Large Language Models (LLMs) can be used to automate many aspects of the educational field. In this paper, we look into the benefits of automating responses to student questions in course discussion forums using our Retrieval-Augmented Generation (RAG)-based LLM pipeline (Edison). Our research questions are: RQ1 How do the responses generated by Edison compare to those of TAs in terms of level of detail and use of examples? RQ2 How does the tone of responses generated by Edison compare to that of TA responses? RQ3 Are responses generated by Edison more self-consistent than TA responses? Our results suggest that Edison generates responses with more detail, examples, positive tone, and self-consistency than TAs. We envision Edison being used as a baseline for TAs to build responses on, reduce response times, and promote equitable feedback.
This research examines the effects of late course enrollment, defined as enrolling after the first lecture, on student performance and the prevalence of Incomplete grades in a large introductory Computer Science (CS0) course at \university. Our CS0 course is unique in that Incomplete grades are awarded to students at risk of failing, allowing them to complete outstanding work in a subsequent term. We analyze whether there is a correlation between late enrollment, poor performance, likelihood of receiving an Incomplete grade, and course engagement. Specifically, we seek to answer the question: How does late enrollment affect students' likelihood of receiving Incomplete grades and their overall academic performance compared to on-time enrollees? Our findings reveal significant disparities in academic performance between on-time enrollees and late enrollees, indicating a need for course policy revisions.
This year's Snap! 9 release represents 10 years since Snap! 4.0 was initially released as a web application. Version 9 includes hundreds of new features, focused on providing students and teachers with new cloud tools, continued development of tools for working with data, and ''quality of life'' enhancements. The Snap! Cloud has been upgraded to support dedicated student and teacher accounts. Teachers can create accounts on behalf of their students, and manage students in their class. Everyone can also choose to ''follow'' users and see a feed of new projects shared with the community. It's also never been easier to work with data! Snap! 9 adds the ability to treat lists as ''dictionaries'' (lists of lists can now be indexed by arbitrary alphanumeric values). A number of blocks have been made variadic (i.e., supporting a variable number of inputs), saving the user the tedium of dragging over multiple copies of the same block. These include and, or, all comparison operators, and if, which adds more intermediate else if cases when the right arrow is clicked. Finally, Parsons Puzzles can be created with a single click. The Generate Puzzle command generates a microworld with only the blocks present in the selected sprite (and hides the answer behind a menu choice). In this demo, we'll share ways that you can use new Snap! features in the classroom, and provide attendees with the tools to build engaging experiences for students. All materials from this demonstration will be shared at https://bjc.link/snap-sigcse24