arXiv subjects

Ethan Dickey

Publications and source records attributed to Ethan Dickey.

15 recordsLinked to original sources

arXiv2026

CodeStylist: Supporting Early Undergraduate Programmers with Course-Aware Code Style Feedback

This innovative practice full paper presents CodeStylist, a web application that provides course-standard-aware code style feedback for early undergraduate programming courses. CodeStylist addresses a common instructional gap: students are expected to follow local conventions for naming, formatting, comments, organization, and readability, but feedback on these expectations is often delayed or inconsistent. Unlike generic linters or general-purpose LLM prompts, CodeStylist supports course-specific standards, multi-file submissions, and file- and line-localized explanations intended to guide revision rather than grade correctness. We report a formative expert review with 18 instructional staff from one early undergraduate programming course. Participants explored the prototype using self-selected code artifacts and completed a survey about response quality, anticipated student use, and redesign priorities. Ratings indicated modest perceived utility but limited trust: perceived correctness averaged 60.7%, response helpfulness averaged 3.50/5, response usefulness averaged 3.33/5, and anticipated student learning averaged 2.61/5. Despite these concerns, 17/18 respondents expected students to use the tool primarily for style checking, often at least weekly. Open-ended feedback showed that respondents valued CodeStylist for making implicit course standards more visible, but were concerned about unreliable output, overreliance, and latency or cost. We interpret these findings as evidence that course-aware style feedback is promising as a pre-submission revision aid, but that future versions should combine deterministic rule checks with LLM-generated explanations, rule citations, and stronger verification support.

cs.CY↗

arXiv2026

BoilerSketch: A TA-Supervised, Diagram-First GenAI Practice for Structured Diagrams in CS1/Early CS2

This innovative practice full paper presents BoilerSketch, a TA-supervised, diagram-first GenAI practice and tablet interface for providing structured visual explanations in CS1 and early CS2 support settings. Large early computing courses routinely face a support bottleneck during labs and office hours because many student questions are best answered with a diagram rather than additional text, yet most AI tutoring tools remain text-forward and unreliable at producing accurate, pedagogically useful visuals. BoilerSketch addresses this gap through a dual-pane interaction model that combines chat with a pen-enabled whiteboard for student sketches and a prompting strategy that constrains the model to generate structured, renderable Mermaid diagrams rather than free-form images. To preserve academic integrity, the system is intentionally scoped to conceptual explanation: it forbids executable code and code-level debugging and uses a human-in-the-loop workflow in which teaching assistants remain accountable supervisors who can monitor sessions and intervene when responses require correction, deeper probing, or escalation to live help. We report a 45-minute expert evaluation with 21 instructional staff from a large programming course who used BoilerSketch on representative questions and completed a post-use survey. Two-thirds rated the system at least moderately helpful for conceptual understanding and at least moderately useful for typical support tasks. Staff saw the strongest value in routine diagram-based explanations and noted limits in diagram depth and applicability to more advanced topics. We conclude with practical guidance for adopting supervised, diagram-first GenAI support in early computing courses, emphasizing scope-and-escalation rules, prompt-as-policy guardrails, and reliable structured diagram rendering.

cs.CY↗

arXiv2026

Is Solving Better Than Evaluating GenAI Solutions?

As Generative AI (GenAI) tools become increasingly capable of generating solutions to computing assignments, the computing education community is exploring pedagogical approaches that emphasize solution evaluation, verification, and critique alongside traditional solution generation. However, evidence regarding the impact of such evaluation-centered tasks on student learning remains limited, particularly in upper-division, theory-heavy courses. We conducted a randomized A/B crossover study (N=220) in a junior-level algorithms course to compare evaluating GenAI-generated solutions with traditional problem solving. Across six assignments, student working groups either solved challenging algorithmic problems directly or evaluated often-flawed GenAI-generated solutions, with roles reversed midway through the semester. We found no statistically significant differences between groups in midterm scores, final exam scores, overall course grades, or exam problems structurally aligned with the homework interventions. Students received significantly higher homework scores when evaluating GenAI-generated solutions, but this localized advantage did not translate into downstream summative gains. Survey data further indicated that most students reported no change in study habits in response to the intervention; however, those who reported adapting their study strategies rated the GenAI-evaluation assignments as significantly more helpful. These findings suggest that GenAI evaluation redistributes student effort from open-ended solution construction toward verification, diagnosis, and judgment, but does not automatically produce stronger conceptual transfer. We conclude that GenAI-evaluation activities can be incorporated into algorithms coursework without broad performance losses, but meaningful learning gains may require deliberate scaffolding that pushes students beyond simple error diagnosis.

