Predictive modelling using Rasch's person-item map: Intrepreting and assessing in Malaysia manufacturing firms

Abstract

One of the more attractive features of the Rasch Model as a paradigm for measurement in the social sciences is that it has the potential to predict the probability of persons' ability on the instrument which was set up on a scale with equal intervals. A person-item map (also known as Wright Map) are used as graphical visualizations of such scales, and it is able for the researcher to interpret. One reason for this is that they are premised on an arbitrary unit of measurement, the logit. By using logit value, person-item maps are able to show simultaneously the hierarchies of both person and items at the same scale (logit). The map analysis also provides clues to how well the content of the item yielded useful information about the firm's ability. This study focuses on final of 88 constructed items, where items must retain their relative difficulty on the equal interval scale (logit), regardless of the ability of the firms that challenges the items. This study also assesses the possibility of firm's ability to predict the firm's performance on a linear scale of measurement and the model could provide meaningful information on the firm's effectiveness in practicing Management Innovation (MI) items.