This document summarizes and analyzes different feedback methods for adjusting manufacturing processes when setup errors result in constant or random offsets from the target value. It considers two cases: when the setup error is an unknown constant offset, and when it is a random variable with unknown mean and variance. The performance of Grubbs' harmonic rule, an integral controller, and a Kalman filter approach are evaluated based on their average integrated square deviation over a number of parts produced. The analysis shows how the choice of initial offset estimate and its uncertainty affect the performance of the adjustment methods.