An Alternative Way of Establishing Measurement in Marketing Research – Its Implications for Scale Development and Validity

Abstract Quantitative consumer and marketing research is looking back on an era of construct operationalization predominantly based on classical test theory as a technical framework of scale development. Rasch measurement theory provides an alternative framework of measurement. Previous studies demonstrated the potential of Rasch measurement for marketing research from a theoretical viewpoint, and reported applications of Rasch measurement models to existing marketing scales. This paper focuses on the fact that the Rasch model explicitly accounts for the different amount of the construct that is needed by the respondents to agree with different items. Each item is characterized by an item parameter, i.e. the item location, that expresses the amount of the property to be measured the item stands for. Whereas the foundation of the Rasch model, i.e. specific objectivity, provide evidence of construct validity of a scale that fits the model, the range of item locations spans the latent dimension and gives insight into the meaning of different levels of the construct and thereby adds to content validity. An empirical example shows that applying Rasch models to existing scales does not reveal the full potential of the model, even if a comprehensive, albeit “classical” item pool is referred to. Consequently, only newly generated items are likely to span a wide range of items providing content validity.