FAQ/RTM222015-10-15 14:23:39PeterWatson12015-10-15 14:20:54PeterWatsonComputing the regression to the mean effect?The below example uses equation (1) of Barnett, AG, van der Pols1, JC and Dobson, AJ (2005) 34 215-220 which uses the example below where only people with a score of 40 or below were sampled at baseline from a population with a mean of 60 and a standard deviation of 15. 1;
Rg[rhox+1]<-(sigma2_w[rhox+1]/m)/sqrt(sigma2_b[rhox+1]+(sigma2_w[rhox+1]/m))*czg; # RTM effect;
}
}
output<-cbind(sigma2_b,sigma2_w,rho,Rl,Rg)
print("The expected RTM effect for a range of baseline samples sizes and rhos")
print(output)
print("sigma2_b=between-subject variance, sigma2_w=within-subject variance")
print("rho=within-subject correlation, Rl=RTM effect (cut-off)");]]>