Session 28 · Phase 4: Statistics & Visualization

Power BI / Tableau — Data Import, Modeling & DAX

Build a self-service BI solution: import and clean data, model relationships, and write DAX measures that power interactive dashboards.

⏱ ~2 hrs 📚 Core content 🎯 High priority

Learning Objectives

Explain what Power BI is and its end-to-end workflow.
Import and clean data with Power Query.
Model relationships with a star schema (facts vs dimensions).
Explain what DAX is and write basic measures.
Distinguish a Measure from a Calculated Column.
Use CALCULATE() and explain row vs filter context.

1. What Power BI Is

Power BI is a business-intelligence tool that turns data into interactive reports and dashboards. The workflow is always the same: Get data → Transform → Model → Visualize → Share.

Get data — connect to sources (Excel, SQL, CSV, web).
Transform — clean and shape with Power Query.
Model — define relationships between tables.
Visualize — build charts and dashboards.
Share — publish to a report server or cloud.

2. Data Import & Cleaning — Power Query

Power Query is the "get and transform" engine. It's where you clean data before it enters the model — the same cleaning concepts from Excel/SQL, but recorded as reusable steps.

Remove duplicates and filter rows.
Split or merge columns.
Replace values and change data types.
Handle missing values.

📝

Note: every step you apply in Power Query is recorded, so it re-runs automatically when the data refreshes — this "repeatable cleaning" is a key selling point in interviews.

3. Data Modeling — the Star Schema

A star schema organizes data into a central fact table (the numbers/events) surrounded by dimension tables (the descriptive categories), connected by relationships.

Table type	Contains	Example
Fact	measures, keys, transactions	Sales (amount, date, product key)
Dimension	descriptive attributes	Product, Customer, Region, Date

Relationships link the fact table to each dimension — this is what lets one dashboard filter across everything.

4. What Is DAX?

DAX (Data Analysis Expressions) is Power BI's formula language — used to create measures and calculated columns. It's like Excel formulas, but designed to work across the whole data model.

Total Sales = SUM(Sales[Amount])

5. Measures vs Calculated Columns — the #1 DAX Question

	Measure	Calculated Column
Stored	not stored — computed on the fly	stored, one value per row
When evaluated	at query/visual time (dynamic)	at data refresh (static)
Best for	aggregations (SUM, AVG, ratios)	row-level flags/categories
Example	`Total Sales = SUM(Sales[Amount])`	`Profit = Sales[Amount] - Sales[Cost]`

💡

Rule of thumb: use a measure for anything you aggregate; use a calculated column for a row-by-row value you need to filter or categorize on. Measures are more memory-efficient.

6. CALCULATE() and Context

CALCULATE() is the most important DAX function — it changes the filter context in which an expression is evaluated.

North Sales = CALCULATE(SUM(Sales[Amount]), Region[Name] = "North")

Row context vs filter context

Row context — "this specific row" (used in calculated columns).
Filter context — "the current set of filters" from slicers/visuals (used in measures).

CALCULATE is powerful precisely because it can override or add to the filter context.

7. Power BI vs Tableau

	Power BI	Tableau
Formula language	DAX	calculated fields (own syntax)
Best known for	tight Excel/ecosystem integration, cost	rich, flexible visualizations
Learning curve	moderate	easy to start

The concepts transfer — both do import → model → visualize. Know one well and you can pick up the other.

📋 Stable content — Reviewed: August 2026

8. Interview Questions (with Model Answers)

The Power BI/DAX questions interviewers ask. Self-test before revealing.

IQ1. What is Power BI, and what's the typical workflow?

Model answer: "Power BI is a BI tool for building interactive reports and dashboards. The workflow is get data, transform with Power Query, model relationships, visualize, and share."

IQ2. What is Power Query, and what's it used for?

Model answer: "Power Query is the data-cleaning engine. I use it to remove duplicates, split columns, replace values, and change types — and every step is recorded so it re-runs on refresh."

IQ3. What is DAX?

Model answer: "DAX (Data Analysis Expressions) is Power BI's formula language. I use it to write measures and calculated columns that compute across the data model."

IQ4. What's the difference between a calculated column and a measure?

Model answer: "A calculated column is stored row by row at refresh time; a measure is computed dynamically at query time and isn't stored. I use measures for aggregations and columns for row-level values."

IQ5. What does CALCULATE() do?

Model answer: "CALCULATE evaluates an expression under a modified filter context. For example, CALCULATE(SUM(Sales[Amount]), Region[Name] = 'North') sums only North sales regardless of other filters."

IQ6. What's the difference between row context and filter context?

Model answer: "Row context refers to a single row (used in calculated columns); filter context is the set of active filters from slicers and visuals (used in measures). CALCULATE modifies the filter context."

IQ7. What is a star schema?

Model answer: "A star schema is a central fact table (transactions/measures) surrounded by dimension tables (categories like product, customer, date), linked by relationships. It's the standard, efficient layout."

IQ8. What's the difference between Power BI and Tableau?

Model answer: "Both do import → model → visualize. Power BI uses DAX and integrates tightly with Excel; Tableau is known for richer, more flexible visualizations and is easier to start with. The core concepts transfer."

IQ9. Which is more memory-efficient — a measure or a calculated column?

Model answer: "A measure, because it isn't stored — it's computed on the fly. A calculated column is stored for every row, so it uses more memory. That's one reason to prefer measures for aggregations."

Hands-On Project: Build a Mini Model

Describe (or build, if you have Power BI installed) a small sales model.

Steps

Import two tables: Sales (fact) and Region (dimension).
In Power Query, remove duplicates and set correct data types.
Create a relationship between Sales and Region on the region key.
Write a measure Total Sales = SUM(Sales[Amount]).
Write a measure North Sales = CALCULATE(SUM(Sales[Amount]), Region[Name] = "North").
Build a bar chart of Total Sales by region.

View Solution / Walkthrough

Get Data → import Sales and Region.
Transform Data → Remove Duplicates → change Amount to decimal, Date to date.
Model view → drag the region key from Sales to Region to create the relationship.
New Measure → Total Sales = SUM(Sales[Amount]).
New Measure → North Sales = CALCULATE(SUM(Sales[Amount]), Region[Name] = "North").
Bar chart → Axis = Region[Name], Values = Total Sales.

Key Takeaways

Power BI workflow: get data → transform → model → visualize → share.

Power Query records your cleaning steps so they re-run on refresh.

Star schema = central fact table + surrounding dimension tables.

Measure = dynamic, not stored; calculated column = stored per row.

CALCULATE() modifies the filter context — the heart of DAX.

Learning Objectives

1. What Power BI Is

2. Data Import & Cleaning — Power Query

3. Data Modeling — the Star Schema

4. What Is DAX?

5. Measures vs Calculated Columns — the #1 DAX Question

6. CALCULATE() and Context

Row context vs filter context

7. Power BI vs Tableau

8. Interview Questions (with Model Answers)

Hands-On Project: Build a Mini Model

Steps

Key Takeaways

Objective Questions — Test Your Understanding