rETHinking data science education in management: Collaborative learning for «Business Analytics»
Data science is redefining organizational decision-making; yet this transformation is fraught−especially in management−with immense methodological complexity. This project seeks to address this issue through a new educational format involving collaborative learning. This improves learning strategies in order to better understand methods; analyze & evaluate results; and create end-to-end use cases.
The project
Owing to the ongoing process of digitalization, managers need to be well skilled in artificial intelligence. However, they also have specific education demands that must be effectively met by our teaching effects. Hence, this Innovedum project aimed at educating future executives in operating, as well as managing, data science technology.
Implementation into teaching practice
We re-designed the course “Business Analytics” as planned. We changed the former “one-course-fits-all” paradigm by separate courses for MSc and MAS students. In addition, there was extensive peer instruction in both courses in order to leverage the distinct expertise across both student bodies. Beyond that, we completed an augmented reality app. This allowed students to approach the learning objectives at their own pace; interact with models and predictions in novel ways; and simplify, as well as visualize, even complex concepts. The introduction of augmented reality to our educational approach facilitated collaborative learning and was further coupled with a shift from the more common “method focus” in data science education towards case-based learning and critical thinking.
Lessons learned and further impacts
Our teaching innovation was published as part of ALEEDSA [1]. The objective was two fold: to validate our working hypothesis that augmented reality is effective for AI education.
For this, ALEEDSA was deployed with a simple case study, in which business analytics students were supposed to model credit risk at a fictitious bank. ALEEDSA was tested based on 21 students from both MSc and MAS backgrounds. Overall, 67 % of the students reported that interactive visualizations in augmented reality enabled them to better understand data. Similarly, 52 % of the students confirmed that ALEEDSA helped them more quickly infer relationships in the data. Moreover, the majority of students (86 %) voiced the opinion that augmented reality can support ML activities for non-experts.
Several additional lessons around our HoloLens app are published as part of the CHI conference:
[1] ALEEDSA: Augmented Reality for Interactive Machine Learning
A Ferrario, R Weibel, S Feuerriegel
CHI EA ’20: Extended Abstracts of the 2020 CHI Conference on Human Factors in Computing Systems, April 2020 Pages 1–8
https://doi.org/10.1145/3334480.3382937
Authors
-
-
Andrea Ferrario
