Is linear regression good enough for determining bridge balance in precision high resistance measurements?

The goodness of linear regression for determining the balance point of a bridge used in precision high resistance measurements is studied, and it is found that the technique may produce a large error. It is shown that, well-correlated data does not guarantee an accurate balance point to be obtained, and the accuracy depends on the sum of squared deviations of datapoints from their mean.