Category: Analytics

Analytics, Business Intelligence, Microsoft Certification, PL-300, Power BI January 17, 2026

Identify and Connect to Data Sources or a Shared Semantic Model (PL-300 Exam Prep)

This post is a part of the PL-300: Microsoft Power BI Data Analyst Exam Prep Hub; and this topic falls under these sections: 
Prepare the data (25–30%) 
    --> Get or connect to data 
        --> Identify and connect to data sources or a shared semantic model

Note that there are 10 practice questions (with answers and explanations) for each section to help you solidify your knowledge of the material. Also, there are 2 practice tests with 60 questions each available on the hub below the exam topics section.

One of the first and most foundational skills tested in the PL-300: Microsoft Power BI Data Analyst exam is the ability to identify, select, and connect to appropriate data sources. This skill lives within the Prepare the data domain and underpins everything that follows—modeling, visualization, and analysis.

A Power BI Data Analyst must understand where data lives, how to connect to it efficiently, and when to reuse existing models instead of building new ones.

Understanding Data Source Types in Power BI

Power BI supports a wide variety of data sources, and the exam expects familiarity with common enterprise and self-service scenarios rather than obscure connectors.

Common Data Source Categories

File-based sources

Excel workbooks
CSV and text files
XML and JSON files
PDF files (structured tables)

Database sources

SQL Server
Azure SQL Database
Azure Synapse Analytics
Oracle
MySQL / PostgreSQL

Cloud and SaaS sources

SharePoint Online lists and files
OneDrive
Dataverse
Azure Blob Storage
Azure Data Lake Storage Gen2

Power Platform and Power BI sources

Power BI datasets (shared semantic models)
Dataflows
Datamarts

Exam tip: You are not expected to memorize every connector—focus on recognizing the correct source for a given scenario.

Choosing the Right Connectivity Mode

When connecting to data, Power BI offers three primary connectivity modes, and the exam frequently tests your understanding of when to use each one.

Import Mode

Data is loaded into the Power BI model
Fast performance
Supports full DAX and modeling capabilities
Requires dataset refreshes

Use when:
Data size is manageable and performance is critical.

DirectQuery Mode

Queries data directly at the source
Minimal data stored in Power BI
Performance depends on the source system
Some modeling and DAX limitations

Use when:
Data is very large or must remain in the source system.

Live Connection

Used with shared semantic models (Power BI datasets) or Analysis Services
No local model created
Modeling is done in the source dataset

Use when:
Connecting to a centrally governed dataset managed by another team.

Connecting to a Shared Semantic Model

A shared semantic model (formerly called a “dataset”) allows analysts to reuse existing data models instead of duplicating logic.

What Is a Shared Semantic Model?

A centrally managed Power BI dataset
Contains curated tables, relationships, measures, and security
Published to the Power BI Service
Designed for reuse across multiple reports

Benefits of Using Shared Semantic Models

Consistent metrics and definitions
Reduced duplication of logic
Centralized governance and security
Faster report development

How to Connect

In Power BI Desktop:

Select Get data
Choose Power BI datasets
Select an existing dataset from the workspace
Create a report using a Live connection

Exam tip:
When connected to a shared semantic model:

You cannot modify relationships
You cannot add calculated tables
You can create report-level measures (thin reports)

Authentication and Access Considerations

The exam also expects awareness of authentication methods, especially in enterprise environments.

Common authentication types include:

Windows authentication
Database authentication
Microsoft Entra ID (Azure AD)
OAuth

Key concept:
Your ability to connect depends on permissions at the source, not just in Power BI.

Identifying the Best Data Source for a Scenario

You may see exam questions that describe a business requirement and ask you to identify the most appropriate data source or connection method.

Example Scenarios

A finance team needs fast, interactive reports → Import mode
A large transactional database updated every minute → DirectQuery
Multiple teams need consistent KPIs → Shared semantic model
Files stored in SharePoint and updated weekly → SharePoint Folder connector

Key Exam Takeaways

For the PL-300 exam, remember the following:

Understand major data source categories
Know Import vs DirectQuery vs Live connection
Recognize when to use shared semantic models
Be aware of limitations when using Live connections
Choose data sources based on performance, governance, and scale

This topic is less about clicking buttons and more about making the right architectural decision—a critical skill for a Power BI Data Analyst.

Power BI Connection Modes – Comparison Table (PL-300 Exam Focus)

Feature / Aspect	Import	DirectQuery	Live Connection
Where data is stored	Loaded into Power BI dataset	Remains in source system	Remains in source dataset or Analysis Services
Performance	Fastest (in-memory)	Depends on source performance	Depends on source dataset
Data refresh required	Yes (scheduled or manual)	No (queries run live)	No (always live)
Data volume support	Limited by dataset size	Very large datasets	Very large datasets
Modeling allowed	Full modeling supported	Limited modeling	No model changes allowed
Create relationships	Yes	Limited	No
Create calculated tables	Yes	Limited	No
Create calculated columns	Yes	Limited	No
Create measures (DAX)	Yes	Yes (with limitations)	Yes (report-level only)
Power Query transformations	Fully supported	Limited (query folding dependent)	Not supported
Row-level security (RLS)	Supported	Supported	Inherited from source
Typical sources	Excel, CSV, SQL, SharePoint	Large SQL, Azure SQL, Synapse	Power BI datasets, Analysis Services
Best use case	High performance, small-to-medium data	Near-real-time or massive data	Reusing governed enterprise models
Exam frequency	Very high	High	Very high

When to Choose Each Mode (Exam Scenarios)

Import Mode

Choose Import when:

Performance is the top priority
Data size is manageable
Full modeling flexibility is needed

📌 Most common and safest answer on the exam unless constraints are stated.

