Does the ICC require that my data are normally distributed? If more information about my data is needed, just ask.
Anyone can suggest me a nonparametric version of the intraclass correlation coefficient (ICC)? My model have the following details: two-way random effects model, consistency, average ratings as unit. ...