Error may be introduced into an analysis by both the biosensor system used to perform the analysis and by errors in the output signal measured by the measurement device of the biosensor. For a reference sample, system error may be determined through the determination of relative error. However, during an analysis of a test sample with the measurement device of the biosensor system, true relative error cannot be known. A pseudo-reference concentration determined during the analysis may be used as a substitute for true relative error. The closer the analysis-determined pseudo-reference analyte concentration is to the reference analyte concentration of the test sample, the more accurate and/or precise the analyte concentration determined by the measurement device using an anchor parameter during compensation. The present invention provides an improvement in the accuracy and/or precision of the analysis determined pseudo-reference concentration through progressive approximation.