4 présentations

• 
  10h30 - 10h55

  Developing Students' Conceptual Understanding through Exploring Key Industrial Engineering Models

  • Scott Flemming, prés., Dalhousie University
  • Clifton Johnston, Dalhousie University

  This paper describes an introductory course in Industrial Engineering (IE) which was created to increase students’ ability to appropriately use fundamental IE models and increase their conceptual understanding of these models. The models introduced are the following: quality control, inventory management, process flow, queuing, human factors, and linear programming. Elementary but detailed problems in each category are presented so that students can focus on exploring the mechanics of IE models, understand their inherent assumptions, and interpret their results well. Students are given a method to determine which model(s) may be most fitting for a particular scenario; that is, they should identify the goals, constraints, variables, and strategies (GCVS) for the context and compare these to typical GCVS for a particular problem category. Student success in problem diagnosis case studies was used as an indicator of their conceptual understanding of these key IE models/problems. Initial analysis suggests that conceptual performance increases when students engage in exploratory modelling tasks (typically using Microsoft Excel).
• 
  10h55 - 11h20

  Generating Multiple Datasets with Differing Statistical Significance for Hands-on Learning

  • Lori Murray, prés., Kings University College at Western University
  • John Wilson, Ivey Business School

  Regression analysis is a fundamental statistical technique used to model and analyze relationships between variables in data analysis. Teaching the importance of regression analysis can be greatly enhanced by providing hands-on learning opportunities with diverse datasets that exhibit different levels of statistical significance. We present a powerful tool for generating multiple datasets with n variables with the same univariate statistics but differing statistical significance. We discuss how these datasets can be used to teach important concepts in regression analysis, such as interpreting regression coefficients, assessing statistical significance, and understanding the impact of different levels of statistical significance on regression results, to better prepare students for real-world data analysis scenarios.
• 
  11h20 - 11h45

  Technology-Enabled Oral Assessments for Large Classes

  • Kyle Maclean, prés., Ivey Business School - Western University
  • Tiffany Bayley, Ivey Business School

  With the advent of advanced AI language models such as ChatGPT, traditional forms of assessment in business education have come under scrutiny, prompting educators to explore alternative methods. Educators are increasingly considering oral exams as they provide an opportunity for students to demonstrate their critical thinking, while avoiding issues with academic integrity. However, conducting these exams for large classes poses significant logistical challenges.
  In this presentation, we share the findings from a pilot study conducted during the Fall of 2020 in a large, required undergraduate course, where students uploaded video responses to short-answer written questions as a proxy for oral exams. We will cover the potential benefits and drawbacks of implementing asynchronous video responses in other courses, as well as the key lessons learned from this pilot study. We will address the practicalities of conducting such assessments, including technological requirements, grading procedures, and student engagement. Moreover, we will explore the potential of this innovative assessment method in fostering critical thinking, communication skills, and adaptability, essential competencies for the ever-evolving business landscape.
• 
  11h45 - 12h10

  Adaptive Stable Matching

  • Federico Bobbio, prés., Université de Montréal
  • Margarida Carvalho, Université de Montréal
  • Ignacio Rios, University of Texas at Dallas
  • Alfredo Torrico, Polytechnique Montréal

  Matchings with preferences are a popular model for many real world applications such as the school admission process. The standard solution concept in this setting is stability, which guarantees that the matching cannot be overridden by some of the agents. The concept of stability is usually built upon the individual preferences of the agents. In particular, we show the existence of a stable assignment when there are at most two siblings in a family and many different entry levels. If there is only one entry level of education, we can show existence under any family size.
  We also provide a mathematical programming formulations to find an exact solution for the problem, and introduce several pre-processing heuristics to speed up the computation. Finally, using both synthetic and real data from Chile, we show that clearinghouses can significantly improve students’ welfare when considering dynamic priorities.