[Teaching Tips] Several Ways Professors Can Learn Generative AI (2024.09.13.)
- 교무팀
- Hit259
- 2024-10-10
Several Ways Professors Can Learn Generative AI Sang-eun Lee, Min-young Ku, Ye-jin Kim |
Recently, Generative AI has been rapidly expanding its influence in various fields, including education. Since last year, the Center for Teaching & Learning Innovation has been holding various workshops on generative AI to give our university’s professors the opportunity to learn about this new wave of technology. From April 2023 to August 2024, the center hosted 11 online workshops using ZOOM, 4 flipped workshops held during the vacation period, which included online video viewing and offline practice, and 1 asynchronous online workshop where recorded video content was provided through iCampus non-regular courses. This teaching tip summarizes the generative AI teaching workshops that have been held from 2023 to the present, reflecting on some of the methods professors have used to learn about generative AI.
1. ZOOM Workshop: A Gateway to the Ever-Evolving Generative AI
The workshops held on ZOOM, free from spatial limitations, had an average of about 100 participants per session. The number of professors who registered for the workshops was about 1.5 times higher than those who actually participated, indicating a high level of interest and active engagement from the professors regarding generative AI. Upon a detailed review of the current situation, we identified the workshop topics with the highest participation and satisfaction rates. The workshop with the most participants, totaling 144, was the first one held on April 26, 2023, titled "Understanding ChatGPT Technology" (Speaker: Professor Jin-Young Park from our university's Department of Artificial Intelligence). Although this workshop was designed to teach the basics of generative AI principles and technology, the content turned out to be more difficult than expected, leading us to create a separate teaching tip summary afterward. The second most attended workshop, held on June 24, 2024, was "Utilizing Generative AI for Education and Research" (Speaker: Professor Tae-Yong Kim from Kyung Hee University), with a total of 125 participants. The third-highest attended workshop was "Hands-on Practice with ChatGPT and Other Generative AIs," held on June 29, 2023, with 117 participants (71 of whom were full-time faculty). This workshop was designed to allow professors to participate via Zoom and practice individually. After COVID-19, as teaching workshops moved to Zoom, it is true that the workshops became free from spatial limitations, allowing many professors to participate simultaneously. However, Professor Tae-Yong Kim from Kyung Hee University mentioned that having over 100 professors participating in each workshop is an unparalleled level of engagement compared to other universities.
The Zoom workshop "Useful GPT-4 Features for Academic Research," held on November 29, 2023, recorded a high satisfaction score of 4.62. Additionally, the workshops "ChatGPT and Changes in University Education" (June 7, 2023) and "Hands-on Practice with ChatGPT and Other Generative AIs" (June 29, 2023) both recorded satisfaction scores above 4.5. These results show that professors were satisfied with workshops that directly related to their research and teaching, and they appreciated not only listening to explanations about generative AI but also having opportunities to practice using it themselves.
2. Flipped Workshops with Offline Practice : Highest Participation Competition, Highest Satisfaction
The offline workshops were held in a flipped learning format, where participants would watch a video in advance for preparation and then engage in hands-on practice during the offline session. The primary goal of this approach was to give professors the opportunity to directly use generative AI themselves. Examples of workshops that included offline practice were the "Flipped Generative AI Practice Workshop" held during the winter vacation in January and the "Self-Learning Generative AI - Offline Practice-Centered Flipped Workshop" held in August. The number of applicants for both workshops was more than double the available spots, so participants had to be selected through a process.
Two workshops had special significance as our university students participated as assistants, helping professors smoothly conduct the practice sessions. When organizing the workshops, we aimed to shift from the usual situation where professors teach and students learn, giving professors the chance to learn while students provided support. This created an opportunity for both parties to better understand each other's perspectives. Although we were initially nervous about the new format of having students assist in the workshop, the satisfaction rate was close to 5.0.