DirectQuery

Choose DirectQuery when:

Data is extremely large
Data must stay at the source
Near real-time reporting is required

📌 Expect questions about performance trade-offs and modeling limitations.

Live Connection

Choose Live Connection when:

Using a shared semantic model
Consistent metrics are required across reports
Centralized governance is in place

📌 Often paired with “thin report” exam scenarios.

High-Value Exam Tips ⭐

If the question mentions enterprise governance or shared datasets → think Live Connection
If it mentions large, frequently changing data → think DirectQuery
If it mentions fast visuals and full modeling → think Import
Live connections do not allow model changes
Import mode requires refresh
DirectQuery relies on query folding

Practice Exam Questions

Go to the Practice Exam Questions for this section.

Analytics, Microsoft Certification, PL-300, Power BI January 17, 2026

Practice Questions: Identify and connect to data sources or a shared semantic model (PL-300 Exam Prep)

This post is a part of the PL-300: Microsoft Power BI Data Analyst Exam Prep Hub; and this topic falls under these sections: 
Prepare the data (25–30%) 
    --> Get or connect to data 
        --> Identify and connect to data sources or a shared semantic model

Below are 10 practice questions (with answers and explanations) for this topic of the exam. 
There are also 2 practice tests for the PL-300 exam with 60 questions each (with answers) available on the hub.

Practice Questions

Question 1

A report requires the fastest possible query performance and the ability to create calculated tables and columns. The dataset size is moderate and updates occur nightly. Which connection mode should you use?

A. DirectQuery
B. Live connection
C. Import
D. Composite model

✅ Correct Answer: C

Explanation:
Import mode stores data in memory, provides the best performance, and supports full modeling capabilities.

Question 2

A company has a centrally governed Power BI dataset with standardized KPIs used across multiple departments. You need to build a report using this dataset without modifying the model. What should you use?

A. Import mode
B. DirectQuery mode
C. Live connection
D. Power BI dataflow

✅ Correct Answer: C

Explanation:
Live connections are used to connect to shared semantic models and enforce centralized governance.

Question 3

Which limitation exists when using a Live connection to a Power BI dataset?

A. Measures cannot be created
B. Relationships cannot be modified
C. Data refresh cannot be scheduled
D. Visual interactions are disabled

✅ Correct Answer: B

Explanation:
When using a Live connection, the data model is read-only and cannot be changed.

Question 4

A dataset must query a transactional database containing billions of rows with near real-time requirements. The data must remain at the source system. Which connection mode is most appropriate?

A. Import
B. Live connection
C. DirectQuery
D. Power BI dataflow

✅ Correct Answer: C

Explanation:
DirectQuery queries the data directly at the source and supports very large datasets.

Question 5

You connect to a shared semantic model in Power BI Service. Which type of measures can you create?

A. Calculated column measures
B. Dataset-level measures
C. Report-level measures only
D. No measures can be created

✅ Correct Answer: C

Explanation:
With a Live connection, only report-level measures are allowed.

Question 6

Which data source scenario is best suited for Import mode?

A. Streaming IoT telemetry
B. Large fact table updated every second
C. Excel files stored in SharePoint and refreshed weekly
D. Enterprise semantic model shared across teams

✅ Correct Answer: C

Explanation:
File-based sources with scheduled refreshes are ideal for Import mode.

Question 7

Which statement about DirectQuery is true?

A. All Power Query transformations are supported
B. Data is cached in memory for performance
C. Query performance depends on the source system
D. Calculated tables are fully supported

✅ Correct Answer: C

Explanation:
DirectQuery sends queries to the source system, so performance depends on the source.

Question 8

When should you connect to a Power BI dataset instead of importing data?

A. When you need to modify relationships
B. When consistent metrics across reports are required
C. When data volume is small
D. When Power Query transformations are required

✅ Correct Answer: B

Explanation:
Shared semantic models ensure consistency and governance across reports.

Question 9

Which authentication method is commonly used when connecting to cloud-based Power BI datasets?

A. Database authentication
B. Windows authentication
C. OAuth / Microsoft Entra ID
D. Anonymous authentication

✅ Correct Answer: C

Explanation:
Power BI Service connections typically use Microsoft Entra ID (OAuth).

Question 10

A Power BI report must use data from both an imported Excel file and a DirectQuery SQL source. Which feature enables this scenario?

A. Live connection
B. Power BI dataflow
C. Composite model
D. Shared semantic model

✅ Correct Answer: C

Explanation:
Composite models allow combining Import and DirectQuery data sources.

Exam Readiness Check ✅

If you can:

Quickly identify Import vs DirectQuery vs Live
Recognize shared semantic model scenarios
Recall modeling limitations by connection type

…then you’re in great shape for this PL-300 objective.

Go back to the PL-300 Exam Prep Hub main page

Analytics, Microsoft Certification, PL-300, Power BI January 17, 2026

Practice Questions: Change data source settings, including credentials and privacy levels (PL-300 Exam Prep)

This post is a part of the PL-300: Microsoft Power BI Data Analyst Exam Prep Hub; and this topic falls under these sections:
Prepare the data (25–30%)
  --> Get or connect to data
    --> Change data source settings, including credentials and privacy levels

Below are 10 practice questions (with answers and explanations) for this topic of the exam.
There are also 2 practice tests for the PL-300 exam with 60 questions each (with answers) available on the hub.

