In scientific research, statistical significance is often reported using p-values, but this metric alone can be misleading when it comes to interpreting the true importance of a finding. This is where effect size comes into play. Effect size provides a quantitative measure of the magnitude of the observed effect, offering a clearer picture of the practical significance of your research findings. In this article, we discuss why effect size is important, the different types of effect size measures, and how to calculate them.

What is Effect Size?

Effect size is a statistical concept that quantifies the magnitude of a phenomenon or difference between groups in a study. Unlike p-values, which only tell us whether an effect exists, effect size measures the strength or practical significance of the effect. Effect size is crucial because it helps researchers and practitioners understand whether a statistically significant result is also meaningful in real-world terms.

For example, a study might find that a new treatment significantly reduces symptoms in a group of patients, but the effect size will tell us how large that reduction is. A small effect size may suggest that while the treatment works, its impact is modest and may not be of practical importance, while a large effect size indicates that the treatment has a significant impact.

Why Use Effect Size?

1. Clarifies Practical Significance: While p-values help determine whether an effect exists, effect size measures the strength of that effect. A result with a small p-value could still have little practical significance if the effect size is small. For example, a p-value of 0.01 might indicate statistical significance, but if the effect size is tiny, it may not be relevant in practice.
2. Comparing Across Studies: Effect size allows researchers to compare results across different studies, even if they use different measurement tools or sample sizes. By standardizing the size of the effect, researchers can make more meaningful comparisons, regardless of the specific methods or scales used in the studies.
3. Understanding Variability: Effect size provides insight into the variability of an effect. Knowing the size of the effect, along with its confidence intervals, helps to understand how consistent and reliable the findings are.
4. Guiding Decision-Making: In applied fields like medicine or social sciences, effect size can guide decision-making. A large effect size in a clinical trial may suggest a treatment is not just statistically significant but also worthy of implementation in real-world settings.
5. Reproducibility: Reporting effect size makes research findings more transparent and reproducible. Instead of just stating whether a result is significant, researchers can share the size of the effect, allowing other scientists to understand the robustness and applicability of the results.

Types of Effect Size Measures

There are several common effect size measures, and their use depends on the type of study or analysis. The most frequently used ones include:

1. Cohen’s d: A measure of effect size

Cohen’s d is one of the most widely used effect size measures, particularly when comparing the means of two groups—such as a treatment group vs a control group. It quantifies the magnitude of the difference between the groups in standard deviation units, allowing for a standardized interpretation of effect size.

How to calculate Cohen’s d:

Cohen’s \(d = \frac{M_1 – M_2}{\sigma_p}\)

Where:

• M₁ and M₂ are the means of each group
• σ_p = pooled standard deviation (the square root of the weighted average of the variances of the two groups)

How to interpret Cohen’s d:

Relative size	Effect size	% of control group below the mean of experimental group
	0	50%
Small	0.2	58%
Medium	0.5	69%
Large	0.8	79%
	1.4	92%

Why use Cohen’s d?

By standardizing the difference between means, Cohen’s d enables meaningful comparisons across studies—even when the original measurements differ in scale.

2. Pearson’s r (for correlation)

Pearson’s r is one of the most commonly used measures of linear correlation between two continuous variables. It quantifies the strength and direction of a linear relationship, producing a value between -1 to +1, where 0 indicates no relationship, and -1 or +1 indicates a perfect negative or positive relationship, respectively.

How to calculate Pearson’s r:

$$r = \frac{cov(X, Y)}{\sigma_X \sigma_Y}$$

Where:

• cov(X,Y) = covariance between variables X and Y
• σ_X= standard deviation of X
• σ_Y = standard deviation of Y

How to interpret Pearson’s r:

• r = 0.1 (small effect)
• r = 0.3 (medium effect)
• r = 0.5 (large effect)

Why use Pearson’s r?

Pearson’s r allows you to assess the degree of linear association between two continuous variables.

3. Cohen’s f² (for regression models)

Cohen’s f² is an effect size measure used in the context of multiple regression. It quantifies how much a specific independent variable contributes to explaining variance in the dependent variable, helping you to assess the practical impact of that predictor.

How to calculate Cohen’s f²:

$$f^2 = \frac{R^2}{1 – R^2}$$

Where:

• R² is the proportion of variance in the dependent variable explained by the model

• 1− R² is the unexplained variance

Interpretation:

• f² = 0.02 (small effect)
• f² = 0.15 (medium effect)
• f² = 0.35 (large effect)

Why use Cohen’s f²?  

Cohen’s f² allows you to evaluate the incremental contribution of a variable within a regression model. It goes beyond statistical significance by providing a sense of practical importance, especially useful when comparing multiple predictors or assessing changes between models.

4. Odds Ratio:

The odds ratio is commonly used effect size measure in studies involving binary outcomes (e.g., success/failure), especially in case-control, cohort, and logistic regression analyses. It quantifies the strength of an association between an exposure (or treatment) and an outcome.

How to calculate the Odds Ratio:

$$Odds Ratio (OR) = \frac{(a/c)}{(b/d)}$$

Where:

• a and c are the number of people with the outcome (success)
• b and d are the number of people without the outcome (failure)
• The odds are the ratio of people with the outcome to those without it, within each group.

How to interpret the Odds Ratio:

Interpretation:

• OR = 1 indicates the event is equally likely in both groups.
• OR > 1 indicates the event is more likely to occur in the exposed (treatment) group.
• OR < 1 indicates the event is less likely to occur in the exposed (treatment) group.

Why use the Odds Ratio?

The odds ratio is particularly useful in:

• Case-control studies, where direct risk calculations are not possible
• Logistic regression models, where it helps estimate the impact of predictors
• Comparing relative likelihoods of outcomes between groups

It provides an intuitive understanding of how strongly an exposure is linked to an outcome.

---

How to Calculate Effect Size

The calculation of effect size varies depending on the type of study and the statistical test being performed. Here’s a general guide to calculating effect size:

1. Choose the appropriate effect size measure: Based on your research design, decide whether you need Cohen’s d, Pearson’s r, Cohen’s f², Odds Ratio, or another measure.
2. Gather the necessary data: This includes group means, standard deviations, correlation values, or regression coefficients, depending on the measure you are using.
3. Apply the formula: Use the formula specific to your chosen measure (e.g., Cohen’s d or Pearson’s r) to calculate the effect size.
4. Interpret the effect size: Once you calculate the effect size, use the commonly accepted thresholds (e.g., for Cohen’s d: small = 0.2, medium = 0.5, large = 0.8) to interpret the magnitude of the effect.
5. Report with confidence intervals: To account for uncertainty, it’s good practice to report effect size along with confidence intervals, which provide a range within which the true effect size is likely to fall.

Conclusion

Effect size is an essential measure for understanding the magnitude of observed effects in research. Unlike p-values, which only tell us if an effect exists, effect size provides a more meaningful understanding of the practical significance of the results. Whether you’re comparing two groups, examining a correlation, or evaluating the impact of a predictor in a regression model, calculating and reporting effect size helps provide context for the findings and enhances the reliability of research conclusions.

By incorporating effect size into your statistical analyses, you ensure that the results you report are not only statistically significant but also practically meaningful, offering valuable insights for decision-making and advancing knowledge in your field.