Exploratory Data Analysis (EDA) in Power BI

Introduction to EDA in Power BI

Exploratory Data Analysis (EDA) is the process of analyzing datasets to summarize their key characteristics using visualization and statistical techniques. In Power BI, EDA helps in:
✅ Understanding data structure
✅ Identifying missing or incorrect values
✅ Finding relationships between variables
✅ Detecting outliers and trends

---

🔹 Step-by-Step Guide to Performing EDA in Power BI

Step 1: Load Data into Power BI

1. Open Power BI Desktop.

2. Click Home → Get Data → Choose a source (Excel, CSV, SQL, etc.).

3. Select your dataset and click Load.

4. Click Transform Data to open Power Query Editor for data cleaning.

---

Step 2: Check Data Types & Rename Columns

1. In Power Query Editor, review data types (Text, Number, Date, etc.).

2. Convert incorrect data types:

   • Right-click on a column → Change Type → Select the correct type.

3. Rename columns for clarity by right-clicking and selecting Rename.

4. Click Close & Apply to save changes.

---

Step 3: Handle Missing and Duplicate Values

1. Check for Null Values:

   • Click Transform → Replace Values (Replace nulls with mean/median/mode).

   • Use Remove Rows to eliminate missing data if necessary.

2. Remove Duplicates:

• Click Remove Duplicates under Transform Tab for unique records

Step 4: Summary Statistics & Data Distribution

1. Create a Table Visualization

   • Go to Visualizations → Table and drag all columns.

   • Check min, max, and average values for numerical data.

2. Use Card Visuals for Key Metrics

   • Total Students, Average Score, Highest Score, Pass Percentage

   • Add a Card visual and drag relevant fields.

3. View Statistical Summary

• Click Modeling → New Measure to calculate summary stats like:

       Write Dax Query

                    Avg_Score = AVERAGE(Data[Total Marks])

Use the Matrix visual to compare numerical columns.

Step 5: Correlation Analysis (Relationship Between Variables)

1. Create a Scatter Plot

   • X-axis: Individual Subject Score

   • Y-axis: Total Marks

   • This helps visualize the strength of relationships between subjects.

2. Use a Correlation Matrix (Custom Visual)

   • Install Correlation Plot from the Power BI Marketplace.

   • Drag numeric columns to generate a heatmap-style correlation matrix.

---

Step 6: Categorical Data Analysis

1. Create a Bar Chart for Categorical Data

   • X-axis: Categories (e.g., Grade, Pass/Fail, Gender)

   • Y-axis: Count of Students

   • Sort data in Descending Order for better visualization.

2. Create a Pie Chart for Grade Distribution

   • Drag Grade to the Legend.

   • Drag Count of Students to Values.

---

Step 7: Filtering & Segmentation

1. Add a Slicer to Filter Data

   • Drag Grade or Subject to the slicer.

   • This helps analyze subsets of data dynamically.

2. Use Drill-Through & Hierarchy Filters

   • Right-click on a visual → Select Drill-Through to see deeper insights.

---

Step 8: Trend Analysis Using Line Charts

1. Use a Line Chart to Analyze Performance Trends

   • X-axis: Time (e.g., Academic Years, Semesters)

   • Y-axis: Average Score / Pass Rate

   • Identify performance improvements or declines over time.

---

Step 9: Generate Insights & Export Report

1. Summarize key findings from EDA.

2. Format visuals for better readability.

3. Export Power BI report:

   • Click File → Export to PDF / Power BI Service to share insights.