Skip to main content
. 2020 Oct 15;10:17372. doi: 10.1038/s41598-020-74057-1

Figure 1.

Figure 1

Illustration of how the variance in estimating X is inversely related to the maximum measurable Pearson correlation coefficient ρ(X,Y)=cov(X,Y)/σXσY1115, while the bias in estimating X does not affect the correlation. Therefore, one should seek to find a minimal variance estimator of X. Note that in the upper panel points were individually shifted in the X dimension to modify the variance of X while points in the lower panel were shifted by the same amount and direction in the X dimension to modify bias. Points have the same Y value across all plots.