In August, we held the "Self-Learning Generative AI - Offline Practice-Centered Flipped Workshop," where we took on a challenge by having professors prepare through pre-study using YouTube videos and not inviting any experts for the offline practice. This workshop was designed to help professors develop skills in using generative AI tools that could be beneficial in teaching and research. The practical session included creating lesson materials using the PPT tool 'Gamma' and building chatbots using GPTs. Both workshops involved students from the design stage as part of a co-design team, where they helped develop the practice topics, participated in workshop rehearsals, and provided assistance during the sessions. After the workshops, satisfaction surveys were conducted, and both the Gamma and GPTs workshops recorded high overall satisfaction scores of over 4.8. However, the satisfaction scores for "applying what was learned in the workshop" were relatively lower for both workshops. In the Gamma workshop, the item with the smallest standard deviation was, "I learned new information through the 'Creating Lesson Materials with Gamma' workshop," with an average score of 4.96 and a standard deviation of 0.19, which were the highest average and lowest deviation among all survey items. For the GPTs workshop, the item with the smallest standard deviation was, "Overall, I was satisfied with the 'Creating Chatbots with GPTs' workshop," with an average of 4.94 and a standard deviation of 0.24, again showing the highest average and smallest deviation.
Additionally, both workshops included common survey questions, such as whether the "Gamma and GPTs practice sessions were helpful for learning generative AI programs without expert guidance" and whether the "student assistants provided sufficient help." Both items recorded overwhelmingly high scores, with averages above 4.9. This suggests that professors found self-learning, supported by assistants, to be an effective way to learn ever-changing and developing generative AI technologies.
Although the offline practice workshops had fewer participants compared to the Zoom workshops, they recorded very high satisfaction rates. This indicates that professors have few opportunities to directly engage with newly emerging technologies and that hands-on practice was effective in helping them understand and utilize generative AI.
3. iCampus Online Workshop: The Most Registered, But the Least Completed Workshop
The workshop held as a non-regular iCampus course was open for about a month, from February 15, 2024, to March 29, 2024, allowing professors to participate at any time at their convenience. The video content consisted of 15 videos recorded by Professor Tae-Yong Kim from Kyung Hee University, who previously conducted workshops, under the title "Research Methods Using Artificial Intelligence." The iCampus online workshop was designed to overcome the limitation of real-time-only participation in previous Zoom and offline workshops by offering it asynchronously. While around 280 participants registered for the workshop at the start, unfortunately, only about 30%, or 85 participants, completed it. Although the convenience of accessing the content anytime is a major advantage of asynchronous online workshops, this suggests that maintaining motivation and watching a large number of videos to the end can be challenging. Additionally, the lack of efforts from the center to help sustain learning over the month-long period was identified as another reason for the low completion rate. Therefore, the Center for Teaching & Learning Innovation recognized the need to encourage continuous participation and plans to provide support to help professors complete their learning, similar to its other online education programs (e.g., T.A. training), when offering future asynchronous workshops.
We have now reviewed the three different types of generative AI teaching development workshops conducted by the Center for Teaching & Learning Innovation and their outcomes. The center experimented with various formats, including Zoom, offline, and iCampus online, to prepare effective workshops that would help professors learn about generative AI. Through participation rates and satisfaction surveys for each workshop, we were able to gather insights into the formats and content.
Among the three formats, the one with the highest participation rate was the real-time Zoom lectures. Workshops held via Zoom, free from spatial limitations, recorded an average of about 100 participants per session. Also, The format with the highest satisfaction was the offline workshops. These workshops recorded a very high satisfaction rate of over 4.8, demonstrating that professors could effectively learn generative AI programs with the help of student assistants, even without experts. Lastly, the asynchronous online workshop highlighted the convenience of being able to access the content anytime but also showed that maintaining motivation to complete all the videos was a challenge.
Moving forward, the Center for Teaching & Learning Innovation plans to strengthen its guidance on incorporating generative AI into teaching. The satisfaction survey results indicated high satisfaction with directly learning how to practically use generative AI. Thus, we plan to design future workshops focusing on specific examples and tools that can be directly applied in university courses, such as writing and project-based learning. In this process, we will enhance practice-centered education to give professors the opportunity to personally experience using generative AI tools, increasing the likelihood of incorporating them into their teaching. Additionally, we aim to discover and create content featuring practical examples that fit the real educational context, providing guidance on generative AI. We will strive to offer more engaging asynchronous workshops by producing highly immersive content and adding interactive elements. Thank you for your interest in the generative AI teaching development workshops provided by the Center for Teaching & Learning Innovation, and we look forward to your continued participation.
Sang-eun Lee, Min-young Ku, Ye-jin Kim (2024), Several Ways Professors Can Learn Generative AI (CTL Teaching Tips #46). Seoul: Sungkyunkwan University, Center for Teaching and Learning Innovation.