Effective Learning Techniques Versus Feeling in Education Policy Development

In the last short talk at Team-Based Learning Collaborative Asia Pacific Community Symposium was by Prof Judy Currey at Deakin University. Professor Currey cited Deslauriers (2019), on the well known phenomenon that students prefer passive learning. It occurred to me that this might also apply to policy makers and legislators, who think that students learn best when listening to a teacher. Fortunately the head of the universities accord review, Professor O’Kane, gets it: "... stressed better tutorial-style teaching to provide for experiential learning and experimentation was needed ...' (Uni graduates lack crucial skills and online study is making it worse, Gus McCubbing and Tom Burton, AFR Aug 22, 2023).

References

Deslauriers L, McCarty LS, Miller K, Callaghan K, Kestin G. Measuring actual learning versus feeling of learning in response to being actively engaged in the classroom. Proc Natl Acad Sci U S A. 2019 Sep 24;116(39):19251-19257. doi: 10.1073/pnas.1821936116. Epub 2019 Sep 4. PMID: 31484770; PMCID: PMC6765278  DOI: 10.1073/pnas.1821936116.

Meet my AI alter ego

Phillip Parannik, and Finn Johnson

At day 2 of the Team-Based Learning Collaborative Asia Pacific Community Symposium we have been learning how to create AI characters to be used as tutors, & simulated clients for students. I was very skeptical of the idea, but in a few minutes I created "Tom", a simulation of me, for teaching computer students. I can now retire. ;-)

This is using Inworld, which was originally created for use in computer games, but is now being used for teaching. The workshop is being run by Phillip Parannik, and Finn Johnson from University of Sydney. The process was very much simpler than the last workshop on AI tutors I took part in, 2018.

If anyone in Canberra would ;like a system built with this, the ANU computer project students could help. 

Team Based Learning Helped with Remote Learning

Larry K. Michaelsen

Greetings from day 2 of the Team-Based Learning Collaborative Asia Pacific Community Symposium. The keynote today is Larry K. Michaelsen, pioneer of Team Based Learning (TBL). He said that teams formed slower online than face to face, also that technology had provided more options for prework. Also Dr Michaelsen reminisced about the marked sense card readers which were used for scanning student responses before laptops and smart phones were used. He pointed out that part of the team bonding process was filling out the one joint response. 

A key point for me was that this team building process could occur in asynchronous mode. That is Dr Michaelsen said that as long as there was a deadline for the team response to be submitted, the participants could be anywhere, arranging their own work. 

Dr Michaelsen suggested AI would have more difficulty answering TBL questions, as they are very specific. He gave the example of "Where would you locate a dry cleaning business in Norman Oklahoma, and why?". So I asked ChatGPT, and it gavce a very generic response, which an assor would not be impressed with. When I asked for more specific locations, it gave answers Dr Michaelsen said that anyone who knows Norman would know are not good locations:

"Determining the ideal location for a dry cleaning business in Norman, Oklahoma would depend on various factors such as target market, competition, accessibility, and zoning regulations. Here are some considerations to help you decide on a suitable location:

1. Commercial Areas: Look for commercial areas with a mix of retail and service-oriented businesses. Areas with a high concentration of offices, hotels, or shopping centers can attract potential customers who may need dry cleaning services.

2. Residential Neighborhoods: Consider placing the dry cleaning business in or near residential neighborhoods, as this would make it convenient for local residents to access your services. Proximity to residential areas can generate regular and loyal customers.

3. Proximity to Offices or Corporate Parks: If you are targeting the professional working population, setting up near offices, corporate parks, or business districts may be advantageous. Many professionals require dry cleaning services for their work attire.

4. Accessibility and Parking: Choose a location with easy access and sufficient parking space. Customers prefer convenience, and a location with ample parking makes it more convenient for them to drop off and pick up their items.

5. Competitor Analysis: Assess the competition in different areas. If there are already established dry cleaning businesses in a particular location, evaluate whether there is room for another competitor or if there's an underserved market elsewhere.

6. Demographics: Understand the demographics of the area. Consider factors such as income levels, age groups, and lifestyles of the residents to ensure that your services align with the needs of the local community.

7. Zoning Regulations: Check local zoning regulations to ensure that the chosen location complies with any restrictions related to operating a dry cleaning business. Some areas may have specific zoning requirements for businesses that use certain chemicals.

8. Visibility and Signage: Choose a location with good visibility from the road, and invest in clear and attractive signage. This will help attract passing traffic and make your business easily identifiable.

9. Cost Considerations: Evaluate the cost of renting or purchasing space in different areas. Balance the costs with the potential customer base and revenue generation of each location.

