attached

Why wrong: Subscriptions send snapshots, not interactive reports.

Embed the report in a SharePoint Online page and share the link

Why wrong: SharePoint requires authentication, external users may not have access.

Power BI Premium capacity assigned to the workspace

Why wrong: Labels are supported in Pro and Premium workspaces.

Power BI Pro license for all users

Why wrong: Labels work with any license if enabled by admin.

Export the report to PDF and email it to the external user.

Why wrong: Exporting is not a sharing method within Power BI service.

Send the external user a direct link to the report via email; they can view it without any additional setup.

Why wrong: External users need to be authenticated as guests or view via Publish to web.

Embed the report in a SharePoint Online page using the Power BI web part.

Why wrong: SharePoint embedding requires the user to be authenticated within the organization.

Add animations to visuals to draw attention.

Why wrong: Animations such as fly-ins or spinning visuals are decorative motion effects that do not aid comprehension. They can trigger vestibular or photosensitive issues, and screen readers ignore motion, so the information is lost. Under accessibility best practices, static and predictable visuals are preferred because animations introduce visual noise without providing any equivalent text or structural benefit.

Use a color-blind friendly palette.

Why wrong: Although color-blind friendly palettes can help some users, they are not considered a direct method to enhance accessibility in the context of this exam; the three correct methods are tab order, high contrast themes, and alt text.

Split columns by delimiter to create more columns.

Why wrong: Splitting columns increases the number of columns, which can increase load time and memory usage, and does not directly improve data quality or reduce load time.

Sort the data by a key column.

Why wrong: Sorting data does not reduce load time; it may even increase processing time. It also does not improve data quality.

Promote headers to use the first row as column names.

Why wrong: Promoting headers improves data quality by correctly labeling columns, but it does not reduce load time as it does not change the volume of data loaded.

Merge queries to combine data from multiple sources.

Why wrong: Merging queries combines data from multiple sources, which can increase complexity and load time, and does not inherently reduce load time.

Define row-level security (RLS) roles.

Why wrong: Row-level security (RLS) defines which rows different users can see, typically by creating roles in the Model view using 'Manage Roles' and writing DAX filter expressions for tables. Query Editor is a data-shaping environment and has no concept of users or security contexts; it cannot be used to set permissions or roles. Configuring RLS generally requires the data model to be completed first.

Create a relationship between two tables.

Why wrong: While you can merge queries in Query Editor to combine data from two tables on a common column, this does not create a true relationship in Power BI's data model. A relationship is a metadata link between distinct tables, defined in the Model view by connecting foreign keys, not by physically reshaping data. Merging effectively denormalizes tables into a single table, which is a different operation from establishing a relationship.

Create a new measure using DAX.

Why wrong: Measures are DAX formulas that are stored in the data model and evaluated dynamically based on report filters and slicers. The Creation of measures is done in the Report view or Model view using the formula bar, or in calculation groups—not in Power Query. Query Editor uses the Power Query M macro language for data transformation, and it has no mechanism to define DAX measures.

Configure incremental refresh to limit data retrieved per query.

Why wrong: Incremental refresh is designed for Import mode, where data is cached in Power BI and partitioned by date or other ranges so only new or changed data is refreshed periodically. DirectQuery does not use an in-memory cache; every visual interaction triggers a live query against the data source, so there is no refresh cycle to schedule or partition. Configuring incremental refresh on a DirectQuery model either is unsupported in the traditional sense or requires a hybrid setup, and it would not limit the data retrieved per query because each visual still submits a full, on-demand SQL query to the source.

Hide columns that are not needed in reports.

Why wrong: Hiding a column in Power BI only prevents it from appearing in the Fields list for report authors; the column remains fully part of the semantic model and its storage and query surface area are unchanged. The DirectQuery engine will still include hidden columns in generated SQL queries when they are referenced by relationships, measures, or other internal metadata structures, so the same volume of data is transferred from SQL Server. Unlike physically removing the column, hiding does nothing to reduce the cost of query generation or data retrieval, making it a governance feature, not a performance optimization.

Add calculated columns to precompute aggregations.

Why wrong: Calculated columns are evaluated during each DirectQuery request because they are not persisted in the database; the DAX expression is translated into query logic that SQL Server must compute on the fly, often over huge tables. This adds significant complexity to the generated SQL, frequently producing correlated subqueries or forcing the engine to materialize entire tables just to compute the new column. Precomputing aggregations requires storing pre-aggregated data, which calculated columns cannot do in DirectQuery—so they increase query overhead rather than reduce it.