Skip to content

Correlation Properties Explorer

Run the Correlation Properties Explorer Fullscreen

Embedding This MicroSim

Place the following line in your website to include this MicroSim in your course:

1
<iframe src="https://dmccreary.github.io/statistics-course/sims/correlation-properties/main.html" height="502px" scrolling="no"></iframe>

About This MicroSim

The Correlation Properties Explorer helps students develop intuition for how the correlation coefficient \( r \) behaves under various transformations and conditions. Rather than memorizing properties, students discover them through direct manipulation and observation.

The scatterplot starts with 8 points showing a moderate positive correlation. Students can drag any point and watch the r-value update in real time. Points are color-coded based on their contribution to the correlation: green for points that strengthen the linear pattern, red for points that weaken it, and blue for neutral contributions.

Key Properties Demonstrated

Property How to Explore
r is bounded (-1 to 1) Try dragging points to extreme positions; r never exceeds this range
r is unitless Use the unit buttons to scale or shift data; r stays exactly the same
r is symmetric Click "Swap Axes" to exchange x and y; r remains unchanged
r is sensitive to outliers Toggle "Add Outlier" to see dramatic r-value changes
Points contribute differently Color gradient shows which points most affect the correlation

Interactive Controls

Control Function
Drag points Move any point to see real-time r-value updates
Add Outlier Toggle an extreme point that dramatically affects correlation
Swap Axes Exchange x and y to demonstrate symmetry property
Unit buttons Apply transformations (scale x2, shift +50, etc.) showing r is unitless
Reset Points Return to the original 8-point dataset
Add Random Pt Grow the dataset with a randomly placed point

Visual Features

  • Color-coded points: Green = strengthens correlation, Red = weakens it, Blue = neutral
  • Real-time r display: Large, prominently displayed correlation coefficient
  • Before/After comparison: See how your last action changed r
  • Strength interpretation: Text description of correlation strength
  • Properties reminder panel: Quick reference for key correlation properties

Lesson Plan

Learning Objective

Students will investigate how changing data affects the correlation coefficient and develop intuition for the mathematical properties of correlation.

Bloom's Taxonomy Level

Analyze (Level 4) - Students analyze how individual data points influence the overall correlation and predict the effects of various transformations.

Prerequisites

  • Understanding of scatterplots and what they represent
  • Basic concept of positive and negative relationships
  • Introduction to the correlation coefficient \( r \)

Suggested Activities

Discovery Phase (15 minutes)

Have students work individually or in pairs to explore each property:

  1. Outlier Impact Challenge
  2. Start with the default dataset and note the r-value
  3. Toggle the outlier on and record the new r-value
  4. Drag the outlier to different positions - can you make r become negative?

  5. Question: "Why does one point have such a big effect?"

  6. Unit Invariance Discovery

  7. Click through each unit button while watching r
  8. Observation: What happens to the axis labels? What happens to r?

  9. Question: "If we measured height in inches vs. centimeters, would the correlation between height and weight change?"

  10. Symmetry Verification

  11. Note the current r-value
  12. Click "Swap Axes" and observe
  13. Question: "Does it matter which variable goes on which axis when calculating r?"

Structured Investigation (10 minutes)

Guide the whole class through these scenarios:

  1. Start with Reset Points
  2. Systematically drag points from the edges toward the center
  3. Predict: "Will r increase or decrease?"
  4. Verify and discuss why

Reflection and Connection (10 minutes)

Discussion questions:

  • "If correlation is unitless, what does that mean for comparing studies done in different countries with different measurement systems?"
  • "A researcher calculates r = 1.15. What can you immediately conclude?"
  • "Why is it important to plot your data before calculating r?"

Assessment Ideas

Formative

  • Ask students to predict what will happen before each manipulation, then verify
  • Have students explain why the outlier affects r so dramatically

Summative

  • Present scenarios and ask students to predict the effect on r (without using the simulation)
  • Given a correlation value and a described transformation, predict the new r

Differentiation

  • Struggling students: Focus on the outlier toggle first; the dramatic change makes the concept concrete
  • Advanced students: Challenge them to create a dataset where adding a point increases r, then another where it decreases r
  • Visual learners: Emphasize the color-coding of point contributions

Common Misconceptions to Address

  1. "Changing units should change r" - Use the unit buttons to show this is false
  2. "Outliers always decrease r" - Show that a well-placed outlier can actually increase r
  3. "The point furthest from others has the biggest effect" - Demonstrate that position relative to the pattern matters more than distance from other points

Connections to Standards

This MicroSim supports learning objectives related to:

  • Calculating and interpreting the correlation coefficient
  • Understanding properties of correlation
  • Recognizing the effect of outliers on statistical measures
  • Distinguishing between correlation and causation (discussed in accompanying text)

References

Note: Remember to capture a screenshot of this MicroSim and save it as correlation-properties.png in this folder for social media previews.