Downloaded 12 times
Uploaded byGalit Shmueli
Reinventing the Data Analytics Classroom
The document discusses the evolution of a business analytics course taught by the author, highlighting changes in teaching methodology from traditional lectures to a flipped classroom approach while incorporating interactive elements. It emphasizes the importance of blending online resources with on-ground activities to enhance student engagement and learning outcomes. The author shares insights from their experience and offers strategies for effectively integrating technology and active learning into the analytics classroom.
More Related Content
Similar to Reinventing the Data Analytics Classroom
Reinventing the Data Analytics Classroom
- 1. Reinventing the Data AnalyticsClassroom International Symposium on University Teaching Innovation, Yunlin U of Sci & Tech, Taiwan, May 2019
- 3. My Research - Statistical& data mining methodology - Interdisciplinary ● To Explain or To Predict? ● Prediction & Causal Models ● Information Quality ● Behavioral Big Data
- 4. Teaching (1993-2019) I neverhad formal teachers’ training Online (Statistics.com, FL MOOC) ● Predictive Analytics ● Forecasting Analytics ● Interactive Data Visualization Engineers (Technion, CMU) ● Industrial Statistics ● Engineering Statistics & Quality Control ● Sampling, Surveys, & Society B-school PhD (UMD, NTHU) ● Scientific Data-Collection ● Research Methods B-school MBA & Executive (UMD, ISB, NTHU) ● Business Analytics using Data Mining ● Forecasting Analytics ● Data, Models, and Decisions ● Statistical Linear Models in Business ● Leveraging Business Through Analytics Started out in 1993 as a TA in Statistics & Probability - Weekly recitation sessions - Before the days of PPT, LMS, ...
- 6. Today I’ll focuson the evolution of my Business Analytics course
- 7. Business Analytics course ●Masters-level business school elective course ● Diverse audience (business, technical & non-tech, with/out work experience) Course Objectives: 1. Learn how business and data mining integrate 2. Hands-on experience with data mining algorithms 3. Become familiar with industry applications 4. Know what you don’t know I developed two BA courses: ○ Data Mining ○ Forecasting I’ve taught with two formats: ○ Traditional (lecture) ○ Flipped
- 8. Through my journeywith this course, I will share: What I learned over the years ● Cultural adaptations (silence/activity; active/passive) ● The effect of classroom space ● The effect of course length ● Technology choices How I tried to achieve course and learning objectives by adding interactivity ● Online, on-ground activities and resources ● How to blend them meaningfully ● Blending classroom with industry ● Needed preparation ● How to increase active/interactive learning into any (analytics) course ● Benefits and Challenges
- 9. In 2014, Ijoined NTHU and taught two courses in Fall… Forecasting - traditional Data mining - flipped
- 10.
- 11. To understand what happened, let’sroll back in time...
- 12. Data Mining forBusiness Designed & taught 2004-2010 MBA elective course 14 weeks (lectures) Main resources: Textbook Handouts (slides) Individual assignments clickers! To increase active/ interactive learning: Team project
- 13. What are clickers? ●Each student has a clicker ● Use clicker to answer instructor’s questions ● See answers immediately ● Confidentiality maintained wireless transmitters (like TV remote)
- 14.
- 15. Business Analytics UsingData Mining Indian School of Business Taught 2010-2013 MBA elective course 5 weeks (10 lectures) Main resources: Textbook Handouts (slides) Individual assignments Team project Clickers 60 students per section 1-2 sections
- 16. In 2011, Iadded an online forum In 2012, I flipped it!
- 17. How to flip? -record videos - create online materials (discussion, reading, quiz,...) What do you do face-to-face in class? How to combine the online + offline resources/efforts in a meaningful way?
- 18. In class 1. Discussion 2.Clicker Qs 3. Hands-on Assignment 4. Team presentations 5. Guest lecture Before class 1. Watch videos 2. Read 3. Discuss online After class 1. Individual assignment 2. Team project + contest (meetings, analysis, ...)
- 20. Student feedback “Supercharged student-facultyinteraction: Instead of the typical case / theory oriented case pre-reads, Prof. Galit posted videos in first person format taking us through the basic concepts and asking us to run through some sample problems. This ensured that during class, everyone had moved beyond the basics and the depth of questions discussed was far more probing than an average course.”
- 21. What I learnedfrom flipping Class time more effective & fun ● Higher-level discussions, closer to industry ● More interaction with students ● Students excited Online adds flexibility, support ● Students don’t “miss out” on new material if sick/absent ● Strengthen peer-to-peer learning using online forum I can build & run my own flipped classroom But: - Heavy set-up efforts - Online time from instructor & students - Students lose a lot by missing on-ground session, or by coming to class unprepared
- 22. Forecasting (lectures) flipped Data mining Traditional-but-active vs.Flipped: Gap bigger in India or Taiwan?
- 23. How to adaptflipped format for Taiwan? Online Examples (videos, discussion cases) Video captions Face-to-face (in class) How to break passive habits? (Why are students passive?) Combine online + offline resources/efforts in meaningful way What are student habits? What makes them learn and perform best?
- 24. Before I showyou my current Taiwan course format, try this experience
- 25. Let’s learn somedata mining: How can a company use prediction? Suppose we have a measurement of interest on many people: ● blood pressure ● diet (veg/non-veg) ● teaching performance score ● daily hours spent on mobile phone Data Mining is used to predict this measurement for new people What should a company predict? - What data do they have? - What challenges & opportunities do they face?
- 26. Exercise (in pairs): Predictingwith EasyCard Data 1) EasyCard has data on ________ (who? what measurements?) 2) We can use EasyCard data to predict _____________ (what? for who?) 3) These predictions can be used by ________________ (who?) to _________________ (do what?)
- 27. 1. EasyCard hasdata on _____________ 2. We can use EasyCard data to predict _______ _____________ 3. These predictions can be used by ____ to _____
- 28.
- 30. Recording videos: MOOCto the rescue!
- 32. Example video (4minutes)
- 33.
- 34.
- 35.
- 36. Team project proposals +final presentations Feedback
- 37. Student workshops (teach newsoftware)
- 38. Guest Session --prepare your guest!
- 39. “Field trip” tocompany ideation/contest/etc.
- 40. Teams collaborate with companythrough course project
- 41. Many trees makea forest
- 42. Sharing with industry ● Industryvisits to classroom Team presentations Meet project team ● Visit to company Meet project team Ideation session Contest ● Online Slack, Skype Sharing online ● Discussion board Canvas ● Team Project Slides Google Slides ● Contests Kaggle Google Sheets Sharing on ground ● Whiteboards with team; with class ● Discussions in teams with everyone ● Team presentations + feedback ● Workshops by students Blending online + offline resources
- 43. What if youcan not get a suitable - Space - Class size - Resources Or you are required to lecture? Solution #1: Use online content to support your lectures / assignments
- 44. What if youcan not get a suitable - Space - Class size - Resources Or you are required to lecture? Solution #2: Create voluntary out-of-class activity
- 45. So… What changedand what did not in the evolution of my course? Changed: Technology Audience, culture Classroom space Delivery mode Me Did not change: Course objectives Course topics Teaching principles & goals Me
- 46. Final Words: Reinvent theclassroom by adding meaningful interactivity - Experiment - Get feedback - Learn from mistakes It takes time, effort & courage But can be worth it!