Practice Questions

Question 1

A dataset refresh succeeds in Power BI Desktop but fails after being published to the Power BI Service. What is the most likely cause?

A. Privacy levels were set incorrectly in Desktop
B. Credentials were not configured in the Power BI Service
C. Query folding was disabled
D. The dataset uses Import mode

✅ Correct Answer: B

Explanation:
Credentials used in Power BI Desktop do not automatically transfer to the Power BI Service. They must be configured separately.

Question 2

Which authentication method is most commonly used when connecting to cloud-based data sources in the Power BI Service?

A. Windows authentication
B. Database authentication
C. OAuth (Microsoft Entra ID)
D. Anonymous authentication

✅ Correct Answer: C

Explanation:
Cloud data sources typically use OAuth through Microsoft Entra ID for secure authentication.

Question 3

What is the primary purpose of privacy levels in Power BI?

A. To encrypt data at rest
B. To restrict report sharing
C. To prevent unintended data leakage when combining sources
D. To improve dataset refresh performance

✅ Correct Answer: C

Explanation:
Privacy levels isolate data sources to prevent sensitive data from being unintentionally shared across sources.

Question 4

A Power BI model combines data from a Private SQL database and a Public web source. What is a potential impact of this configuration?

A. Scheduled refresh will be disabled
B. Query folding may be prevented
C. Calculated measures will not work
D. Row-level security will be removed

✅ Correct Answer: B

Explanation:
When combining sources with different privacy levels, Power BI may prevent query folding to enforce data isolation.

Question 5

Which privacy level should be assigned to internal corporate data sources?

A. Public
B. Anonymous
C. Organizational
D. Private

✅ Correct Answer: C

Explanation:
Organizational privacy level is intended for internal company data that can be shared within the organization.

Question 6

Which statement about the “Ignore privacy levels” option is true?

A. It applies to both Desktop and Service
B. It permanently disables privacy enforcement
C. It only applies in Power BI Desktop
D. It improves dataset refresh performance in the Service

✅ Correct Answer: C

Explanation:
Ignoring privacy levels is a Desktop-only setting and is not honored during refresh in the Power BI Service.

Question 7

When are credentials typically required to be re-entered?

A. When a report is viewed by another user
B. When the dataset owner changes
C. When a visual is modified
D. When a slicer is added

✅ Correct Answer: B

Explanation:
Changing dataset ownership often requires credentials to be re-entered to maintain refresh functionality.

Question 8

Which scenario requires the use of an on-premises data gateway?

A. Connecting to Azure SQL Database
B. Connecting to a Power BI dataset
C. Connecting to an on-premises SQL Server from the Service
D. Connecting to SharePoint Online

✅ Correct Answer: C

Explanation:
On-premises data sources require a gateway to allow the Power BI Service to access them.

Question 9

What happens if incorrect credentials are configured in the Power BI Service?

A. The report will not open
B. Visuals will display blank data
C. Scheduled refresh will fail
D. The dataset will be deleted

✅ Correct Answer: C

Explanation:
Incorrect credentials cause scheduled refresh failures, even if reports still open using cached data.

Question 10

Which action should you take first when a scheduled refresh suddenly starts failing?

A. Rebuild the data model
B. Change privacy levels
C. Check and update credentials in the Service
D. Disable query folding

✅ Correct Answer: C

Explanation:
Most unexpected refresh failures are caused by expired or changed credentials and should be checked first.

Exam Readiness Summary ✅

If you can confidently:

Distinguish Desktop vs Service credentials
Explain why privacy levels exist
Identify refresh and gateway issues
Understand how privacy levels affect query folding

…then you are well prepared for this PL-300 exam objective.

Go back to the PL-300 Exam Prep Hub main page

Analytics, Business Intelligence, Data Education & Training, Microsoft Certification, PL-300, Power BI January 17, 2026

Practice Questions: Choose between DirectQuery and Import (PL-300 Exam Prep)

This post is a part of the PL-300: Microsoft Power BI Data Analyst Exam Prep Hub; and this topic falls under these sections:
Prepare the data (25–30%)
   --> Get or connect to data
      --> Choose between DirectQuery and Import

Below are 10 practice questions (with answers and explanations) for this topic of the exam.
There are also 2 practice tests for the PL-300 exam with 60 questions each (with answers) available on the hub.

Practice Questions

Question 1

A Power BI report must deliver the fastest possible visual response. The dataset is moderate in size and refreshed once per day. Which connectivity mode should you choose?

A. DirectQuery
B. Live connection
C. Import
D. Composite model

✅ Correct Answer: C

Explanation:
Import mode stores data in memory, providing the fastest performance and full modeling capabilities.

Question 2

A report must show up-to-the-minute transaction data from a large operational database. Data must remain in the source system. What is the best option?

A. Import
B. DirectQuery
C. Live connection
D. Power BI dataflow

✅ Correct Answer: B

Explanation:
DirectQuery retrieves data directly from the source in real time and avoids importing large datasets.

Question 3

Which limitation is most commonly associated with DirectQuery?

A. No scheduled refresh support
B. Reduced modeling and DAX capabilities
C. Inability to use row-level security
D. Inability to connect to SQL Server

✅ Correct Answer: B

Explanation:
DirectQuery limits certain modeling features, including calculated tables and some DAX expressions.

Question 4

A dataset contains a small product lookup table and a very large fact table that updates continuously. What is the most appropriate solution?

