This is not a quiz. There are no right or wrong answers given to you here. These questions are a mirror — answer each one honestly to yourself, and let your honest answer guide where you need to grow.
Answer Yes only when you can back it up with action — when you have done it, built it, or explained it clearly. A No is not a failure. It is the most honest and useful thing you can say to yourself. Every No is a door that opens into exactly the knowledge you need next.
Open Power BI Desktop and use Get Data → Excel to load a workbook. Then move the file to a different folder and practice using Data Source Settings to update the path. This is a daily real-world skill — master the file path relationship before anything else.
Install a local MySQL or use a free cloud database. Connect Power BI using Get Data → MySQL. Write a simple SELECT statement to pull only the columns you need. Practice limiting rows with WHERE. Understanding what you import matters as much as how you import it.
Use Get Data → Web to connect to a public JSON or CSV URL. Then try connecting to a SharePoint list if you have access. Study the difference between Anonymous, Basic, and OAuth authentication. Scheduled refresh is what makes a Power BI report stay alive — understand it early.
Find or create a deliberately messy dataset. In Power Query, practice Remove Duplicates, Replace Values, Fill Down, Change Type, and Trim. Each of these solves a specific real-world data problem. Do not move on until you can do all five from memory.
Take a wide table with month columns (Jan, Feb, Mar...) and Unpivot it into a clean date-value structure. Then take two similar tables and Append them. Then take two related tables and Merge them on a key column. These three operations cover 80% of real restructuring scenarios.
Use Add Column → Conditional Column to create a category based on a value range (e.g., Low / Medium / High sales). Then open the Advanced Editor and read the M code that was generated. Modify one line. You do not need to master M — but you must not fear it.
In Power Query Editor, go to View → Column Quality and Column Profile. Load any dataset and read what those panels tell you. Identify one column with errors and one with nulls. Practice switching between profiling the top 1,000 rows versus the entire dataset — and understand why that distinction matters.
Take two tables that share a key column (e.g., Customer ID). Check that both columns have the same data type, the same formatting (no leading spaces, consistent casing), and no orphaned values. A relationship built on mismatched keys will silently produce wrong numbers — this check prevents that.
After your next data preparation session, write a short plain-language summary: what the raw data looked like, what you changed, and why. Show it to someone outside of data. If they understand it, your preparation is documented well. This skill separates a technician from a trusted analyst.
Draw a star schema by hand before touching Power BI. Place a Sales fact table in the center. Surround it with dimension tables: Date, Customer, Product, Region. Draw the lines connecting them. When you can explain why each line goes in one direction only, you understand the foundation of every reliable model.
In Power BI's Model View, create a relationship between two tables. Change the cross-filter direction from Single to Both and observe how your visuals change. Then change it back. That observable difference is cardinality and filter direction made real — not theory.
Deliberately create an ambiguous relationship in a model — two paths between the same two tables — and watch Power BI warn you. Then resolve it by making one relationship inactive and using USERELATIONSHIP() in a DAX measure to activate it when needed. Experiencing the problem is the fastest way to understand the solution.
Write a simple Total Sales = SUM(Sales[Amount]) measure. Place it in a table visual with a Year slicer. Watch the number change as you filter. That change is filter context working. Now try to explain in one sentence why the number changed — and why that is fundamentally different from an Excel cell formula. That explanation is your milestone.
Build these five measures on a real dataset, one by one: Total Revenue (SUM), Average Order Value (AVERAGE), Number of Orders (COUNT), Unique Customers (DISTINCTCOUNT), and Revenue for Product Category X (CALCULATE with a filter). Each one solves a real business question. Do not move on until all five work correctly.
Build a proper Date Table in your model first — time intelligence fails without one. Then write three measures: Sales YTD using TOTALYTD, Sales Last Year using SAMEPERIODLASTYEAR, and YoY Growth % combining both. Place them side by side in a line chart. When a business leader can see trend at a glance, that is time intelligence delivering real value.
Open View → Performance Analyzer in Power BI Desktop. Record while your report loads. Find the visual with the highest DAX query time. Copy that DAX query into DAX Studio (free tool) and read the query plan. You do not need to optimize it yet — just understand what the model is being asked to do. Awareness precedes improvement.
Study this decision: Import copies data into Power BI's engine — fast, flexible, but requires scheduled refresh. DirectQuery queries the source live — always current, but slower and limited in DAX. Build the same simple report in both modes on the same dataset. Feel the difference in speed. Then write one sentence explaining when you would choose each in a real project.
In Power BI Desktop, go to Modeling → Manage Roles. Create a role called "Region East" with a DAX filter on your Region column: [Region] = "East". Then use View As Role to test it. You should see only East data. This is the skill that allows one report to safely serve an entire organization without exposing data it should not.
Take one dataset and answer five different business questions using five different visuals: a bar chart for comparison, a line chart for trend, a pie or treemap for composition, a scatter plot for relationship, and a histogram for distribution. For each, write one sentence: "I chose this visual because..." That sentence is the skill — not the chart.
Take a default Power BI bar chart and apply this discipline: remove the legend if the axis already explains the categories, turn off gridlines that add no information, set a descriptive title that states the insight (not just "Sales by Region" but "East Region Leads in Q3 Revenue"), and use a single consistent color with one accent color for the highest value. Less visual noise means faster understanding.
Build a table visual with a sales column. Apply conditional formatting → Background color using a color scale from red (low) to green (high). Then apply a rule-based format that highlights any value below target in red. Show it to someone — if their eye goes immediately to the problem areas without you pointing, the formatting is working as intended.
Build a report page with three visuals and one date slicer. Click a bar in one visual — observe how the others respond. That is cross-filtering. Now go to Format → Edit Interactions and change one visual to not respond to the others. Understand why you would intentionally break cross-filtering for certain visuals — this is one of the most practical report design decisions you will make.
Create a second report page. In the Visualizations pane → Drill through field well, add a Category field. Go back to your summary page, right-click a category bar, and select Drill through. You should land on the detail page filtered to that category. This single feature transforms a static report into a self-service investigation tool.
Create two bookmarks: one showing a chart view, one showing a table view of the same data. Add two buttons labeled "Chart" and "Table." Assign each bookmark to its button using Action → Bookmark. Now you have a toggle. This technique can hide/show entire panels, simulate page navigation, and create report experiences that feel like applications — not spreadsheets.
Publish any report from Power BI Desktop using Home → Publish. In the Power BI Service, create a new Workspace and move your report there. Add one other person as a Viewer. Understand the four workspace roles: Admin controls everything, Member can publish, Contributor can edit, Viewer can only read. This access structure is how organizations govern their data at scale.
In the Power BI Service, go to your dataset's Settings → Scheduled Refresh. Configure a daily refresh. If your source requires an on-premises connection, research the On-Premises Data Gateway. A report that requires manual republishing every day is a report that will eventually show stale data to a decision-maker — scheduled refresh is what makes a report a living asset, not a static file.
Understand this distinction clearly: a Report is multi-page, interactive, built in Desktop — it is the analytical workspace. A Dashboard is a single canvas in the Service made by pinning tiles from reports — it is the executive summary. Build one of each from the same data. Pin your three most important KPI visuals to a Dashboard. Ask yourself: if a leader had 30 seconds, what would they need to see? That question defines what belongs on a Dashboard.
Future phase: All 27 terminal skills, user responses, and progress states will be stored in a MySQL database — enabling persistent progress tracking, personalized learning paths, and organizational skill reporting across the full E-Scholar platform.