Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
Tackling Professorial Expert Bias: The Role of ChatGPT in Simplifying Financial Accounting Exam Texts
13
Zitationen
4
Autoren
2024
Jahr
Abstract
ABSTRACT Experts tend to produce complex texts but mistakenly consider them an easy read for their audience. Accounting professors likely exhibit such expert bias in formulating exams. Consequently, students struggle with comprehension, giving rise to high failure rates and a shortage of talent for the profession. This study explores the potential of ChatGPT for overcoming expert bias and improving financial accounting exam texts to increase comprehension and performance and reduce frustration and confusion. We ask three key questions: (1) Does text modification with ChatGPT lead to increased perceived comprehension?, (2) Does text modification with ChatGPT result in better performance?, and (3) Does perceived comprehension mediate the relationship between ChatGPT-based text modification and performance? We find that students benefit from ChatGPT-based modification through increased perceived comprehension and reduced confusion and frustration. Causal mediation analysis suggests that improved perceived comprehension attributes significantly to exam performance, but we cannot document an overall positive effect. JEL Classifications: A2; A22; M4; M41; I23.
Ähnliche Arbeiten
Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI
2019 · 8.214 Zit.
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
2019 · 8.071 Zit.
High-performance medicine: the convergence of human and artificial intelligence
2018 · 7.429 Zit.
Proceedings of the 19th International Joint Conference on Artificial Intelligence
2005 · 5.776 Zit.
Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI)
2018 · 5.418 Zit.