This post is a part of the PL-300: Microsoft Power BI Data Analyst Exam Prep Hub; and this topic falls under these sections:
Model the data (25–30%)
 --> Optimize model performance
    --> Improve Performance by Identifying and Removing Unnecessary Rows and 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. Why does removing unused columns improve Power BI model performance?

A. It reduces the number of visuals that can be created
B. It decreases the number of DAX measures required
C. It reduces memory usage in the VertiPaq engine
D. It disables automatic relationships

✅ Correct Answer: C

Explanation:
Power BI stores data in the VertiPaq in-memory engine. Each column consumes memory—even if it is hidden. Removing unused columns reduces the model’s memory footprint, leading to faster refreshes and better query performance.

---

2. Where should unnecessary rows ideally be removed to achieve the best performance improvement?

A. In DAX measures
B. In calculated tables
C. In report-level filters
D. In Power Query

✅ Correct Answer: D

Explanation:
Removing rows in Power Query prevents the data from ever being loaded into the model. This reduces model size and improves refresh and query performance. Filtering data after it’s loaded (DAX or visuals) does not reduce memory usage.

---

3. Which type of column typically has the greatest negative impact on model size?

A. Numeric columns with repeated values
B. Boolean columns
C. High-cardinality text columns
D. Date columns

✅ Correct Answer: C

Explanation:
High-cardinality text columns (e.g., IDs, comments, URLs) compress poorly in VertiPaq and significantly increase model size. Removing unused high-cardinality columns is a key performance optimization technique.

---

4. A column is hidden in the model but not used in any visuals or calculations. What impact does it have on performance?

A. No impact, because it is hidden
B. It improves compression automatically
C. It still consumes memory
D. It is removed at refresh time

✅ Correct Answer: C

Explanation:
Hidden columns are still loaded into the model and consume memory. Hiding a column does not improve performance—removing it does.

---

5. Which scenario best justifies removing historical rows from a fact table?

A. The data refresh fails intermittently
B. Reports only require the last three years of data
C. Measures return incorrect totals
D. Relationships are inactive

✅ Correct Answer: B

Explanation:
If reports only require recent data, removing older historical rows reduces table size and improves performance without affecting business requirements.

---

6. Why is removing unnecessary columns from fact tables especially important?

A. Fact tables rarely participate in relationships
B. Fact tables usually have the fewest rows
C. Fact tables typically have the most rows
D. Fact tables do not support compression

✅ Correct Answer: C

Explanation:
Fact tables often contain millions of rows. Any unnecessary column in a large fact table significantly increases memory usage and query processing time.

---

7. Which action improves performance without reducing the model size?

A. Removing unused columns in Power Query
B. Filtering rows in Power Query
C. Using report-level filters
D. Removing columns from the source system

✅ Correct Answer: C

Explanation:
Report-level filters limit what is displayed but do not remove data from the model. The data is still loaded and stored in memory.

---

8. Which of the following columns is most likely safe to remove if not used?

A. Date key used in relationships
B. Surrogate key not used in any relationship
C. Foreign key linking to a dimension
D. Numeric measure column

✅ Correct Answer: B

Explanation:
Surrogate keys that are not used in relationships, visuals, or calculations add no analytical value and only increase model size.

---

9. Why is early filtering in Power Query recommended?

A. It increases DAX calculation accuracy
B. It enables query folding and reduces data transfer
C. It hides rows from report users
D. It automatically creates relationships

✅ Correct Answer: B

Explanation:
Filtering early allows Power BI to push filtering logic back to the data source (query folding), reducing the amount of data transferred, loaded, and stored in the model.

---

10. A Power BI model is slow, and analysis shows many unused descriptive columns in dimension tables. What is the BEST action?

A. Hide the columns
B. Convert them to measures
C. Remove the columns in Power Query
D. Create calculated columns instead

✅ Correct Answer: C

Explanation:
Removing unused descriptive columns in Power Query permanently reduces model size and improves performance. Hiding columns does not provide performance benefits.

---

Final Exam Tips

• Remove, don’t hide unused columns
• Filter rows in Power Query, not visuals
• Focus on high-cardinality columns and large fact tables
• Smaller models = faster performance (this is a recurring PL-300 theme)

---

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