KembaraXtra – Psychology: Cohen’s d

Definition

Cohen’s d is a standardized measure of effect size used to express the difference between two means in units of standard deviation. In its common form, it is calculated by subtracting one group mean from the other and dividing the result by a pooled estimate of standard deviation. Because the result is standardized, it allows differences measured on different scales to be compared more meaningfully.

Statistical Mechanism

A common formula is d = (M_1 - M_2)/s_p, where M_1 and M_2 are the two sample means and s_p is the pooled standard deviation. The sign indicates the direction of the difference, while the absolute value indicates its magnitude. Jacob Cohen suggested conventional benchmarks of approximately 0.20 for a small effect, 0.50 for a medium effect, and 0.80 for a large effect, although these should not be applied mechanically.

Role in Psychological Research

Cohen’s d is widely used in experimental, clinical, educational, and social psychology. It complements statistical significance testing by indicating how large a difference is rather than merely whether it is unlikely under a null hypothesis. It is especially useful in meta-analysis, where results from different studies must be placed on a common scale. Reporting confidence intervals around d improves interpretation by showing uncertainty.

Limitations and Interpretation

The meaning of a particular value depends on the research context, measurement reliability, population, and practical consequences. A statistically small effect may be important when applied to many people, while a large effect may be of limited importance if the outcome is trivial. Cohen’s d can also be biased in small samples, in which case a corrected measure such as Hedges’ g may be preferable.

Summary

Cohen’s d expresses the difference between two means in standard-deviation units. It provides a standardized estimate of effect magnitude and is widely used alongside significance tests. Conventional benchmarks offer rough guidance but should not replace contextual interpretation. Confidence intervals and study quality remain essential.