Course contentsShow
← Data Science
Lesson 386 of 2,14510. Hypothesis Tests for MeansFree lesson

Effect Size Interpretation Guidelines

Understanding small, medium, and large effect size conventions and their context-dependent nature.

Effect Size Interpretation Guidelines

You'll learn how to interpret Cohen's d values using standard conventions while understanding when these rules of thumb apply—and when they don't.

Understanding the Conventions

Now that you can calculate Cohen's d, you need to interpret what those numbers actually mean. Jacob Cohen proposed widely-used benchmarks:

  • Small effect: d ≈ 0.2
  • Medium effect: d ≈ 0.5
  • Large effect: d ≈ 0.8

These guidelines suggest that a Cohen's d of 0.5 means the two group means differ by half a standard deviation—a medium-sized difference.

Context is King

Here's the crucial part: these are rules of thumb, not absolute truths. What counts as "large" depends entirely on your field and question.

Example 1: In educational research, improving test scores by d = 0.3 might be considered substantial and meaningful, because changing human learning outcomes is inherently difficult.

Example 2: In pharmaceutical trials, a pain medication with d = 0.3 might be too weak to matter clinically, even if statistically significant. Patients need noticeable relief.

Example 3: In physics experiments measuring fundamental constants, even d = 0.1 could be enormous, suggesting a major discovery.

Think Practically, Not Just Statistically

Effect size interpretation should always involve:

  • Domain expertise: What's typical in your field?
  • Practical significance: Does this difference matter in the real world?
  • Cost-benefit analysis: Is the effect large enough to justify intervention costs?

A statistically significant result with d = 0.1 might pass your hypothesis test but be too small to care about practically.

Key Takeaway: Cohen's benchmarks (0.2 small, 0.5 medium, 0.8 large) provide starting points for interpretation, but always contextualize effect sizes within your specific domain—a "small" effect in one field may be groundbreaking in another.