Course contentsShow
← Data Science
Lesson 651 of 2,14517. Categorical Predictors and InteractionsFree lesson

Interpreting Continuous × Continuous Interactions

Understand how to interpret coefficients when both predictors in an interaction are continuous variables.

Interpreting Continuous × Continuous Interactions

What You'll Learn

You'll learn how to read and explain coefficients when two continuous variables interact, meaning the effect of one depends on the level of the other.

The Core Idea

When you include an interaction between two continuous predictors (like height * weight), the coefficient for each main effect no longer represents a fixed, universal slope. Instead, the effect of one variable changes depending on the value of the other.

Consider a model predicting energy expenditure:

energy = β₀ + β₁·height + β₂·weight + β₃·(height × weight)

Here's how to interpret each coefficient:

  • β₁ (height coefficient): The effect of height when weight = 0 (often not meaningful in practice)
  • β₂ (weight coefficient): The effect of weight when height = 0
  • β₃ (interaction coefficient): How much the effect of height changes for each one-unit increase in weight (or vice versa)

The Key Insight

The actual slope for height isn't just β₁—it's β₁ + β₃·weight. As weight increases, the effect of height on energy changes by β₃ units. Similarly, the slope for weight is β₂ + β₃·height.

Practical Example

If β₃ = 0.5, it means: for every additional kilogram of weight, the effect of height increases by 0.5 units. Taller people might benefit more from additional weight in terms of energy expenditure.

This is like saying "the value of studying extra hours depends on how much sleep you get"—neither factor works in isolation.

Key Takeaway: In continuous × continuous interactions, main effect coefficients represent effects at zero values of the other predictor, while the interaction coefficient tells you how rapidly one predictor's effect changes as the other predictor increases.