Figure 1
From: The impact of measurement errors in the identification of regulatory networks
P-value distribution under the null hypothesis ( β 1 = 0) in independent data with standard deviation of the measurement error equal to 0.6 and sample size equal to 400 (model described in Simulations section, simulation I). This simulation was performed 10,000 times. (A) Standard Ordinary Least Squares (non uniform distribution); (B) Corrected Ordinary Least Squares (uniform distribution).