Easily determine the effect size (Cohen's f²) for your multiple regression models. Input your full and reduced model R-squared values to quantify the practical significance of adding predictors to a statistical model. This crucial tool helps researchers and statisticians interpret the magnitude of an effect beyond just statistical significance.
Formula:
Cohen's f² Effect Size for Multiple Regression
The formula calculates the effect size (f²) when assessing the contribution of a set of predictors (or a single predictor) to a multiple regression model by comparing a full model to a reduced model.
f² = (R²full - R²reduced) / (1 - R²full)
Where:
R²full = The R-squared value of the full (larger) regression model, including all predictors of interest.
R²reduced = The R-squared value of the reduced (smaller) regression model, which excludes the specific predictors whose effect size you want to calculate.
Interpretation of f² (Cohen, 1988):
- f² ≈ 0.02: Small effect size
- f² ≈ 0.15: Medium effect size
- f² ≈ 0.35: Large effect size