Seth Poulsen

Seth Poulsen, Shubhang Kulkarni, Geoffrey Herman et al.

Proof Blocks is a software tool that allows students to practice writing mathematical proofs by dragging and dropping lines instead of writing proofs from scratch. Proof Blocks offers the capability of assigning partial credit and providing solution quality feedback to students. This is done by computing the edit distance from a student's submission to some predefined set of solutions. In this work, we propose an algorithm for the edit distance problem that significantly outperforms the baseline procedure of exhaustively enumerating over the entire search space. Our algorithm relies on a reduction to the minimum vertex cover problem. We benchmark our algorithm on thousands of student submissions from multiple courses, showing that the baseline algorithm is intractable, and that our proposed algorithm is critical to enable classroom deployment. Our new algorithm has also been used for problems in many other domains where the solution space can be modeled as a DAG, including but not limited to Parsons Problems for writing code, helping students understand packet ordering in networking protocols, and helping students sketch solution steps for physics problems. Integrated into multiple learning management systems, the algorithm serves thousands of students each year.

Gavin Eddington, Christopher Warren, Seth Poulsen et al.

Background and Context. Computer programming often involves extended periods of sustained activity and mobile phone notifications introduce frequent opportunities for interruption. Prior work demonstrates that suppressing phone notifications may reduce these disruptions. Objectives. Our primary research question is: How does suppressing phone notifications affect students' task engagement and productivity while programming? Method. We report on a replication and methodological extension study conducted in a CS1 course involving 22 students. Using a within-subject design, selected programming assignments were randomly designated for enabling notification suppression. Phone state logs were synchronized with millisecond-resolution IDE keystroke data to measure student attention and focus when in the control and notification-suppression conditions. Findings. Assignments completed with notification suppression enabled significantly lower break rates and longer intervals of focus compared to assignments completed in the control condition for many, but not all, students. This study provides evidence that notification suppression is associated with measurable differences in programming engagement and behavior. We also find a remarkable bimodality in the effect across students -- many students are positively affected, a small number are negatively affected, and very few experience little or no effect. This finding is consistent with other studies in diverse disciplines. Implications. Our results show that, for many students, phone notification suppression tools, such as Do Not Disturb, can improve attention and focus. Implications apply to educational settings (do-not-disturb as an intervention) and scholarship (understanding the effects of phone distraction).

Peter Fowles, Erik Falor, Sulove Bhattarai et al.

Background and Context: Large Language Models (LLMs) are more accessible and accurate than ever before, raising significant concerns for computing educators. One major concern is students using LLMs to bypass the effort needed to understand concepts and metacognitive strategies essential for success in computer science. Objectives: We contribute a unique approach to assessing and building up student understanding through weekly oral code review assessments. These formative assessments incentivize students to understand their submitted code, regardless of whether or not the code was generated by AI tools. We also use a flipped classroom to provide time for students to learn concepts outside of class and provide ample time for students to schedule code review interviews. Methods: For this paper, we collected data from three semesters. We analyze student exam scores, keystroke logs, and surveys to understand how the new course policies affected student learning, behavior, and attitudes. Findings: Pairwise comparison of exam results reveals a statistically insignificant increase in average scores for Fall 2025 compared to previous semesters. Keystroke logs show a significant increase in characters pasted per total characters input into coding assignments in Fall 2025, pointing towards higher AI usage. Survey results show positive student sentiment towards code reviews at the end of Fall 2025, with nearly all negative feedback being addressable through better scheduling and more rigorous TA training. Implications: Oral code reviews with a flipped classroom appear to be effective at mitigating harms of LLM use while providing space for students to freely experiment with these tools. Our work suggests that students in Fall 2025 still show adequate understanding of material covered in written exams, despite dramatic increases in LLM usage for coding assignments.