10. Future Growth: Consider the potential for future growth and development in the chosen location. An area that is experiencing growth may attract more customers over time.

Ultimately, a comprehensive analysis of the local market and careful consideration of these factors will help you determine the most suitable location for your dry cleaning business in Norman, Oklahoma."

One of the delegates suggested occasionally tell the students to turn off their gadgets. Dr Michaelsen suggested then asking studnts to reflect on the differences in the experience. However, I suggest the instructors need to also reflect on what technology they are dependent on: would the be willing to turn off the lights and air-conditioning in the room, turn off their hearing aid, take off their glasses? One generations new tech is just routine for the next. While is is tempting to think of the good old days, some technology is so useful it is worth becoming dependent on it.  

Online Team Based Learning

Greetings from the first workshop at Team-Based Learning Collaborative Asia Pacific Community Symposium. Rosa Howard, University of Sydney, is talking on "Adapting TBL to Changing Times". This highlighted the benefits of providing text base input, as opposed from verbal. This raises an interesting issue as to if graduates need to be able to confidently talk in front of large groups. If it becomes routine to use computer mediated communication in the workplace talking to a group becomes an obsolete skill, like handwriting.

The Fear of Failure & TBL

Greetings from the first workshop at Team-Based Learning Collaborative Asia Pacific Community Symposium. This is using Intedashboard software and ChatGPT. Something which slowly dawned on me during the morning session was that the medical educators were using Team Based Learning (TBL) not as a general term for learning in teams, but a very specific methodology. I felt a bit like the student who was not told there was to be a test.  At the start of the workshop we were asked to lo into Intedashboard, and was immediately asked a question I didn't understand, at which point I had the unprepared student panic. Having recovered from this I struggled with some of the acronyms MCQ (multiple choice questions), ILO (no idea). 

The workshop was modeled on TBL, and this was a familiar format for me. What I found fascinating were examples of giving ChatGPT very complex requests to create course material. Doing this never occurred to me, no more so that I would get a random person off the street and ask them to do so. Habitually I would consider the qualifications and experience of any person, company or organisation. But ChatGPT is essentially a black box. 

As a team we then used ChatGPT to create objectives for a learning exercise. It was interesting to see the range of experience with ChatGPT, with some participants having never used it, while others routinely used it to create course content. My team got a bit naughty, asking ChatGPT to write the evaluation of what we thought of ChatGPT. Another team accidentally ended up with a rubric for assessing learning objectives. One team found what ChatGPT too wordy, but another went further and told it to be less wordy, which it then was.

In terms of a TBL exercise I found this like the previous development exercises. The first is the worry of getting the software used to work. The next is how rushed everything seems to be. There is then the worry of getting the exercise done in time. My approach would always be to produce a draft answer, then refine it. However, that is an approach difficult to get a group to do. 

ps: One use for ChatGPT I hadn't thought of was suggested b y one of our team: planning an itinerary for a trip. 

Team Based Learning & AI

Professor Simon Buckingham Shum

Greetings from the University of Sydney where I am taking part in the Team-Based Learning Collaborative Asia Pacific Community Symposium. Most of the staff are from USyd medical schools, but I fell at home as we are talking about experiential learning. The keynote is Professor Simon Buckingham Shum, Director of the UTS  Connected Intelligence Centre, being provocative about AI. He suggested universities need to have their own generative AI tools, as it is not acceptable for staff and  students to have to enter their sensitive information into offshore commercial ones. One practical use is in debriefing students after they participate in a simulation. What is most useful about the Professor's analysis is that it treats AI as a tool, not a problem, and asks us how we might use it.

Professor Buckingham Shum pointed to a paper he wrote about how one such tool carried out a very good analysis of a complex policy paper (2024). There was one flaw in the Professor's analysis: he claimed the AI was in error by referring to an "Argument from ignorance" which did not exist. It turns out that this is known to philosophers such as John Locke, and politicians such as Donald Rumsfeld. When I asked about this the Professor suggested that students need to be gently introduced to the limitations of the AI, so they are not awed by it.

I am looking forward to workshops, and short talks later today and tomorrow, with people not only from Australia, but around the region.

ps: One area where TBL AI might be used is in helping students meet diffuse learning outcomes.

Reference

Buckingham Shum, S. (2024). Generative AI for Critical Analysis: Practical Tools, Cognitive Offloading and Human Agency. 1st International Workshop on Generative AI for Learning Analytics: 14th International Learning Analytics and Knowledge Conference (LAK’24), March 18-22, 2024, Kyoto, Japan [PDF]