pgen.1007081.g003.tif (284.59 kB)
Outcomes were simulated to be unrelated to the exposure (bottom plot, showing false positive rates on the y-axis) or causally influenced by the exposure (top plot, showing true positive rates on the y-axis) with varying degrees of measurement imprecision applied to the exposure variable (x axis).
figure
posted on 2017-11-17, 18:35 authored by Gibran Hemani, Kate Tilling, George Davey SmithResults for MR and CIT were compared for varying sample sizes (columns of boxes).