A test of the effects of retail store rating context on the stability of alternative multientity scaling methods

Abstract

Although multientity scaling has been advocated as a method of increasing sampling efficiency, reducing questionnaire size, increasing rating discrimination, and reducing halo effects, recent research indicates multientity scaling procedures may be vulnerable to context induced instability. This study extends the research on context effects in multientity scaling. The results of two experiments support previous research that indicates entity-based scaling, when compared to attribute-based scaling, may be less vulnerable to context induced rating instability. In addition, the findings of both experiments indicate that oral data collection represents a promising method of increasing the stability of entity-based multientity scaling.