cs.CY↗

arXiv2026

Stable Voting is PSPACE-Complete

Stable Voting and Simple Stable Voting, introduced by Holliday and Pacuit, are Condorcet-consistent voting rules defined recursively: a candidate wins if they would win after removing some opponent they beat, taking the pair with the largest margin first. The computational complexity of winner determination under these rules has been an open question. We resolve this problem: winner determination is PSPACE-complete under both Stable Voting and Simple Stable Voting.

cs.GT↗

arXiv2026

Universal Complex Quantum-Like Bits from Hermitian Weighted Graphs

We study when block-coupled regular graphs can realize prescribed complex quantum-like (QL) bit states as exact synchronized eigenstates. Two regular subgraphs $G_A$ and $G_B$ supply normalized all-ones eigenvectors $V_A$ and $V_B$, and algebraically regular bipartite couplings reduce the full graph-supported operator exactly to a $2\times 2$ effective block on $\mathcal S=\operatorname{span} \{ \lvert 0\rangle, \lvert 1\rangle \}$. Within this reduction we prove that two natural symmetric complexifications are not universal for state synthesis under a real-spectrum requirement: complex symmetric coupling with real diagonal regularities forces the target computational basis amplitude ratio $r=\omega_2/\omega_1$, for $\lvert \psi\rangle = \omega_1\lvert 0\rangle + \omega_2\lvert 1\rangle$, to satisfy $r^2\in\R$, while real symmetric coupling with complex diagonal regularities forces $r+1/r\in\R$. Replacing complex symmetry by Hermitian coupling removes this phase obstruction. For any nonbasis target state, any prescribed real eigenvalue, and any prescribed nonzero signed spectral gap, a Hermitian weighted coupling realizes the target exactly. Additionally, an independently tuned directed-coupling model gives a second universality mechanism. We then pass from continuous effective parameters to finite weighted graphs with entries in $\{0, \pm1, \pm i\}$ (the fourth roots of unity and zero), characterize the balanced discrete coupling lattice by perfect matchings, and show that exact discrete Hermitian realizations are dense in the synchronized pure-state space. These results give a state-synthesis universality taxonomy for complex QL-bits and identify Hermitian conjugate pairing as the robust structural mechanism that supports arbitrary complex amplitudes with real two-level spectra.

quant-ph↗

arXiv2025

Owlgorithm: Supporting Self-Regulated Learning in Competitive Programming through LLM-Driven Reflection

We present Owlgorithm, an educational platform that supports Self-Regulated Learning (SRL) in competitive programming (CP) through AI-generated reflective questions. Leveraging GPT-4o, Owlgorithm produces context-aware, metacognitive prompts tailored to individual student submissions. Integrated into a second- and third-year CP course, the system-provided reflective prompts adapted to student outcomes: guiding deeper conceptual insight for correct solutions and structured debugging for partial or failed ones. Our exploratory assessment of student ratings and TA feedback revealed both promising benefits and notable limitations. While many found the generated questions useful for reflection and debugging, concerns were raised about feedback accuracy and classroom usability. These results suggest advantages of LLM-supported reflection for novice programmers, though refinements are needed to ensure reliability and pedagogical value for advanced learners. From our experience, several key insights emerged: GenAI can effectively support structured reflection, but careful prompt design, dynamic adaptation, and usability improvements are critical to realizing their potential in education. We offer specific recommendations for educators using similar tools and outline next steps to enhance Owlgorithm's educational impact. The underlying framework may also generalize to other reflective learning contexts.

