1 post / 0 new
avsharda's picture


I was wondering if someone could guide me to estimate errors-in-variables, or EIV, correlation coefficient. I have the reliability coefficient (k), which I got from the formula,

k = 1 - Var(error)/[1.414 * Var (X)]
where Var(error) is the variance of individual repeated measurements (estimated from precision study). I have data with errors in measurement of the dependent variable as well. So I wish to get rid of the attenuation of correlation (just by Pearson correlation) by correcting for the errors-in-variables. Can someone please guide me to the next step. Thanks.