Skip to content

This template is for educators to clone/fork and turn into their own AI-guided learning activity.

License

Notifications You must be signed in to change notification settings

LiamOMueller/SABER_2024

 
 

Repository files navigation

SABER 2024: Developing No-Code, AI-Enhanced, Online Study Tools for Biology Students: A Beginner's Workshop

Workshop Presenters:

  • Keefe Reuther, an Assistant Teaching Professor in Ecology, Behavior, and Evolution at UC San Diego. He co-developed the course, Data Analysis and Design for Biologists. His research focuses on creating and evaluating tools, including generative AI, that enhance learners' science process skills. Currently, he has established a SABER Special Interest Group (SIG) on generative AI and has presented several seminars and posters on the subject in the past year.
  • Liam O’Connor Mueller, is an Assistant Teaching Professor in Molecular Biology at UC San Diego where he co-developed the course, Data Analysis and Design for Biologists. His research relevant to this proposal is focused on developing automated tools for improving instruction, which he presented a poster on at SABER West in 2023.
  • Grace Constantian is a senior undergraduate at UC San Diego majoring in Bioinformatics
  • Albert Nguyen is a senior undergraduate at UC San Diego majoring in Human Biology and minoring in Computer Science

Workshop Abstract:

In the era of generative artificial intelligence (GenAI) tools, like ChatGPT, educators face new challenges. They must integrate these transformative technologies into curricula to enhance student learning and prepare students with the skills they will need to leverage GenAI in industry after graduation (Agathokleous et al., 2023). Discipline-Based Education Research (DBER) on the use of this technology is crucial, as it provides evidence-based strategies for its appropriate use (Lodge et al., 2023; Yeralan & Ancona Lee, 2023). This hands-on workshop will guide participants through the development of an interactive, AI-driven web application that provides formative feedback to students as well as a framework for evaluating its effectiveness. Targeted at individuals eager to learn about generative AI tools such as ChatGPT, this workshop requires no prior technical knowledge or experience in web design or computer programming. By the session's conclusion, participants will have created their own web application, which they can use to provide the most advanced ChatGPT model at no cost to all their students. This method takes advantage of OpenAI's API function, offering enhanced privacy and giving educators and researchers significantly more control over the creation and evaluation of learning activities. In line with the principles of open educational resources (OER) and open-source code, a proof-of-concept web application developed by the workshop facilitators will be presented as a foundational example (https://schemastudy.streamlit.app/). Anchored in schema theory, this online active learning activity encourages students to engage deeply with definitions, connections, and applications of course concepts (Duschl, 2019). Participants are presented with a randomly selected term from a list compiled by the instructor and are prompted to construct a schema around it. Using student responses and instructor-prepared prompts, a real-time conversation with a ChatGPT tutor trained to offer precise formative feedback and engage in Socratic questioning is initiated (Wang et al., 2023; Lloyd et al., 2022). Workshop attendees will employ ChatGPT to curate their term lists and schemas, as well as to personalize and test their instructions for both students and the chatbot. With the aid of thorough documentation, participants will create GitHub and Streamlit accounts, clone a public repository, adapt it with their curated content, and launch a live web application. After setting up their version of the app template, participants will simulate student interactions, allowing them to use a design-based research approach to iteratively refine feedback accuracy and quality (Scott et al., 2020). The workshop will conclude with a structured discussion on the ethical integration of AI tools into biology education, focusing on potential impacts, challenges, and future directions for DBER (Bai̇doo-Anu & Ansah, 2023; Farrelly & Baker, 2023). At the end of the session, attendees will have a customizable web application to share with their students and a strategy for ethically leveraging the power of generative AI in biology education.

Workshop Learning Objectives:

By the end of the workshop, participants will be able to:

  • Generate a tailored list of 20 biology-specific terms and definitions using ChatGPT, demonstrating competency in curating course-relevant content.
  • Use the tailored list generated above with ChatGPT to build an interactive study activity that students can use to practice with your tailored list at multiple levels of Bloom’s Taxonomy.
  • Implement the setup and customization of a Streamlit web application by creating a GitHub repository and connecting it to ChatGPT via the OpenAI API.
  • Assess the web application's feedback accuracy by conducting simulated student interactions, applying strategies to improve content relevance and pedagogical value.
  • Propose strategies for integrating AI-enhanced tools into biology education, reflecting on potential impacts, ethical considerations, and future applications in a structured discussion.

Important Notes

  1. Pre-Workshop Preparation: Participants should have a Google account, a GitHub account, an OpenAI account to generate an API key, and a Streamlit account.
  2. Workshop Duration: 3 hours
  3. Workshop Format: The workshop will be a combination of live demonstrations, hands-on activities, and group discussions.
  4. DO NOT EDIT THE APP.PY FILE: This file contains the code for the web application. Participants should only edit the term list in the config.py file.

About

This template is for educators to clone/fork and turn into their own AI-guided learning activity.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%