Effect Size Calculator for ANOVA (Eta Squared)

A simple, powerful tool to calculate a measure of effect size using ANOVA results. Instantly find Eta Squared (η²) to understand the practical significance of your findings.

What is a Measure of Effect Size in ANOVA?

When you perform an Analysis of Variance (ANOVA), the p-value tells you whether the difference between your group means is statistically significant. However, it doesn’t tell you the size or magnitude of that difference. That’s where effect size comes in. An effect size is a quantitative measure of the magnitude of a phenomenon. For ANOVA, a common and intuitive effect size is Eta Squared (η²).

Eta Squared represents the proportion of the total variance in the dependent variable that is associated with or explained by the independent variable(s). In simpler terms, it tells you how much of the “story” your model explains. A value of 0.15, for example, means that 15% of the variance in your outcome can be attributed to your experimental manipulation or grouping factor. This is crucial for understanding the practical significance of your results beyond just statistical probability. To properly interpret your data, you need to be able to calculate a measure of effect size using ANOVA outputs.

The Formula for Eta Squared (η²): A Key ANOVA Effect Size

The formula to calculate Eta Squared is straightforward and uses values directly from your ANOVA output table. It’s the ratio of the variance explained by the model to the total variance in the data.

η² = SS_between / SS_total

Where SS_total = SS_between + SS_within. This makes the full formula:

η² = SS_between / (SS_between + SS_within)

Description of Variables for the Eta Squared Formula

Variable	Meaning	Unit	Typical Range
η²	Eta Squared, the effect size.	Unitless ratio	0 to 1
SSbetween	Sum of Squares Between Groups (also called SSeffect or SSmodel). Represents the variance explained by the independent variable.	Squared units of the dependent variable	0 to +∞
SSwithin	Sum of Squares Within Groups (also called SSerror or SSresidual). Represents the unexplained or error variance.	Squared units of the dependent variable	0 to +∞
SStotal	Total Sum of Squares. Represents the total variance in the data.	Squared units of the dependent variable	0 to +∞

For more complex models with multiple factors, researchers often use Partial Eta Squared. For a one-way ANOVA, Eta Squared and Partial Eta Squared are identical.

Practical Examples of Calculating Effect Size

Example 1: Educational Intervention Study

A researcher tests three different teaching methods (Group A, B, C) on student exam scores. After running a one-way ANOVA, they get the following output:

• Inputs:
  • Sum of Squares Between Groups (SS_between): 1500
  • Sum of Squares Within Groups (SS_within): 4500
• Calculation:
  • SS_total = 1500 + 4500 = 6000
  • η² = 1500 / 6000 = 0.25
• Result: An Eta Squared of 0.25. This is considered a large effect size and means that 25% of the variance in exam scores can be explained by the different teaching methods.

Example 2: Pharmaceutical Drug Trial

A study compares the effectiveness of a new drug versus a placebo for reducing blood pressure. The ANOVA results show a small, but statistically significant, effect.

• Inputs:
  • Sum of Squares Between Groups (SS_between): 80
  • Sum of Squares Within Groups (SS_within): 1200
• Calculation:
  • SS_total = 80 + 1200 = 1280
  • η² = 80 / 1280 = 0.0625
• Result: An Eta Squared of 0.0625. This is a medium effect size. While the drug works, it only accounts for 6.25% of the total variance in blood pressure changes, suggesting other factors are also highly influential. Check out our guide on interpreting statistical significance to learn more.

How to Use This ANOVA Effect Size Calculator

This calculator is designed for simplicity and accuracy. Follow these steps to calculate a measure of effect size using your ANOVA data:

1. Locate Your ANOVA Output: Open the results table from your statistical software (like SPSS, R, or JASP).
2. Enter SS_between: Find the “Sum of Squares” value for your independent variable (often labeled “Between Groups” or with the variable’s name) and enter it into the first field.
3. Enter SS_within: Find the “Sum of Squares” value for the error term (often labeled “Within Groups,” “Error,” or “Residual”) and enter it into the second field.
4. Interpret the Results: The calculator instantly provides the Eta Squared (η²) value. Use the interpretation guide (Small, Medium, Large) to understand its practical significance. The results also show intermediate values like Total Sum of Squares and Cohen’s f for a more complete picture. The chart provides a visual breakdown of the variance.

Key Factors That Affect Effect Size in ANOVA

Several factors can influence the calculated effect size. Understanding them helps in designing better experiments and interpreting results accurately.

• Magnitude of Group Differences: The larger the actual difference between the means of the groups you are comparing, the larger the SS_between will be, leading to a larger effect size.
• Within-Group Variability: If the data points within each group are tightly clustered around their mean (low variance), the SS_within will be small. A smaller error term results in a larger effect size. High variability or “noise” can obscure a real effect.
• Sample Size: While effect size is less sensitive to sample size than p-values, very small samples can produce unstable and biased estimates. Specifically, Eta Squared is known to be a positively biased estimator, especially in small samples.
• Number of Groups: In a one-way ANOVA, adding more groups can change the SS_between and affect the overall effect size calculation.
• Measurement Error: Imprecise or unreliable measurement of the dependent variable increases the within-group variance (SS_within), which in turn reduces the calculated effect size.
• Confounding Variables: If another variable that influences the outcome is not controlled for, it can increase the error variance and decrease the effect size of the variable of interest. You can learn more about controlling variables in our article on advanced experimental design.

Frequently Asked Questions (FAQ) about ANOVA Effect Size

What is considered a small, medium, or large effect size for Eta Squared?

According to common guidelines (originally from Cohen), the interpretation is typically: Small effect: η² ≈ 0.01; Medium effect: η² ≈ 0.06; Large effect: η² ≥ 0.14. These are not strict rules but helpful benchmarks.

What is the difference between Eta Squared (η²) and Partial Eta Squared (η²p)?

In a one-way ANOVA (one independent variable), they are identical. In a factorial ANOVA (with multiple independent variables), Partial Eta Squared calculates the variance explained by one factor while controlling for the other factors. Eta Squared calculates the proportion of the total variance explained by that one factor.

Can Eta Squared be negative?

No. Since Sums of Squares are always non-negative (they are sums of squared values), Eta Squared will always range from 0 to 1.

How does effect size differ from a p-value?

A p-value tells you the probability that you would observe your data if there were no real effect (the null hypothesis was true). It measures statistical significance. Effect size, on the other hand, measures the magnitude or practical importance of the effect, regardless of its statistical significance. A tiny, unimportant effect can be statistically significant with a large enough sample size.

Where do I find the Sum of Squares values?

You find them in the standard output table generated by statistical software when you run an ANOVA. Look for a column labeled “Sum of Squares” or “SS” and rows corresponding to your effect (“Between”) and the error (“Within” or “Residual”).

Is Eta Squared a biased measure?

Yes, Eta Squared is known to be an upwardly biased estimator of the population effect size, meaning it tends to overestimate the true effect, especially with smaller sample sizes. Alternative measures like Omega Squared (ω²) are less biased.

What is Omega Squared (ω²)? Should I use it instead?

Omega Squared (ω²) is another effect size measure that corrects for the bias found in Eta Squared. It provides a more conservative and often more accurate estimate of the population effect size. While Eta Squared is more common, many researchers recommend reporting Omega Squared, especially for smaller studies. Explore our Omega Squared calculator for more details.

Can I calculate an effect size from an F-statistic and degrees of freedom?

Yes, you can. The formula is η² = (F * df_between) / ((F * df_between) + df_within). Our calculator uses the Sum of Squares as it is a more direct and common method for this particular effect size.