cs.CY↗

arXiv2025

Programmable Quantum-Like bits from Signed Regular Graphs

Extending upon observations of the emergence of quantum-like (QL) states from classical complex synchronized networks, this work adds mathematical rigor to the analysis of single QL bits constructed from adjacency-matrix eigenvectors. First, we rigorously show that symmetric construction of such networks (regular undirected/symmetric bipartite graph $G_C$ connecting two regular undirected subgraphs $G_A,\,G_B$) leads to an equal superposition of the $|+\rangle, |-\rangle$ Hadamard states (with basis $|0\rangle,\,|1\rangle$ set from eigenvectors of the subgraphs), and provide an analysis of sufficient conditions on the network for construction of such states. Second, we prove two methods to construct arbitrary single qubit states $|\psi\rangle = a|0\rangle + b|1\rangle,\, |a|^2+|b|^2=1$, and give switching lemmas for their boundaries: (i) by detuning the two subgraphs regularities and (ii) by asymmetrically allowing the bipartite connection matrix $C$ to be directed and detuning those regularities. Although motivated by using complex synchronized networks for quantum information storage and computation, the proofs for these methods rely only on the structure of the graph embedded in the adjacency matrix. Thus, synchronization is unnecessary; QL bits arise when edge weights are unit (or near-unit) and subgraphs are regular. Results on combinations of random k-regular graphs (more precisely Erd\H{o}s-R\'enyi graphs) may be independently interesting.

quant-ph↗

arXiv2025

The Failure of Plagiarism Detection in Competitive Programming

Plagiarism in programming courses remains a persistent challenge, especially in competitive programming contexts where assignments often have unique, known solutions. This paper examines why traditional code plagiarism detection methods frequently fail in these environments and explores the implications of emerging factors such as generative AI (genAI). Drawing on the author's experience teaching a Competitive Programming 1 (CP1) course over seven semesters at Purdue University (with $\approx 100$ students each term) and completely redesigning the CP1/2/3 course sequence, we provide an academically grounded analysis. We review literature on code plagiarism in computer science education, survey current detection tools (Moss, Kattis, etc.) and methods (manual review, code-authorship interviews), and analyze their strengths and limitations. Experience-based observations are presented to illustrate real-world detection failures and successes. We find that widely-used automated similarity checkers can be thwarted by simple code transformations or novel AI-generated code, while human-centric approaches like oral interviews, though effective, are labor-intensive. The paper concludes with opinions and preliminary recommendations for improving academic integrity in programming courses, advocating for a multi-faceted approach that combines improved detection algorithms, mastery-based learning techniques, and authentic assessment practices to better ensure code originality.

cs.CY↗

arXiv2025

Evaluating Performance Consistency in Competitive Programming: Educational Implications and Contest Design Insights

Competitive programming (CP) contests are often treated as interchangeable proxies for algorithmic skill, yet the extent to which results at lower contest tiers anticipate performance at higher tiers, and how closely any tier resembles the ubiquitous online-contest circuit, remains unclear. We analyze ten years (2015--2024) of International Collegiate Programming Contest (ICPC) standings, comprising five long-running superregional championships (Africa \& Arab, Asia East, Asia West, North America, and Northern Eurasia), associated local regionals of North America and Northern Eurasia, and the World Finals. For 366 World Finalist teams (2021--2024) we augment the dataset with pre-contest Codeforces ratings. Pairwise rank alignment is measured with Kendall's $\tau$. Overall, superregional ranks predict World Final ranks only moderately (weighted $\tau=0.407$), but regional-to-superregional consistency varies widely: Northern Eurasia exhibits the strongest alignment ($\tau=0.521$) while Asia West exhibits the weakest ($\tau=0.188$). Internal consistency within a region can exceed its predictive value for Worlds -- e.g., Northern Eurasia and North America regionals vs. superregionals ($\tau=0.666$ and $\tau=0.577$, respectively). Codeforces ratings correlate more strongly with World Final results ($\tau=0.596$) than any single ICPC tier, suggesting that high-frequency online contests capture decisive skill factors that many superregional sets miss. We argue that contest organizers can improve both fairness and pedagogical value by aligning problem style and selection rules with the formats that demonstrably differentiate teams, in particular the Northern-Eurasian model and well-curated online rounds. All data, scripts, and additional analyses are publicly released to facilitate replication and further study.

