Effect Size
A measure of the magnitude or practical importance of a finding, independent of sample size.
Also known as: Magnitude of effect, Practical significance, Cohen's d
Category: Concepts
Tags: statistics, research, measurement, analysis, evidence
Explanation
Effect size measures the magnitude of a difference or relationship - how big an effect actually is, not just whether it exists. Unlike statistical significance (which depends heavily on sample size), effect size indicates practical importance. Common measures include: Cohen's d (standardized difference between means), correlation coefficient r, odds ratios, and percentage differences. Why effect size matters: with large enough samples, trivially small effects become 'significant'; effect size reveals if the effect is meaningful. Small effect (d=0.2): barely noticeable, may not be practically important. Medium effect (d=0.5): noticeable to careful observer. Large effect (d=0.8): obvious to casual observer. Best practices: always report effect size alongside significance, interpret in context (small effects can matter if they're easy to implement), and compare to other interventions in the field. Effect size thinking asks: 'how much?' rather than 'is there an effect?' For knowledge workers, understanding effect size helps: evaluate whether research findings matter practically, compare interventions fairly, and avoid being impressed by 'significant' findings that are actually tiny.
Related Concepts
← Back to all concepts