| Abstract: | This paper sets out to test a popular feeling in University circles that lecturers can predict with some accuracy the final degree classification of students they have taught or known on a personal level. The results of an experiment carried out in the Economics department of the University of Bradford suggest that the magnitude of the forecasting error for these predictions is larger than the staff concerned might have anticipated, especially when compared with some alternative “naive” models. Part of the explanation of this gap might result from the fallacy of retrospective inference from results to causes. |
