Journal Article


Grade expectations: How well can past performance predict future grades?

Abstract

Students in the UK apply to university with teacher-predicted examination grades, rather than actual results. These predictions have been shown to be inaccurate, and to favour certain groups, leading to concerns about teacher bias. We ask whether it is possible to improve on the accuracy of teachers’ predictions by predicting pupil achievement using prior attainment data and machine learning techniques. While our models do lead to a quantitative improvement on teacher predictions, substantial inaccuracies remain. Our models also underpredict high-achieving state school pupils and low socio-economic status pupils, suggesting they have more volatile education trajectories. This raises questions about the use of predictions in the UK system.

Attached files

Authors

Wyness, Gill
Macmillan, Lindsey
Anders, Jake
Dilnot, Catherine

Oxford Brookes departments

Department of Accounting, Finance and Economics

Dates

Year of publication: 2022
Date of RADAR deposit: 2022-08-10


Creative Commons License This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License


Related resources

This RADAR resource is Identical to Grade expectations: How well can past performance predict future grades?

Details

  • Owner: Joseph Ripp
  • Collection: Outputs
  • Version: 1 (show all)
  • Status: Live
  • Views (since Sept 2022): 101