Procedures for Revising Management Judgments Forecasts

Abstract

Forecast improvement is often approached by attempt ing to find the "best" model for a given situation. Less attention has been paid to the possibility of examining past prediction errors for patterns that may suggest forecast adjustments for the future. This empirical study involves one firm's management judgment forecasts for product sales and the attempts made to improve their accuracy by removing certain types of bias. In three of the five series examined, error reduction averaging close to thirty percent occurs as a result of the adjustment procedure. The other two series proved to be relatively free from bias and were therefore not in need of the correction method described.