Effect size interpreter

An R function that helps in the interpretation of statistical effect sizes such as correlation, eta-squared (ANOVA) or Cohen's coefficients (w or d).

Usage

You must invoke effectSizeInterpreter function with two arguments:

• e: a number with your effect size.
• type: the effect size type: 'correlation', 'etaSquared', 'oddsRatios', 'd' (or 'tTest'), 'w' (or 'chiSquare').

> effectSizeInterpreter(0.97, type="correlation")
[1] "Nearly perfect"
> effectSizeInterpreter(0.90, type="correlation")
[1] "Very large"
> effectSizeInterpreter(0.56, type="correlation")
[1] "Large"

> effectSizeInterpreter(0.56, type="etaSquared")
[1] "Large"
> effectSizeInterpreter(0.12, type="etaSquared")
[1] "Moderate"
> effectSizeInterpreter(0.01, type="etaSquared")
[1] "Small"

References

The following references have been used to construct the scales used in this function:

• http://www.sportsci.org/resource/stats/effectmag.html
• http://imaging.mrc-cbu.cam.ac.uk/statswiki/FAQ/effectSize