Stephen Downes

Knowledge, Learning, Community

This is a good article, well-structured and clear, about the use of AI in assessment. The authors identify five key issues with the standard assessment paradigm (SAP): assessment design is onerous, they only provide a discrete snapshop, they uniformly assign the same tasks to everyone regardless of prior knowledge, they are inauthentic, and they are antiquated, evaluating skills that are obsolete. The authros then offer an account of how AI-based assessment address easch of these issues in turn, creating (I think) a compelling value proposition for them. They then outline some concerns to be addressed: the sidelining of expertise and opacity of accountability, limitation of the pedagogical role of assessment, the use of surveillance technology (especially for 'stealth assessment'), and assessing only for a set of traditional male-centered western values.

[Direct link]


Stephen Downes Stephen Downes, Casselman, Canada

Creative Commons License.

Copyright 2022
Last Updated: May 15, 2022 09:12 a.m.