A psychometric assessment of an instrument to measure a service firm's customer-based corporate reputation

Abstract

Over the years, marketers have been cautioned repeatedly about the use of measuring instruments that do not demonstrate sufficient evidence of reliability and validity, particularly construct validity. The use of marketing instruments that do not demonstrate sufficient evidence of construct validity can lead to invalid results and erroneous conclusions. In short, construct validity is not just a "nice to have" in marketing research - it is at the heart of marketing decision-making. This study assesses the psychometric properties of a recently published instrument by Walsh and Beatty (2007) to measure the customer-based corporate reputation of a service firm, and illustrates how, by using an approach proposed by Steenkamp and Trijp (1991) to assess the uni-dimensionality of the underlying dimensions of a multi-dimensional construct, one can enhance the construct validity of such an instrument. The result is a shortened, 17-item instrument with superior psychometric properties compared to the longer 31-item instrument.