Title: Correcting for attenuation due to measurement error

Abstract: I present a frequentist method for quantifying uncertainty when correcting correlations for attenuation due to measurement error. The method is conservative but has far better coverage properties than the methods currently used when sample sizes are small. I recommend the use of confidence curves in favor of confidence intervals when this method is used. I introduce the R package "attenuation" which can be used to calculate and visualize the methods described in this paper.