A. Import both tables
B. Use DirectQuery for both tables
C. Use a composite model
D. Use a live connection

✅ Correct Answer: C

Explanation:
Composite models allow importing small static tables while using DirectQuery for large, frequently updated tables.

Question 5

Which factor has the greatest impact on report performance when using DirectQuery?

A. Number of visuals on the page
B. Power BI Desktop version
C. Performance of the source system
D. Dataset refresh frequency

✅ Correct Answer: C

Explanation:
DirectQuery sends queries to the source system, so performance depends heavily on the source’s ability to handle queries.

Question 6

When is Import mode generally not recommended?

A. When modeling flexibility is required
B. When dataset size exceeds practical memory limits
C. When reports need fast interactivity
D. When refresh can occur on a schedule

✅ Correct Answer: B

Explanation:
Very large datasets may exceed memory constraints, making Import impractical.

Question 7

Which statement about data freshness is true?

A. Import mode always shows real-time data
B. DirectQuery requires scheduled refresh
C. Import mode relies on dataset refresh
D. DirectQuery stores data in memory

✅ Correct Answer: C

Explanation:
Import mode displays data as of the last refresh, while DirectQuery retrieves data at query time.

Question 8

A Power BI report must support complex DAX measures and calculated tables. Data updates hourly and does not need real-time accuracy. What should you choose?

A. DirectQuery
B. Import
C. Live connection
D. Streaming dataset

✅ Correct Answer: B

Explanation:
Import mode supports full DAX and modeling flexibility and is appropriate when real-time data is not required.

Question 9

Which scenario is the best candidate for DirectQuery?

A. Monthly financial reporting
B. Historical trend analysis
C. Real-time inventory monitoring
D. Static reference data

✅ Correct Answer: C

Explanation:
Real-time or near-real-time monitoring scenarios are ideal for DirectQuery.

Question 10

Why might a Power BI Data Analyst avoid DirectQuery unless necessary?

A. It cannot connect to cloud data sources
B. It disables report sharing
C. It can negatively impact performance and modeling flexibility
D. It does not support security

✅ Correct Answer: C

Explanation:
DirectQuery introduces performance dependencies on the source system and limits modeling features, making Import preferable when possible.

Exam Readiness Check ✅

You’re well prepared for this PL-300 objective if you can:

Identify real-time vs scheduled refresh needs
Balance performance vs flexibility
Recognize large-scale data scenarios
Explain why DirectQuery is chosen—not just when

Go back to the PL-300 Exam Prep Hub main page

Analytics, Microsoft Certification, PL-300, Power BI January 17, 2026

Practice Questions: Select the appropriate column data type (PL-300 Exam Prep)

This post is a part of the PL-300: Microsoft Power BI Data Analyst Exam Prep Hub; and this topic falls under these sections:
Prepare the data (25–30%)
   --> Transform and load the data
      --> Select the appropriate column data type

Below are 10 practice questions (with answers and explanations) for this topic of the exam.
There are also 2 practice tests for the PL-300 exam with 60 questions each (with answers) available on the hub.

Practice Questions

1. Choosing the Correct Type for IDs

A column contains values such as 100245, 100312, and 100587. These values are used only to uniquely identify records. What data type should you use?

A. Whole Number
B. Decimal Number
C. Text
D. Fixed Decimal Number

Correct Answer: C

Explanation:
Identifiers should be stored as Text, not numeric types, since they are not used in calculations and may contain leading zeros in the future.

2. Fixing Incorrect Aggregation Behavior

A column representing revenue does not aggregate correctly in visuals. The column’s data type is set to Text. What should you do?

A. Change the data type to Decimal Number
B. Replace null values
C. Remove duplicate rows
D. Create a calculated column

Correct Answer: A

Explanation:
Text columns cannot be aggregated. Changing the data type to Decimal Number enables proper numeric calculations.

3. When to Use Fixed Decimal Number

Which scenario is BEST suited for the Fixed Decimal Number data type?

A. Product quantity
B. Currency values requiring precision
C. Customer ID numbers
D. Percentage values used in visuals

Correct Answer: B

Explanation:
Fixed Decimal Number is designed for financial data where precision is critical, such as currency.

4. Handling Dates Imported as Text

A column contains date values stored as text due to regional formatting differences. What is the BEST way to convert them?

A. Change data type directly to Date
B. Replace values manually
C. Change data type using locale
D. Create a calculated column

Correct Answer: C

Explanation:
Using Change Type → Using Locale ensures Power BI interprets dates correctly based on regional formats.

5. Preventing Data Type Errors

You are experiencing errors after converting a column’s data type. What should you do FIRST?

A. Remove rows with errors
B. Replace errors with null
C. Clean and validate the source values
D. Disable automatic type detection

Correct Answer: C

Explanation:
Errors often occur because the data is not clean. Always clean data before converting types.

6. Relationship Creation Failure

You are unable to create a relationship between two tables on a key column. What is the MOST likely reason?

A. The column contains duplicate values
B. The column contains null values
C. The data types of the columns do not match
D. The column is hidden

Correct Answer: C

Explanation:
Relationship columns must have matching data types for Power BI to create the relationship.

7. Choosing Between Date and Date/Time

A column contains dates only, with no time component. Which data type is MOST appropriate?

A. Text
B. Date/Time
C. Date
D. Date/Time/Timezone

Correct Answer: C

Explanation:
Using Date avoids unnecessary time components and improves filtering and grouping accuracy.

8. Automatic Type Detection Issues

Power BI automatically assigns a Whole Number type to a column that should be text. Why is this a problem?

