How fine is C-OAR-SE? A generalizability theory perspective on Rossiter's procedure

In an important contribution to the measurement literature, Rossiter [Rossiter, J. R. (2002). The C-OAR-SE procedure for scale development in marketing. International Journal of Research in Marketing, 19 (December), 305-335] proposes a new procedure, labeled as C-OAR-SE, for the development of scales to measure constructs in marketing. In this commentary on the procedure, we first take a historical perspective on scale development in marketing to suggest that while Rossiter's procedure makes us refocus on the conceptualization of constructs, it has the potential to create an important gap by advocating against empirical validation of constructs. We then suggest that multivariate generalizability theory integrates the two competing perspectives by requiring a balanced emphasis on conceptual rigor and empirical validation of constructs.