Chenyan Zhao, Mariana Silva, Seth Poulsen

Providing evaluations to student work is a critical component of effective student learning, and automating its process can significantly reduce the workload on human graders. Automatic Short Answer Grading (ASAG) systems, enabled by advancements in Large Language Models (LLMs), offer a promising solution for assessing and providing instant feedback for open-ended student responses. In this paper, we present an ASAG pipeline leveraging state-of-the-art LLMs. Our new LLM-based ASAG pipeline achieves better performances than existing custom-built models on the same datasets. We also compare the grading performance of three OpenAI models: GPT-4, GPT-4o, and o1-preview. Our results demonstrate that GPT-4o achieves the best balance between accuracy and cost-effectiveness. On the other hand, o1-preview, despite higher accuracy, exhibits a larger variance in error that makes it less practical for classroom use. We investigate the effects of incorporating instructor-graded examples into prompts using no examples, random selection, and Retrieval-Augmented Generation (RAG)-based selection strategies. Our findings indicate that providing graded examples enhances grading accuracy, with RAG-based selection outperforming random selection. Additionally, integrating grading rubrics improves accuracy by offering a structured standard for evaluation.

Chenyan Zhao, Mariana Silva, Seth Poulsen

In mathematical proof education, there remains a need for interventions that help students learn to write mathematical proofs. Research has shown that timely feedback can be very helpful to students learning new skills. While for many years natural language processing models have struggled to perform well on tasks related to mathematical texts, recent developments in natural language processing have created the opportunity to complete the task of giving students instant feedback on their mathematical proofs. In this paper, we present a set of training methods and models capable of autograding freeform mathematical proofs by leveraging existing large language models and other machine learning techniques. The models are trained using proof data collected from four different proof by induction problems. We use four different robust large language models to compare their performances, and all achieve satisfactory performances to various degrees. Additionally, we recruit human graders to grade the same proofs as the training data, and find that the best grading model is also more accurate than most human graders. With the development of these grading models, we create and deploy an autograder for proof by induction problems and perform a user study with students. Results from the study shows that students are able to make significant improvements to their proofs using the feedback from the autograder, but students still do not trust the AI autograders as much as they trust human graders. Future work can improve on the autograder feedback and figure out ways to help students trust AI autograders.

Seth Poulsen, Mahesh Viswanathan, Geoffrey L. Herman et al.

Proof Blocks is a software tool which enables students to write proofs by dragging and dropping prewritten proof lines into the correct order. These proofs can be graded completely automatically, enabling students to receive rapid feedback on how they are doing with their proofs. When constructing a problem, the instructor specifies the dependency graph of the lines of the proof, so that any correct arrangement of the lines can receive full credit. This innovation can improve assessment tools by increasing the types of questions we can ask students about proofs, and can give greater access to proof knowledge by increasing the amount that students can learn on their own with the help of a computer.

Alan T. Sherman, Geoffrey L. Herman, Linda Oliva et al.

We reflect on our ongoing journey in the educational Cybersecurity Assessment Tools (CATS) Project to create two concept inventories for cybersecurity. We identify key steps in this journey and important questions we faced. We explain the decisions we made and discuss the consequences of those decisions, highlighting what worked well and what might have gone better. The CATS Project is creating and validating two concept inventories---conceptual tests of understanding---that can be used to measure the effectiveness of various approaches to teaching and learning cybersecurity. The Cybersecurity Concept Inventory (CCI) is for students who have recently completed any first course in cybersecurity; the Cybersecurity Curriculum Assessment (CCA) is for students who have recently completed an undergraduate major or track in cybersecurity. Each assessment tool comprises 25 multiple-choice questions (MCQs) of various difficulties that target the same five core concepts, but the CCA assumes greater technical background. Key steps include defining project scope, identifying the core concepts, uncovering student misconceptions, creating scenarios, drafting question stems, developing distractor answer choices, generating educational materials, performing expert reviews, recruiting student subjects, organizing workshops, building community acceptance, forming a team and nurturing collaboration, adopting tools, and obtaining and using funding. Creating effective MCQs is difficult and time-consuming, and cybersecurity presents special challenges. Because cybersecurity issues are often subtle, where the adversarial model and details matter greatly, it is challenging to construct MCQs for which there is exactly one best but non-obvious answer. We hope that our experiences and lessons learned may help others create more effective concept inventories and assessments in STEM.