A. It increases model size
B. It prevents the column from being filtered
C. It can remove leading zeros
D. It causes duplicate values

Correct Answer: C

Explanation:
Numeric types do not preserve leading zeros, which can be critical for codes or IDs.

9. Best Practice for Applying Data Types

When should you apply data type changes in Power Query?

A. Immediately after loading the data
B. After creating relationships
C. After cleaning and transforming the data
D. After publishing the report

Correct Answer: C

Explanation:
Applying data types after cleaning prevents conversion errors and ensures consistent transformations.

10. Diagnosing Model Performance Issues

A Power BI model is larger than expected and performs poorly. One cause is unnecessary Date/Time columns where only Date is needed. What should you do?

A. Convert Date/Time columns to Date
B. Convert Date columns to Text
C. Remove the columns
D. Disable auto date/time

Correct Answer: A

Explanation:
Using Date instead of Date/Time reduces model size and improves performance when time data is not required.

✅ PL-300 Exam Takeaways

Data type selection affects calculations, relationships, and performance
IDs should almost always be Text
Clean data before converting data types
Know when to use Decimal vs Fixed Decimal
Expect subtle, behavior-based questions on the exam

Go back to the PL-300 Exam Prep Hub main page

Analytics, Microsoft Certification, PL-300, Power BI January 17, 2026

Practice Questions: Create and transform columns (PL-300 Exam Prep)

This post is a part of the PL-300: Microsoft Power BI Data Analyst Exam Prep Hub; and this topic falls under these sections:
Prepare the data (25–30%)
   --> Transform and load the data
      --> Create and transform columns

Below are 10 practice questions (with answers and explanations) for this topic of the exam.
There are also 2 practice tests for the PL-300 exam with 60 questions each (with answers) available on the hub.

Practice Questions

1. Choosing the Correct Tool

You need to split a column named FullName into FirstName and LastName before loading the data into the model. What is the BEST approach?

A. Create two DAX calculated columns
B. Use Split Column in Power Query
C. Use a measure
D. Use a visual-level calculation

Correct Answer: B

Explanation:
Structural changes like splitting columns should be done in Power Query before the data loads into the model.

2. Fixing Inconsistent Text Data

A column contains values with inconsistent casing (e.g., “north”, “North”, “NORTH”). What transformation should you apply?

A. Replace values
B. Trim
C. Format → Uppercase
D. Conditional column

Correct Answer: C

Explanation:
Formatting text to Uppercase (or Lowercase) standardizes values efficiently and avoids manual replacements.

3. Extracting Date Components

You need a column containing only the year from an existing Date column. What is the BEST option?

A. Create a DAX measure
B. Extract → Year in Power Query
C. Convert the date to text
D. Use a conditional column

Correct Answer: B

Explanation:
Power Query provides built-in Extract options for date components such as Year, Month, and Day.

4. Creating Business Logic Columns

You want to classify sales as “High” when the amount is greater than 1,000 and “Low” otherwise. What should you use?

A. Replace values
B. Split column
C. Conditional column
D. Index column

Correct Answer: C

Explanation:
Conditional columns are designed for simple business logic and can be created through the UI without writing code.

5. Removing Hidden Characters

A text column contains hidden, non-printable characters that affect filtering. Which transformation should you apply?

A. Trim
B. Clean
C. Replace errors
D. Capitalize Each Word

Correct Answer: B

Explanation:
Clean removes non-printable characters that are not visible but can cause data issues.

6. Merging Text Values into One Column

You need to combine City and State columns into a single column called Location. What should you use?

A. Split Column
B. Replace Values
C. Merge Columns
D. Custom Column

Correct Answer: C

Explanation:
Merge Columns combines multiple columns into one using a specified delimiter.

7. When to Use a Custom Column

When should you create a Custom Column instead of a Conditional Column?

A. When no logic is required
B. When simple IF/ELSE logic is needed
C. When more complex formulas or functions are required
D. When renaming columns

Correct Answer: C

Explanation:
Custom columns use Power Query M code, allowing more complex logic than conditional columns.

8. Power Query vs DAX Calculated Columns

Which scenario is BEST suited for a DAX calculated column instead of a Power Query column?

A. Cleaning inconsistent text values
B. Splitting a column by delimiter
C. Creating a column dependent on model relationships
D. Removing leading spaces

Correct Answer: C

Explanation:
DAX calculated columns can reference model relationships, while Power Query columns cannot.

9. Creating Row Identifiers

You need to add a sequential number to each row to preserve order. What should you use?

A. Conditional column
B. Custom column
C. Index column
D. Replace values

Correct Answer: C

Explanation:
An Index Column assigns a sequential number to each row and is commonly used for sorting or tracking order.

10. Best Practice for Column Creation

Which is the BEST practice when creating and transforming columns?

A. Create all columns using DAX
B. Transform columns after publishing the report
C. Clean data before creating derived columns
D. Avoid renaming columns

Correct Answer: C

Explanation:
Derived columns should be created from clean, validated data to avoid compounding errors.

✅ PL-300 Exam Takeaways

Use Power Query for structural transformations
Use DAX calculated columns when logic depends on model context
Know when to split, merge, extract, or derive columns
Expect scenario-based questions testing tool choice and order of operations

Go back to the PL-300 Exam Prep Hub main page

Analytics, Business Intelligence, Microsoft Certification, PL-300, Power BI January 17, 2026

Practice Questions: Pivot, Unpivot, and Transpose Data (PL-300 Exam Prep)

This post is a part of the PL-300: Microsoft Power BI Data Analyst Exam Prep Hub; and this topic falls under these sections:
Prepare the data (25–30%)
   --> Transform and load the data
      --> Pivot, Unpivot, and Transpose Data

Below are 10 practice questions (with answers and explanations) for this topic of the exam.
There are also 2 practice tests for the PL-300 exam with 60 questions each (with answers) available on the hub.

