Big Data May Help Evaluate Teacher Performances
Often there is much talk of improving student grades and performances in schools and colleges. However, we still follow an outdated method of assessing teacher performances. A new reliable method of evaluating teacher quality based on principles of econometrics is being developed by Jiaying Gu, Assistant Professor of University of Toronto, Department of Economics.
evaluation uses the linear shrinkage principle that assumes that distribution
of quality follows a bell curve. According to Jiaying Gu, a recipient of 2018
Polanyi Prize in Economic Science for her novel statistical methods in
evaluating teachers, the data set in such a system is too diverse to be
analysed one way. The differences in behaviour of individual teachers cannot be
judged easily such as innate ability or personal preference. Big data may give
quantities of nuanced information to tackle datasets that are difficult to
observe using traditional methods. In evaluation based on student grades, the
teaching style cannot be accurately factored in due to difficulty in observing
it. Therefore, Gu is working on datasets in elementary schools in USA with the
aim of accurately evaluating the quality of teachers.
The research outcome will have far-reaching impact on educational policy and help improve our educational system at various levels.