Chi-Square Calculation Step-by-Step
About This MicroSim
Practice calculating chi-square statistics step by step! This visualization breaks down the formula into its component parts with visual feedback showing how each category contributes to the final statistic.
How to Use
- Click Next and Back buttons to move through calculation stages
- Toggle Auto to automatically step through the calculation
- Click on any observed count in the table to edit it
- Use Preset buttons to load example datasets (Candy, Dice)
- Color bars show which categories contribute most to the chi-square value
The Chi-Square Formula
The chi-square statistic measures how far observed counts are from expected counts:
Where: - O = observed count for each category - E = expected count for each category - The sum is taken over all categories
Key Insights
- Positive and negative differences both matter - squaring removes the sign
- Larger departures from expected contribute more to the chi-square
- Dividing by E standardizes - a difference of 5 matters more when E is small
- The largest contributor shows where data differs most from expectations
Lesson Plan
Learning Objective
Students will practice calculating chi-square statistics by working through each component of the formula with visual feedback (Bloom's Taxonomy: Applying).
Warmup Activity (3 minutes)
Load the Candy preset. Ask: "Looking at the bar chart, which category seems to have the biggest difference between observed and expected?" Then step through to verify.
Main Activity (12 minutes)
- Step through the Candy example one stage at a time
- At each stage, have students predict the next value before revealing
- Discuss: Which category contributes most? Why?
- Load the Dice example and repeat
Discussion Questions
- Why do we square the differences?
- Why do we divide by the expected count?
- If two categories have the same absolute difference (O - E), which contributes more - the one with larger E or smaller E?
Practice Problems
Have students edit the observed values to create: 1. A case with chi-square very close to 0 (all O close to E) 2. A case with a very large chi-square (extreme departures)