Practice Questions

Question 1

A dataset contains monthly sales stored in separate columns named Jan, Feb, Mar, and so on. You need to create a time-series visual that supports filtering by month. What should you do?

A. Pivot the Month columns
B. Unpivot the Month columns
C. Transpose the table
D. Create calculated columns in DAX

Correct Answer: B

Explanation:
Unpivoting converts repeated value columns into a single attribute–value pair, producing a normalized structure suitable for time-series analysis.

Question 2

You want to convert unique values from a column into individual columns, displaying aggregated results per category. Which transformation should you use?

A. Unpivot Columns
B. Transpose Table
C. Pivot Columns
D. Group By

Correct Answer: C

Explanation:
Pivoting turns values from one column into multiple columns, often requiring an aggregation function such as Sum or Count.

Question 3

Which transformation flips all rows into columns and columns into rows?

A. Pivot
B. Unpivot
C. Group By
D. Transpose

Correct Answer: D

Explanation:
Transpose rotates the entire table structure and is typically used when the table orientation is incorrect.

Question 4

After transposing a table, the first row now contains the desired column headers. What should you do next?

A. Unpivot Columns
B. Rename columns manually
C. Promote the first row to headers
D. Create a calculated table

Correct Answer: C

Explanation:
Promoting the first row to headers finalizes the transformation after transposing data.

Question 5

Which scenario best indicates that unpivoting is required?

A. Values in a column need to be summarized
B. Categories should become column headers
C. Repeating attribute columns must become rows
D. Data needs to be grouped by multiple keys

Correct Answer: C

Explanation:
Unpivoting is used to normalize data where multiple columns represent the same attribute (such as months or measures).

Question 6

What is a potential negative impact of pivoting data in Power Query?

A. Increased DAX complexity
B. Reduced model performance due to many columns
C. Loss of relationships
D. Automatic removal of measures

Correct Answer: B

Explanation:
Pivoting can significantly increase the number of columns, bloating the model and negatively affecting performance.

Question 7

You receive a dataset where column headers are stored in the first column instead of the first row. Which transformation is most appropriate?

A. Unpivot
B. Pivot
C. Transpose
D. Group By

Correct Answer: C

Explanation:
Transposing the table re-orients rows and columns so headers can be promoted correctly.

Question 8

Which transformation often requires you to specify an aggregation function?

A. Transpose
B. Unpivot
C. Pivot
D. Remove Columns

Correct Answer: C

Explanation:
Pivoting may result in multiple values for a new column and therefore requires aggregation (Sum, Count, Average, etc.).

Question 9

You want to keep identifier columns (such as ProductID and Region) while converting all other columns into rows. Which option is best?

A. Pivot Selected Columns
B. Unpivot Other Columns
C. Transpose Table
D. Group By

Correct Answer: B

Explanation:
Unpivot Other Columns preserves key identifier columns while normalizing all remaining columns.

Question 10

Which statement reflects a PL-300 best practice for reshaping data?

A. Always pivot data before loading
B. Use transpose for large fact tables
C. Normalize data before modeling when possible
D. Avoid unpivoting to improve performance

Correct Answer: C

Explanation:
Normalized (unpivoted) data supports better filtering, relationships, and DAX flexibility, which aligns with PL-300 modeling principles.

Final Exam Tips for This Topic

Identify whether the data is wide or tall
Pivot = rows → columns
Unpivot = columns → rows
Transpose = rotate entire table
Watch for aggregation requirements when pivoting
The exam emphasizes choosing the right transformation, not UI memorization

Go back to the PL-300 Exam Prep Hub main page

Analytics, Business Intelligence, Data Education & Training, Microsoft Certification, PL-300, Power BI January 17, 2026

Practice Questions: Convert Semi-Structured Data to a Table (PL-300 Exam Prep)

This post is a part of the PL-300: Microsoft Power BI Data Analyst Exam Prep Hub; and this topic falls under these sections:
Prepare the data (25–30%)
   --> Transform and load the data
      --> Convert Semi-Structured Data to a Table

Below are 10 practice questions (with answers and explanations) for this topic of the exam.
There are also 2 practice tests for the PL-300 exam with 60 questions each (with answers) available on the hub.

Practice Questions

Question 1

You load a JSON file into Power BI. The resulting table contains a single column where each row shows List. What is the first step to analyze the data?

A. Expand the column
B. Convert the list to a table
C. Promote headers
D. Split the column by delimiter

Correct Answer: B

Explanation:
Lists must be converted into tables before they can be expanded or analyzed as rows.

Question 2

A column in Power Query displays Record in each row. What does this indicate?

A. The column contains duplicated values
B. The column contains nested structured fields
C. The column contains multiple rows per record
D. The column contains untyped data

Correct Answer: B

Explanation:
A Record represents a nested structure with named fields that can be expanded into columns.

Question 3

Which Power Query action is used to expose fields stored inside a record?

A. Convert to Table
B. Pivot Column
C. Expand Column
D. Transpose Table

Correct Answer: C

Explanation:
Expanding a record reveals its internal fields as individual columns.

