Basics: How to Do Office Hours and Tutoring Well
This TA training continued the discussion-format innovation. The "10-minute lecture" was replaced with brief self-introductions.
The introduction format: "I'm xxx, a TA for course xxx." Then answer a simple question about tutoring, e.g. "I prefer online/offline tutoring."
A series of discussion questions was prepared in advance:
- What's needed to do one tutoring session well? To schedule a whole semester of tutoring?
- Does "tutoring" also have a hierarchy? How to divide tutoring between LLMs, the department tutoring center, and course TAs?
- Online vs. offline tutoring: why do some teachers/TAs dislike WeChat-group tutoring? Is the office-hour system suitable for learning?
- If a questioning student lacks some prerequisite and it's hard to explain at once, what to do?
- If you suddenly remember a mistake in a previous answer, what to do?
- How to avoid being inadvertently passive-aggressive?
- Some students don't proactively seek TAs when stuck. What causes this? How to improve?
Finally, briefly shared the SIGCSE 2024 paper: Brief, Just-in-Time Teaching Tips to Support Computer Science Tutors.
This time we also tried, for some discussion questions, letting an LLM give an initial answer as a basis for TA discussion.
Related research (computing education & peer tutoring)
- Brief, Just-in-Time Teaching Tips to Support Computer Science Tutors (SIGCSE 2024) — short, immediately-usable teaching tips for TAs.
- Inside the Mind of a CS Undergraduate TA: A Firsthand Account of Undergraduate Peer Tutoring in Computer Labs (Julia M. Markel and Philip J. Guo, SIGCSE 2021, Best Paper Award) — insight into undergraduate TAs' thoughts and practices during peer tutoring in labs.
- A Study of Undergraduate Learning Assistants (ULAs) in Computer Science (SIGCSE 2024) — a study of undergraduate learning assistants.
- The Role of Probing and Clarifying Questions for Teaching Fellows in Computer Science — how TAs guide students' growth through probing and clarifying.
- Novices' Perceptions of Web-Search and AI for Programming (SIGCSE 2024) — how novices view using search engines and AI for programming, directly relevant to tutoring.
- Bringing a Visual Memory Model to VS Code — a tool bringing a "visual memory model" to VS Code, helping understand student code and debugging.
(Source: the tutor-related papers compiled for the 2024 Spring "How to do regular tutoring well.")