Level 4 exam — Microsoft Power BI

Exam — fresh questions

12 questions to validate level 4 with Zineb, a management controller at an agri-food SME in Algiers: design and themes, row-level security, dynamic parameters, optimization and team architecture.
Answer everything, then click Check: your score and explanations will appear.

4.1 · Recall

1. What is the purpose of a Power BI report theme?

To apply palette and fonts to the whole report in one go To filter the visible rows according to the signed-in user To measure the time of each visual on the active page To simulate an increase in revenue with a slider

The theme is a JSON of colors and fonts applied in one go: you set it early, export it, reuse it.

4.1 · Application

2. Zineb wants only the top city to stay in color. What does she do?

She colors each city with a different shade by hand She applies conditional formatting (fx) She changes the visual to a scatter chart without a legend She adds a city slicer in front of the visual

Color means something: conditional formatting (fx) drives it by the data — only the top city stays in accent.

4.1 · Arbitration

3. Recolor each visual or set a theme: what does Zineb keep?

Recolor each visual, it's simpler to maintain Remove all colors to stay neutral Set the theme once and export it to reuse it Switch the report to dark mode to decide

A theme set once then exported applies everywhere and is reusable — it's standardization that pays off.

4.2 · Recall

4. What does row-level security (RLS) filter?

The fonts and colors of the published report The number of visuals shown on each page The calculation speed of the model's measures The visible rows according to the signed-in user

RLS filters rows by user: the same report gives each person their own data scope.

4.2 · Application

5. The Algiers manager must only see their region. Where to place the DAX filter?

On the Region dimension, in an RLS role On the fact table, sale line by sale line On each visual of each page by hand On the theme, to hide the other regions

You filter the DIMENSION: through the relationship, the filter propagates to the facts — more reliable than filtering the facts.

4.2 · Application

6. Zineb tests her RLS with her own account and sees everything. Conclusion?

RLS is broken, it must be rebuilt That's normal: the author sees everything; test "View as" You must first publish the report for RLS to act You must switch the connection to DirectQuery

The author is not subject to RLS: you test with "View as" (or even a specific email for dynamic).

4.3 · Application

7. How to swap the measure shown in the same visual?

By duplicating the page for each measure By coloring the bars according to the value With a field parameter driven by a slicer By writing one measure per visual by hand

The field parameter lets a slicer swap the measure or the axis: a single visual, several angles.

4.3 · Application

8. Zineb wants to simulate a 10% revenue increase without changing the data. Which tool?

Conditional formatting on the bars An RLS filter restricting the sales table A JSON theme applied to the whole report A what-if parameter read by SELECTEDVALUE

What-if creates a slider whose value (SELECTEDVALUE) feeds a simulation measure — without touching the data.

4.4 · Application

9. Where to start to diagnose a slow report?

With the Performance Analyzer, which times each visual By changing to a more powerful computer By adding RAM to the workstation By converting the whole model to DirectQuery

You don't guess a slowdown: you time it first with the Performance Analyzer (View).

4.4 · Arbitration

10. A visual's time is in "DAX query". Where to act first?

On the visual, by reducing the number of categories On the model or the measure, at the source (Power Query) On the machine, by changing the graphics card On the theme, by lightening the color palette

Slow DAX query = model/measure to review: you act first on the model, at the source, not on the visual.

4.5 · Arbitration

11. What does a dataflow allow?

To simulate a scenario using a slider To display the report on a mobile phone To share Power Query preparation in the cloud To test an RLS role before publishing

The dataflow centralizes Power Query preparation in the Service: prepare once, consume everywhere.

4.5 · Arbitration

12. Several reports redo the same cleanup. What is the right reflex?

Copy the cleanup into each isolated .pbix file Let each report define its own revenue Create an RLS role per report author Create a dataflow and a single central model

A repeated cleanup becomes a dataflow, and the key measures a source-of-truth model consumed by live connection.

What's next?

Score ≥ 80%: you master design, security and performance — head for Level 5, publishing and distributing in the Service.
Score between 50% and 80%: review the failed section(s) (4.1 themes, 4.2 RLS, 4.3 parameters, 4.4 performance, 4.5 architecture), reread the course, then retake the test.
Score < 50%: reread all of level 4 while applying each 48h challenge (house theme, RLS role, what-if, timing, team roadmap) before trying again.
And after that? Keep your theme, your RLS role and your roadmap: level 5 publishes and distributes them in the Service.