Calculate Kappa-Squared Mediation Effect Size

by Ian Rothmann Jr.

Kappa-squared: Mediation Effect Size

This page provides an implementation of the k2 effect size as devised in:
Preacher, K. J. & Kelley, K. (2011). Effect sizes measures for mediation models: Quantitative strategies for communicating indirect effects. Psychological Methods, 16(2), 93-115.

Although R tools exist, this tool will help you calculate k2 from any package and would thus help you to calculate it for latent variables. If you are using Mplus, you can get the required variances and covariances of latent variables from the TECH4 output.

Data entry

Please enter variances, covariances and unstandardised regression coefficients below. X represents your independent variable, M represents your proposed mediator and Y represents your outcome variable.

Variance of X
Variance of M
Variance of Y
Covariance of XY
Covariance of XM
Covariance of MY
Beta of X on M (a)
Beta of M on Y (b)

Results

k2=0*
M(a)=0.2	a between {-1,2}
M(b)=0.2	b between {-1,2}
M(ab)=0.2	ab between {-1,2}

* Preacher and Kelley (2011) suggest that it makes sense to interpret k2 values in the same light as the coefficient of determination (R2), i.e. with the guidelines of Cohen (1988), where small, medium, and large effect sizes are stated as 0.01, 0.09, and 0.25 respectively.

Please let us know who you are!

We are planning an online nifty statistics studio that will help you calculate all kinds of useful statistics like this one!

We are also planning to launch the brand new myresearchsurvey.com online questionnaire builder in the near future, complete with statistical tools for preliminary analyses during data gathering.

We would like to notify you of these important events, and promise to not abuse your email address!