Jul 2026· Annual Conference on Innovation and Technology in Computer Science Education· pp. 443-449· 0 citations· 48 references
Computer Science
TL;DR
This paper empirically study how different algorithmic implementations of the same underlying task affect students' ability to reason about good abstractions through method extraction, and shows that interleaved functional composition is more difficult to decompose.
Abstract
Program decomposition is a core skill in computer science that overlaps with program comprehension, refactoring, and design. Decomposition takes many forms, with one common task at the introductory level being the identification and extraction of meaningful abstraction into separate functions. However, there is limited empirical evidence about how the algorithmic structuring of code affects the difficulty of producing a meaningful decomposition. In this paper, we empirically study how different algorithmic implementations of the same underlying task affect students' ability to reason about good abstractions through method extraction. Grounded in a recent framework on code structuring, we design three functionally equivalent versions of the same task that differ only in how two latent functional patterns relate to each other: sequentially, hierarchically, or interleaved. Using a large-scale controlled study with 994 introductory programming students, we examine students' effort in decomposing the programs, the approaches they adopt, and their perceptions of whether the resulting decomposition improves readability and understanding. Our results show that interleaved functional composition is more difficult to decompose, providing initial empirical support for the framework's hypothesized ordering of composition pattern difficulty. Students also perceive their decompositions of the interleaved version to be less easy to read and understand. We also find that students describe many different approaches of decomposing, and we discuss several implications for teaching and future research.
Introduction We investigate how expressive inline code comments written in various developer styles, functional to progressively poetic, philosophical, and misleading, affect large language model (LLM) behavior during code optimization. Methods In this pilot study, we used a controlledmerge sort implementation across five stylistic variants and evaluated GPT-5 and Claude Opus 4.1 under standardized console prompts, isolating the effect of embedded comment semiotic variation. Seven expert developers (three senior, four mid-level) scored model outputs against adapted ISO/IEC 25010 criteria and novel LLM suggestibility index (LSI) framework. Results Semiotic character of comments measurably altered code quality, with consensus-score reliability ICC(2, k) = 0.65–0.81 for six of seven dimensions; single-rater Krippendorff's α = 0.232 reflects substantial interpretive variability. Claude exhibited higher interpretive sensitivity (mean behavioral divergence 4.00; SD 1.16), while GPT-5 maintained stronger architectural fidelity (mean divergence 3.58; SD 1.26). Reflective comments (philosophical, conversational) were associated with Claude's highest maintainability scores in our panel (both M = 4.00, ~8% above stock M = 3.71), while the same philosophical comments reduced GPT-5 maintainability (M = 2.86), suggesting asymmetric model responses to expressive context. Conclusions These findings position inline comments as model-sensitive latent semantic prompts, with implications for AI-in-the-loop development and design of comment conventions for AI-assisted maintenance.
Angela N. Johnson· Frontiers in Artificial Inte...· 0 citations
It is demonstrated that models and code achieve comparable overall correctness, and thus models alone may be sufficient in model-centric scenarios where access to code is limited or unavailable, and a consistent structure-behavior comprehension gap is revealed.
Iris Reinhartz-Berger, Monique Snoeck· Journal of Software and Syst...· 0 citations
Conceptual modeling supports the design, analysis, and communication of the properties of complex systems, yet conceptual models can be difficult to understand when domain-level abstractions must be encoded through low-level constructs required mainly for semantic conformity. Prior work has mainly improved how existing individual constructs are visually represented. We shift the focus from individual constructs to recurring configurations of constructs, and propose abstract notation engineering as a language-agnostic method for replacing such configurations with higher-level, semantically transparent constructs. The method comprises pattern identification, pattern formalization, visual notation design, and empirical validation. We instantiate it for Dynamic Condition Response (DCR) graphs, where common workflow patterns require elaborate low-level configurations. The resulting extension, DeCleaR, replaces such configurations with compact pattern-based abstractions. The results of our empirical validation show that DeCleaR improves perceived empirical quality, pragmatic quality, and user preference over standard DCR graphs.
Amine Abbad-Andaloussi, D. Schütt, Kasper Solhøj Jørgensen et al.· arXiv.org· 0 citations
This study empirically evaluates whether cost-efficient Large Language Models (LLMs) can be trusted to generate enterprise code to a written specification. Three models (Gemini Flash 3, GPT-5.4 mini and Claude Haiku 4.5) were asked to solve 992 algorithmic problems as Java Spring Boot service methods conforming to a mandated signature and data-transfer-object specification, crossing four model and agentic coding tool combinations with two prompt variants to yield eight configurations, with iteration forbidden and hardcoded answers explicitly prohibited. Eight problem statements were withheld to probe how models respond to missing input. The 7,593 resulting methods were classified by an eight-class outcome taxonomy describing what each does about producing an answer, then deployed and executed, giving 7,936 measured requests joined to that classification. Structural conformance approached ceiling, yet 38.4% of methods do not compute the value they returned and only 12.9% of returned answers were correct. Conditioning on outcome class shows that response reliability and correctness are inversely related, whereas genuinely computing methods answered least often and were correct 19.3%. Limitations include single generation runs per configuration, partial harness coverage, single-pass timing, syntactic classification, and probable corpus contamination.
