The incoherence of Rasch measurement: A critical comparison between measurement in psychology and physics

Probabilistic models for measurement were originally developed by Georg Rasch (1901-1980) in an attempt to establish the measurement of psychological properties on ratio scales. Rasch defended the view that there exists a consistent and general structural analogy between measurement in physics and psychology. Though in theory the two approaches are comparable, it is argued that at a crucial point in experimental practice there is a divergence between psychology and physics. While psychologists mainly trust in statistical fit as a means to achieve the objective of measurement, physicists focus on getting phenomena under control under experimentally constrained conditions. Derived measurement as applied in physics may be regarded as a way out of the impasse, but given the empirical hindrances encountered when implemented in psychology, it is concluded that Rasch models are inappropriate as tools to attain measurement in psychology.

Automated summary

Probabilistic models for measurement, developed by Georg Rasch, aimed to establish measurement of psychological properties on ratio scales. However, there is a divergence between psychology and physics in terms of measurement practices. While physicists focus on experimental control, psychologists rely on statistical fit. It is concluded that Rasch models are not suitable for measurement in psychology due to empirical hindrances.