The Rasch distribution: A discrete, general form of the Gauss distribution of uncertainty in scientific measurement
The final of three public lectures by David Andrich, professor emeritus and world-renowned expert on Rasch Measurement Theory.
About the lecturer
Professor emeritus David Andrich is a highly regarded assessment specialist from the University of Western Australia. He is particularly renowned for his expertise in Rasch measurement theory, a mathematical framework used to analyze and interpret data from tests and surveys. Andrich has worked with the famous mathematician Georg Rasch, and his contributions to the field have been widely recognized.
Read more about David Andrich >>
The lecture
The Rasch distribution: A discrete, general form of the Gauss distribution of uncertainty in scientific measurement
Key points:
- Studies of random distributions of replicated measurements of an object culminated in the quadratic exponential law of Gauss (normal distribution) for continuous distributions
- The Rasch distribution of replicated measurements also culminates in the quadratic exponential law, but for discrete distributions
- Applications of the Rasch distribution to two instruments from the social sciences are shown
Three questions to David Andrich
What first sparked your interest in modern test theory?
"I had a degree in mathematics and mathematics applied to classical physics. Measurement is an integral part of physics. Later I learned classical test theory, but it had no connection to measurement in physics. Rasch measurement theory applied to the social sciences connected to what I had learned about measurement in physics."
How does Rasch measurement theory differ from other measurement theories?
"Classical test, and item response, and representational measurement theories, are theories only in the sense of a body of coherent knowledge and principles. They are not scientific theories in the sense of explaining measurement. Rasch measurement theory does. In particular in relation to classical and item response theories, the case for the models of the theory are a-priori to any study of data."
What can attendees expect to learn from your upcoming lectures, and how might they benefit from attending?
"In addition to learning about Rasch measurement theory, they will learn to use software that operationalises the theory for analysis of questionnaire and educational and psychological test data. They will learn how to assess, in a detailed, diagnostic way, the degree to which their questionnaires or tests produce measurements, and where any problems are present. "
Other lectures in the series
May 15: The contributions of Georg Rasch to measurement in
social sciences