The Rasch probabilistic model for measuring affective responses to product features

Analysis for eliciting affective responses to product features can present shortcomings if fundamental measurement principles are violated. This paper overviews the concepts of an approach for human-centred product design to overcome the problem, namely the Rasch probabilistic model. The Rasch model consists of a systematic search process currently applied with intensive computing support. The paper shows that the model could enable the development of a frame of reference for measurement structures. This is possible because Rasch analysis validates a scale through of calibration, assessing invariance of items (independent variables), potential item bias, validity of category scoring system and fit of items to the model. As a result calibrated scales might be used in customised applications without loss of comparability.