Question 4

An API response loads as a table with a column containing lists of values. What is the correct transformation sequence?

A. Expand → Promote Headers
B. Convert to Table → Expand
C. Split Column → Fill Down
D. Group By → Expand

Correct Answer: B

Explanation:
Lists must be converted into tables first, after which they can be expanded.

Question 5

After expanding nested data, you notice duplicate rows in your fact table. What is the most likely cause?

A. Incorrect data type
B. Expanding without understanding data granularity
C. Missing relationships
D. Failure to promote headers

Correct Answer: B

Explanation:
Expanding nested structures without considering the grain can duplicate rows and inflate fact tables.

Question 6

You import an Excel file where headers appear in multiple rows instead of a single row. What is the most appropriate approach?

A. Expand the column
B. Convert the table to a list
C. Transpose the table and promote headers
D. Group rows by column

Correct Answer: C

Explanation:
Transposing realigns rows and columns so headers can be promoted properly.

Question 7

Which Power Query feature is most useful when category labels appear only once and apply to multiple rows below?

A. Replace Values
B. Fill Down
C. Unpivot Columns
D. Merge Queries

Correct Answer: B

Explanation:
Fill Down propagates header or category values to related rows, common in semi-structured spreadsheets.

Question 8

Why is it recommended to expand only required fields when converting semi-structured data?

A. To reduce report refresh frequency
B. To improve visual formatting
C. To reduce model size and complexity
D. To enable DirectQuery mode

Correct Answer: C

Explanation:
Expanding unnecessary fields increases model size and can negatively impact performance and usability.

Question 9

Which transformation should be completed before creating relationships in the data model?

A. Creating measures
B. Flattening semi-structured data
C. Formatting visuals
D. Applying row-level security

Correct Answer: B

Explanation:
Relationships require clean, tabular data. Semi-structured data must be flattened first.

Question 10

Which statement best reflects a PL-300 best practice for handling semi-structured data?

A. Leave nested data unexpanded until report creation
B. Use DAX to flatten semi-structured data
C. Normalize and flatten data in Power Query
D. Always transpose semi-structured tables

Correct Answer: C

Explanation:
Power Query is the correct place to normalize and flatten semi-structured data before modeling and analysis.

Final Exam Tips for This Topic

Recognize lists vs records vs tables
Lists → Convert to table
Records → Expand
Inspect data grain before expanding
Clean data before flattening
This topic is about recognition and transformation choices, not memorizing UI clicks

Go back to the PL-300 Exam Prep Hub main page

Analytics, Business Intelligence, Data Education & Training, Microsoft Certification, PL-300, Power BI January 17, 2026

Practice Questions: Create Fact Tables and Dimension Tables (PL-300 Exam Prep)

This post is a part of the PL-300: Microsoft Power BI Data Analyst Exam Prep Hub; and this topic falls under these sections:
Prepare the data (25–30%)
   --> Transform and load the data
      --> Create Fact Tables and Dimension Tables

Below are 10 practice questions (with answers and explanations) for this topic of the exam.
There are also 2 practice tests for the PL-300 exam with 60 questions each (with answers) available on the hub.

Practice Questions

Question 1

A table contains SalesAmount, Quantity, ProductName, ProductCategory, CustomerName, and OrderDate. Which columns should remain in the fact table?

A. ProductName, ProductCategory
B. CustomerName, OrderDate
C. SalesAmount, Quantity
D. ProductName, CustomerName

Correct Answer: C

Explanation:
Fact tables store numeric measures that are aggregated, such as SalesAmount and Quantity. Descriptive attributes belong in dimension tables.

Question 2

What is the primary purpose of a dimension table?

A. Store transaction-level data
B. Provide descriptive context for facts
C. Improve visual formatting
D. Store calculated measures

Correct Answer: B

Explanation:
Dimension tables provide descriptive attributes (such as names, categories, and dates) that are used to filter and group fact data.

Question 3

Which relationship type is most appropriate between a dimension table and a fact table?

A. Many-to-many
B. One-to-one
C. One-to-many
D. Bi-directional

Correct Answer: C

Explanation:
A dimension table contains unique keys, while the fact table contains repeated foreign keys, creating a one-to-many relationship.

Question 4

You create a Product dimension table but forget to remove duplicate ProductID values. What issue is most likely?

A. Measures will return blank values
B. Relationships cannot be created correctly
C. Visuals will fail to render
D. DAX functions will not work

Correct Answer: B

Explanation:
Dimension tables must have unique key values. Duplicates prevent proper one-to-many relationships.

Question 5

Which schema design is recommended by Microsoft for Power BI models?

A. Snowflake schema
B. Flat table schema
C. Galaxy schema
D. Star schema

Correct Answer: D

Explanation:
The star schema is recommended for performance, simplicity, and easier DAX calculations in Power BI.

Question 6

Where should fact and dimension tables typically be created?

A. In DAX measures
B. In Power Query during data preparation
C. In visuals after loading data
D. In the Power BI Service

Correct Answer: B

Explanation:
Fact and dimension tables should be shaped in Power Query before loading into the data model.

Question 7

A model uses the same Date table for Order Date and Ship Date. What type of dimension is this?

A. Slowly changing dimension
B. Degenerate dimension
C. Role-playing dimension
D. Bridge table

Correct Answer: C

Explanation:
A role-playing dimension is used multiple times in different roles, such as Order Date and Ship Date.

Question 8

Which is a valid reason not to split a dataset into fact and dimension tables?

