Good Science with Structural Equation Models: The Case of Non-Financial Performance Measurement Systems
Category : Paper Series EN
Issue No. 39 (July – August 2006)
Dr. Pavlos A. Vlachos, and Dr. Adam P. Vrechopoulos
Performance measurement systems play a major role in evaluating the accomplishment of organizational objectives and compensating managers and front-office employees. Traditionally, performance measurement systems emphasize financial metrics such as earnings and accounting returns, giving much less attention to non-financial metrics like customer and employee value, satisfaction and most importantly trust. Recently, Wharton Accounting Professors’ Ittner and Larcker (2003) suggest that financial data have limitations as a measure of company performance, further arguing that non-financial measures may be better suited for measuring performance but are difficult to understand. In this article, we first discuss the notion of non-financial performance measurement systems and then discuss the details of an analytical methodology that enables their application, namely structural equation models with latent variables. Concluding we present the case of an investment bank that wants to incorporate the notion of social capital in the process of evaluating potential investment projects, and discuss structural equation methodologies for doing so.