Accuracy of Relative Weights on Multiple Leading Performance Measures: Effects on Managerial Performance and Knowledge

Abstract

Many firms that use multiple lead measures in their performance measurement systems do not validate the causal model linking these measures to future financial outcomes, and the cause-and-effect relationships in the model are often left to subjective estimates that may be prone to errors. Using an experiment, this study examines how the accuracy of assumptions about the relative importance of lead measures in a causal model affects managerial knowledge and performance when managers are given the opportunity to learn over multiple periods. The results show that having inaccurate relative weights on lead measures improves knowledge and performance, relative to not having any weights. Furthermore, knowledge and performance are similar under accurate versus inaccurate relative weights, providing no support for the biasing effects of inaccurate relative weights. The findings suggest that, at least under certain circumstances, managers benefit even if they are given inaccurate relative weights on lead measures, and they are able to correct those inaccuracies to reach a comparable level of knowledge and performance as if they had been given accurate relative weights.

Keywords: causal model, lead measures, relative weights, knowledge

JEL Classification: D83, M1, M41, M52