Three papers accepted to SIGCSE Technical Symposium 2027

I’m thrilled that I’m part of three papers accepted to the premier CS education conference, SIGCSE Technical Symposium 2027.

Two of my doctoral students, Kayleigh Stallings and Nicole Tian, each have lead-author papers.

The third paper is led by Lin Zhu, a doctoral student advised by my colleague Dr. Lijun Ni at the University at Albany.

1. CT-as-Direction: A Scoping Review of Computational Thinking in CS Education for AI-Assisted Coding by Kayleigh Stallings, Ismaila Temitayo Sanusi, and Fred Martin.

   Kayleigh’s insight is that as students use LLMs to write code we need new understandings of computational thinking and ways to observe and measure it. She presents a synthesizes 14 empirical studies in AI-assisted CS education and introduces the concept of “CT-as-direction,” arguing that student work in decomposing a problem, giving directions to an LLM, and evaluating whether AI output reflects the intended algorithm, are all forms of computational thinking.

   I’m excited about Kayleigh’s work and this paper. I think it’s poised to have real impact on our field.

2. Playing to Win or Learning to Understand: Lessons from a Q-learning Game for Middle Schoolers by Nicole Tian, Ismaila Temitayo Sanusi, and Fred Martin.

   Nicole led the design and implementation of AI Hoops, a web-based tool that introduces players to Q-learning via a fun and challenging basketball shooting game.

   The player attempts to make “hoops” by varying the angle and power of each shot. Then, then they can have a reinforcement learning agent take shots, causing a display of the Q-table that’s built.

   Many students moved from describing AI as learning from data toward describing it as improving through repeated training.

   Interestingly, students who were more focused on shooting baskets themselves exhibited less learning about how the AI works—hence the paper title, “Playing to Win or Learning to Understand.”

   Nicole’s game is available now for immediate online use: AI Hoops. Congratulations, Nicole!

3. Building Teacher AI Fluency in Middle Schools: Can Artificial Intelligence (AI) Enhance Subject-Area Understanding? by Lin Zhu, Ismaila Temitayo Sanusi, Hailey Muñiz, Deepti Tagare, Lucretia M. Fraga, Diane Schilder, Fred Martin, and Lijun Ni.

   This paper is the first full publication from our collaborative NSF-funded project, Building Capacity for Teacher and Student AI Fluency in Middle Schools in Texas and New York, where we are working with teachers to develop their understandings of AI while inviting them to create new activities for their students that unite concepts in AI with the subject-area material they teach.

   From our paper abstract:

   “Cross-case analysis revealed that teachers did more than simply add AI to their lessons. By engaging in model building and testing, teachers improve their understanding of how training data shapes outputs and why models can fail. Reviewing AI-generated outputs prompted discussions of accuracy, verification, and the role of human judgment. Ethics- and civics-oriented activities connected AI to concerns about broader social impacts, including concerns about privacy… Overall, the findings suggest that meaningful opportunities exist for teachers and students to engage with AI concepts within subject-area contexts.”

   The full team of co-authors contributed to the direct work with teachers, making meaning from the data we collected, and the paper itself. Thank you to my colleagues and our students: at the University at Albany, Dr. Lijun Ni and her doctoral student Lin Zhu; at UT San Antonio, Dr. Ismaila Sanusi, Dr. Deepti Tagare, and undergraduate student Hailey Muñiz; at the University of the Incarnate Word (San Antonio), Dr. Lucretia M. Fraga; and our project evaluator Diane Schilder. Thank you for leading the submission, Lin!

The papers will be presented at the 58th SIGCSE Technical Symposium in Sacremento, CA from February 17 through 20, 2027. See you there!