A. The dataset is extremely small and static
B. The dataset contains numeric measures
C. The model requires relationships
D. The data will be refreshed regularly

Correct Answer: A

Explanation:
For very small or simple datasets, splitting into facts and dimensions may add unnecessary complexity.

Question 9

What is the primary performance benefit of separating fact and dimension tables?

A. Faster visual rendering due to fewer measures
B. Reduced memory usage and simpler filter paths
C. Automatic indexing of columns
D. Improved DirectQuery support

Correct Answer: B

Explanation:
Star schemas reduce duplication of descriptive data and create efficient filter paths, improving performance.

Question 10

Which modeling mistake often leads to the unnecessary use of bi-directional relationships?

A. Using too many measures
B. Poor star schema design
C. Too many dimension tables
D. Using calculated columns

Correct Answer: B

Explanation:
Bi-directional relationships are often used to compensate for poor model design. A clean star schema usually requires only single-direction filtering.

Final Exam Tips for This Topic

Measures → Fact tables
Descriptive attributes → Dimension tables
Use Power Query to shape tables before modeling
Ensure unique keys in dimension tables
Prefer star schema over flat or snowflake models
Know when not to over-model

Go back to the PL-300 Exam Prep Hub main page

Analytics, Business Intelligence, Data Education & Training, Microsoft Certification, PL-300, Power BI January 17, 2026

Practice Questions: Identify when to use reference or duplicate queries and the resulting impact (PL-300 Exam Prep)

This post is a part of the PL-300: Microsoft Power BI Data Analyst Exam Prep Hub; and this topic falls under these sections:
Prepare the data (25–30%)
   --> Transform and load the data
      --> Identify when to use reference or duplicate queries and the resulting impact

Below are 10 practice questions (with answers and explanations) for this topic of the exam.
There are also 2 practice tests for the PL-300 exam with 60 questions each (with answers) available on the hub.

Practice Questions

Question 1

You have a query that cleans and standardizes sales data. You need to create several dimension tables from this cleaned dataset. Which option should you use?

A. Duplicate the query for each dimension
B. Reference the query for each dimension
C. Import the source data multiple times
D. Merge the query with itself

✅ Correct Answer: B

Explanation:
Referencing allows multiple tables to inherit the same cleaned logic from a single base query. This ensures consistency and reduces repeated transformation steps, which is a recommended best practice for production models.

Question 2

What is the primary difference between a referenced query and a duplicated query?

A. Referenced queries refresh faster
B. Duplicated queries do not support transformations
C. Referenced queries depend on the original query
D. Duplicated queries cannot be loaded to the model

✅ Correct Answer: C

Explanation:
A referenced query is dependent on its source query and will reflect any changes made to it. A duplicated query is an independent copy with no dependency.

Question 3

A change made to a base query causes multiple downstream queries to fail during refresh. What is the most likely reason?

A. The downstream queries were duplicated
B. The downstream queries were referenced
C. The model relationships were deleted
D. The data source credentials expired

✅ Correct Answer: B

Explanation:
Referenced queries rely on the base query. If a breaking change is introduced (such as removing or renaming a column), all dependent referenced queries may fail.

Question 4

When should you duplicate a query instead of referencing it?

A. When you want transformations to stay consistent
B. When creating multiple dimension tables
C. When experimenting with major changes
D. When reducing refresh dependencies

✅ Correct Answer: C

Explanation:
Duplicating a query is ideal when testing or experimenting, because changes will not affect other queries or downstream dependencies.

Question 5

Which impact is most commonly associated with excessive query duplication?

A. Improved refresh reliability
B. Reduced data volume
C. Increased maintenance effort
D. Better data lineage visibility

✅ Correct Answer: C

Explanation:
Duplicating queries can lead to repeated transformation logic, making the model harder to maintain and increasing the risk of inconsistent data shaping.

Question 6

How does Power BI’s View Lineage represent referenced queries?

A. As independent branches
B. As disconnected tables
C. As upstream and downstream dependencies
D. As hidden queries

✅ Correct Answer: C

Explanation:
Referenced queries appear as downstream dependencies in View Lineage, clearly showing how data flows from base queries to derived queries.

Question 7

You want to ensure that a change to data cleansing logic automatically applies to all derived tables. What should you do?

A. Duplicate the query
B. Reference the query
C. Disable query loading
D. Create calculated tables

✅ Correct Answer: B

Explanation:
Referencing ensures that any change to the base query propagates to all dependent queries automatically.

Question 8

Which of the following is a common mistake when using referenced queries?

A. Using them for experimentation
B. Using them for dimension creation
C. Forgetting that changes propagate downstream
D. Using them to centralize data cleaning

✅ Correct Answer: C

Explanation:
A frequent mistake is forgetting that changes to a referenced base query can unintentionally affect multiple dependent queries.

Question 9

Which approach generally results in a cleaner and more maintainable data model?

A. Duplicating all queries
B. Referencing a well-designed base query
C. Importing data separately for each table
D. Performing transformations in DAX

✅ Correct Answer: B

Explanation:
Using a base query with referenced downstream queries centralizes transformation logic and simplifies maintenance, which aligns with Microsoft’s recommended modeling practices.

Question 10

Which scenario best illustrates when NOT to use a referenced query?

A. Creating a product dimension
B. Applying consistent formatting rules
C. Testing a new transformation approach
D. Creating multiple tables from a single source

✅ Correct Answer: C

Explanation:
Referenced queries should not be used when testing or experimenting with transformations, because changes may impact other dependent queries. Duplicating is safer in this case.