Content area
Full Text
ABSTRACT
The construct of reputation has attracted much attention among researchers and practitioners alike. The growing body of literature in this area has led to a wide variety of measurement approaches, albeit most publications do not point out the epistemic nature of reputation as a construct. It is not clear (1) if reputation is a formative or reflective construct and (2) how it should be conceptualized using a formative approach. The incorrect specification of the construct would produce misleading results, which in turn would provide no solid basis for reputation management. Based on a study among German consumers, this paper conceptualizes reputation as a formative construct consisting of ten indicators. The findings suggest that this conceptualization is a solid method on which to build reputation measures, which in turn are an important step for efficient reputation management.
KEYWORDS: formative indicators, reflective indicators, reputation measurement, reputation scales, structural equation modelling
INTRODUCTION
There has been a resurgence of interest in the reputation construct among researchers and practitioners (Fombrun, 2003; Dowling, 2001). The growing body of literature has led to an abundance of different definitions of corporate reputation. Gotsi and Wilson (2001) for instance highlight numerous viewpoints (Mahon, 2002; Wartick, 2002). They conclude that corporate reputation is 'a stakeholder's overall evaluation of a company over time' (Gotsi and Wilson, 2001: 29).
From an empirical point of view, corporate reputation has remained a rather controversial issue. For instance, Nguyen and Leblanc (2001) diagnose a lack of consensus on valid scales for reputation; they therefore propose to use direct measures. Unfortunately, this docs not lead to practical insights for reputation management because the sources of a good or bad reputation do not become evident. Effective measures are still considered the biggest hurdles for reputation management (Larkin, 2003).
In addition to the lack of consensus on the dimensions of reputation, recent publications challenge the common approach of specifying complex constructs such as reputation in general (Rossiter, 2002; Jarvis et al., 2003). Given the developments increasing the usability of structural equatuion modelling in the social sciences (Bollen, 1989; Baumgartner and Homburg, 1996), the relevance of this discussion for the conceptualization of reputation needs to be investigated in more detail.
In structural equation modelling, every construct or latent variable is assigned a...