cs.CY↗

arXiv2025

A Scaffolded GenAI Lab in Early Undergraduate CS: A Mixed-Methods, Multi-Course Evaluation

Background and Context. Generative AI (GenAI) tools are increasingly used in programming courses, but we have limited evidence about how brief instruction can foster responsible, learning-oriented use. Objectives. We evaluate "AI-Lab", a scaffolded GenAI literacy intervention, asking how students' self-reported GenAI usage and their openness and comfort using GenAI for conceptual, debugging, and homework tasks change after participation. Methods. Across two semesters in three CS courses and one first-year engineering course at a U.S. university, we deployed the "AI-Lab" (pre-lab orientation, in-class critique of GenAI outputs, and a required homework reflection), collecting paired pre/post surveys (Perception N=831; Usage N=826) and six post-intervention focus groups; primary inferential analyses used the three CS courses (N=778 and 773, respectively). We analyzed survey shifts with paired non-parametric tests and focus groups via thematic analysis. Findings. Openness increased for conceptual questions and homework help, and comfort increased for conceptual, debugging, and homework scenarios; self-reported frequency of GenAI use for homework and projects remained stable, while self-reported use for debugging increased. Focus group participants described adopting more iterative prompting strategies, becoming more skeptical of correctness, and articulating clearer boundaries around integrity and dependence. Implications. A short, structured intervention can shift students' reported comfort with and willingness to use GenAI and influence the strategies they describe for engaging with it without increasing overall self-reported use on graded work. These results motivate future work triangulating surveys with behavioral traces and learning measures.

cs.CY↗

arXiv2024

BoilerTAI: A Platform for Enhancing Instruction Using Generative AI in Educational Forums

Contribution: This Full paper in the Research Category track describes a practical, scalable platform that seamlessly integrates Generative AI (GenAI) with online educational forums, offering a novel approach to augment the instructional capabilities of staff. The platform empowers instructional staff to efficiently manage, refine, and approve responses by facilitating interaction between student posts and a Large Language Model (LLM). This contribution enhances the efficiency and effectiveness of instructional support and significantly improves the quality and speed of responses provided to students, thereby enriching the overall learning experience. Background: Grounded in Vygotsky's socio-cultural theory and the concept of the More Knowledgeable Other (MKO), the study examines how GenAI can act as an auxiliary MKO to enrich educational dialogue between students and instructors. Research Question: How effective is GenAI in reducing the workload of instructional staff when used to pre-answer student questions posted on educational discussion forums? Methodology: Using a mixed-methods approach in large introductory programming courses, human Teaching Assistants (AI-TAs) employed an AI-assisted platform to pre-answer student queries. We analyzed efficiency indicators like the frequency of modifications to AI-generated responses and gathered qualitative feedback from AI-TAs. Findings: The findings indicate no significant difference in student reception to responses generated by AI-TAs compared to those provided by human instructors. This suggests that GenAI can effectively meet educational needs when adequately managed. Moreover, AI-TAs experienced a reduction in the cognitive load required for responding to queries, pointing to GenAI's potential to enhance instructional efficiency without compromising the quality of education.

cs.CY↗

arXiv2023

Quantum Voting and Violation of Gibbard-Satterthwaite's Impossibility Theorem

In the realm of algorithmic economics, voting systems are evaluated and compared by examining the properties or axioms they satisfy. While this pursuit has yielded valuable insights, it has also led to seminal impossibility results such as Arrow's and Gibbard-Satterthwaite's Impossibility Theorems, which pose challenges in designing ideal voting systems. Enter the domain of quantum computing: recent advancements have introduced the concept of quantum voting systems, which have many potential applications including in security and blockchain. Building on recent works that bypass Arrow's Impossibility Theorem using quantum voting systems, our research extends Quantum Condorcet Voting (QCV) to counter the Gibbard-Satterthwaite Impossibility Theorem in a quantum setting. To show this, we introduce a quantum-specific notion of truthfulness, extend ideas like incentive compatibility and the purpose of onto to the quantum domain, and introduce new tools to map social welfare functions to social choice functions in this domain.

