Stephanie Kelley, PhD
The Role of Explainable Artificial Intelligence聽in Influencing Loan Officer Behaviour
Scotiabank Professor in Innovations in Business Technology, Department of Finance, Information Systems and Management Science, Saint Mary鈥檚 University
Date:聽Wed, February 12, 2025
Time: 11:00 am - 12:00 pm (AST)
Location:
听滨苍-笔别谤蝉辞苍: "I" Building, room 222
听痴别苍耻别:
聽 聽 MS Team
聽 聽聽
聽 聽 Meeting ID: 210 687 611 371
聽 聽 Passcode: DjeDzk
Abstract:
We conduct a series of lab-in-the-field and lab experiments to investigate the preferences of real AI users (loan officers) with respect to the tailoring of explainable artificial intelligence (XAI), and whether XAI can help increase AI model recommendation adherence. We first conduct a choice-based conjoint (CBC) survey (Study #1) in which we vary the XAI approach, the type of underlying AI model (developed by data scientists at a large institutional lender in Asia with real data on the exact loans that our experimental subjects issue), the number of features in the聽visualization, the applicant aggregation level, and the lending outcome. We analyze the survey data using Hierarchical Bayes method, generating preference estimates (part-worth utilities) for each AI user and at the sample level across every attribute combination. We observe that (i) the XAI approach is the most important factor, more than any other attribute, (ii) AI users prefer certain combinations of XAI approaches and models to be used together, (ii) users prefer nine or six features in the XAI visualizations, (iv) users do not have preferences for the applicant aggregation level, (v) their preferences do not change across positive or negative lending outcomes. Next, we conduct a lab experiment (Study #2) in which we vary the XAI visualization accompanying an AI model recommendation. We find that (vi) more preferred XAI visualizations lead to greater AI model recommendation adherence, driven by the most important factor, the XAI approach. We then conduct a series of profitability simulation to determine how adherence behaviour impact firm profitability. We show how firms can strategically tailor XAI visualizations to drive greater use of AI models for improved operational efficiency.
Speaker Biography:
Dr. Stephanie Kelley is the Scotiabank Professor in Innovations in Business Technology and an Assistant Professor of Management Science at the Sobey School of Business at Saint Mary鈥檚 University, where she studies the ethics of artificial intelligence (AI). She investigates both algorithmic and governance solutions and currently has projects on explainable AI, human-AI collaboration, measuring AI governance, and AI ethics audits. She has worked with several organizations to implement AI ethics including the Monetary Authority of Singapore, OSFI, Lululemon, Stewart McKelvey, and several Canadian banks. Before returning to academia, she held various marketing, sales, and analytics roles in the pharmacy and home hygiene industry at Reckitt Benckiser.
Industrial Engineering seminar series contact person:
Prof. Hamid Afshari
email: Hamid.Afshari@dal.ca
General Enquiry:
Ms. Susan Russell McGrath
Tel: 902.494.3125
email: iegrad@dal.ca