A Multi-level Analysis of Mistrust/Trust Formation in Algorithmic Grading - Responsible AI and Analytics for an Ethical and Inclusive Digitized Society Access content directly
Conference Papers Year : 2021

A Multi-level Analysis of Mistrust/Trust Formation in Algorithmic Grading

Abstract

While the concept of trust continues to grow in importance among information systems (IS) researchers and practitioners, an investigation of mistrust/trust for- mation in algorithmic grading across multiple levels of analysis has so far been under researched. This paper proposes a multi-level model for analyzing the for- mation of mistrust/trust in algorithmic grading. More specifically, the model ex- amines multiple levels at play by considering how top-down forces may stimulate mistrust/trust at lower levels, but also how lower-level activity can influence mis- trust/trust formation at higher levels. We briefly illustrate how the model can be applied by drawing on the case of the Advanced Level student fiasco in the United Kingdom (UK) that came to head during August 2020, whereby an algorithm was used to determine student grades. Although the paper positions trust as a multifaceted concept, it also acknowledges the importance of researchers to be mindful of issues pertaining to emergence, duality, context, and time.
Fichier principal
Vignette du fichier
512902_1_En_61_Chapter.pdf (375.48 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03648163 , version 1 (21-04-2022)

Licence

Attribution

Identifiers

Cite

Stephen Jackson, Niki Panteli. A Multi-level Analysis of Mistrust/Trust Formation in Algorithmic Grading. 20th Conference on e-Business, e-Services and e-Society (I3E), Sep 2021, Galway, Ireland. pp.737-743, ⟨10.1007/978-3-030-85447-8_61⟩. ⟨hal-03648163⟩
20 View
3 Download

Altmetric

Share

Gmail Facebook X LinkedIn More