cs.GT↗

arXiv2023

Innovating Computer Programming Pedagogy: The AI-Lab Framework for Generative AI Adoption

Over the last year, the ascent of Generative AI (GenAI) has raised concerns about its impact on core skill development, such as problem-solving and algorithmic thinking, in Computer Science students. Preliminary anonymous surveys show that at least 48.5% of our students use GenAI for homework. With the proliferation of these tools, the academic community must contemplate the appropriate role of these tools in education. Neglecting this might culminate in a phenomenon we term the "Junior-Year Wall," where students struggle in advanced courses due to prior over-dependence on GenAI. Instead of discouraging GenAI use, which may unintentionally foster covert usage, our research seeks to answer: "How can educators guide students' interactions with GenAI to preserve core skill development during their foundational academic years?" We introduce "AI-Lab," a pedagogical framework for guiding students in effectively leveraging GenAI within core collegiate programming courses. This framework accentuates GenAI's benefits and potential as a pedagogical instrument. By identifying and rectifying GenAI's errors, students enrich their learning process. Moreover, AI-Lab presents opportunities to use GenAI for tailored support such as topic introductions, detailed examples, corner case identification, rephrased explanations, and debugging assistance. Importantly, the framework highlights the risks of GenAI over-dependence, aiming to intrinsically motivate students towards balanced usage. This approach is premised on the idea that mere warnings of GenAI's potential failures may be misconstrued as instructional shortcomings rather than genuine tool limitations. Additionally, AI-Lab offers strategies for formulating prompts to elicit high-quality GenAI responses. For educators, AI-Lab provides mechanisms to explore students' perceptions of GenAI's role in their learning experience.

cs.CY↗

arXiv2023

GAIDE: A Framework for Using Generative AI to Assist in Course Content Development

This paper introduces "GAIDE: Generative AI for Instructional Development and Education," a novel framework for using Generative AI (GenAI) to enhance educational content creation. GAIDE stands out by offering a practical approach for educators to produce diverse, engaging, and academically rigorous materials. It integrates GenAI into curriculum design, easing the workload of instructors and elevating material quality. With GAIDE, we present a distinct, adaptable model that harnesses technological progress in education, marking a step towards more efficient instructional development. Motivated by the demand for innovative educational content and the rise of GenAI use among students, this research tackles the challenge of adapting and integrating technology into teaching. GAIDE aims to streamline content development, encourage the creation of dynamic materials, and demonstrate GenAI's utility in instructional design. The framework is grounded in constructivist learning theory and TPCK, emphasizing the importance of integrating technology in a manner that complements pedagogical goals and content knowledge. Our approach aids educators in crafting effective GenAI prompts and guides them through interactions with GenAI tools, both of which are critical for generating high-quality, contextually appropriate content. Initial evaluations indicate GAIDE reduces time and effort in content creation, without compromising on the breadth or depth of the content. Moreover, the use of GenAI has shown promise in deterring conventional cheating methods, suggesting a positive impact on academic integrity and student engagement.

cs.CY↗

arXiv2022

A Relative Church-Turing-Deutsch Thesis from Special Relativity and Undecidability

Beginning with Turing's seminal work in 1950, artificial intelligence proposes that consciousness can be simulated by a Turing machine. This implies a potential theory of everything where the universe is a simulation on a computer, which begs the question of whether we can prove we exist in a simulation. In this work, we construct a relative model of computation where a computable \textit{local} machine is simulated by a \textit{global}, classical Turing machine. We show that the problem of the local machine computing \textbf{simulation properties} of its global simulator is undecidable in the same sense as the Halting problem. Then, we show that computing the time, space, or error accumulated by the global simulator are simulation properties and therefore are undecidable. These simulation properties give rise to special relativistic effects in the relative model which we use to construct a relative Church-Turing-Deutsch thesis where a global, classical Turing machine computes quantum mechanics for a local machine with the same constant-time local computational complexity as experienced in our universe.

cs.CC↗