Rookie Mistakes: Measuring Software Quality in Student Projects to Guide Educational Enhancement

Marco De Luca, Sergio Di Martino, Sergio Di Meglio, Anna Rita Fasolino, Luigi Libero Lucio Starace, Porfirio Tramontana · 2025

Read Rookie Mistakes: Measuring Software Quality in Student Projects to Guide Educational Enhancement by Marco De Luca, Sergio Di Martino, Sergio Di Meglio, Anna Rita Fasolino, Luigi Libero Lucio Starace, Porfirio Tramontana (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-12488.txt.

Introduction

Marco De Luca and co-authors present a 2025 paper on measuring software quality in student projects. Courses push students to ship working programs, and quality lessons get squeezed. Teachers must choose which quality topics deserve scarce time, but evidence mostly covers tiny exercises by raw beginners.

The team analyzed eighty-three team projects written by 172 students across four runs of an object-oriented programming course. Two analysis tools scanned the code for smells and architectural anti-patterns. Recurring flaws stood out clearly at this intermediate level. The authors turn the patterns into concrete guidance for educators: which mistakes to teach against first, and how course design can steadily raise the quality bar.

Catalog record

CREATORMarco De Luca, Sergio Di Martino, Sergio Di Meglio, Anna Rita Fasolino, Luigi Libero Lucio Starace, Porfirio Tramontana
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE39 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-12488

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Rookie Mistakes: Measuring Software Quality in Student Projects to Guide Educational Enhancement Marco De Luca, Sergio Di Martino, Sergio Di Meglio, Anna Rita Fasolino, Luigi Libero Lucio Starace, Porfirio Tramontana Abstract When teaching Programming and Software Engineering in Bachelor’s Degree programs, the emphasis on creating functional software projects often overshadows the focus on software quality, a trend that aligns with ACM curricula recommendations. Software Engineering courses are typically introduced later in the curriculum, and can generally allocate only limited time to quality-related topics, leaving educators with the challenge of deciding which quality aspects to prioritize. In this decision, the literature offers limited guidance, as most existing studies focus on code written by novice students and small code units, making it unclear whether those findings extend to intermediate-level students with foundational object-oriented programming skills working on more complex software projects. To address this gap, we analyze 83 object-oriented team projects developed by 172 university students across 4 different editions of the Object-Oriented Programming course. We apply a static analysis pipeline used in prior research to assess software quality, combining SonarQube and ArchUnit to detect code smells and architectural anti-patterns. Our findings highlight recurring quality issues and offer concrete evidence of the challenges students face at this stage, providing valuable guidance for educators aiming to continuously improve Software Engineering curricula and promote quality-oriented development practices. 1 Introduction In Bachelor’s Degree programs, such as Computer Science and Computer Engineering, the focus of programming and software engineering education often leans more towards the ability to develop functional projects rather than emphasizing software quality. This aligns with the ACM Computer Science and Computer Engineering curriculum recommendations for undergraduate degrees [ 7 ] . According to these guidelines, software quality is typically only lightly addressed in three-year Bachelor’s programs. In these programs, introductory CS1 courses concentrate primarily on programming skills, with basic concepts of software quality introduced later, primarily in Software Engineering courses. However, even within Software Engineering courses, the time dedicated to software quality teaching is limited. Educators face the challenge of determining which aspects of software quality to prioritize within these constraints. Furthermore, there is a significant gap in the literature regarding research into the foundational software quality challenges specifically faced by intermediate-level students when developing software projects. This lack of insight leaves instructors of more advanced Software Engineering courses with little direction on how to prioritize quality aspects in their curricula. In this study, we present an empirical investigation of software quality in a set of 83 projects developed by 172 Computer Science students involved in an Object-Oriented Programming (OOP) course taken in two consecutive academic years. Our study leverages a static code analysis pipeline validated in prior research [ 9 ] , applying state-of-the-art tools such as SonarQube 1 1 1 SonarQube, available at https://www.sonarsource.com/products/sonarqube/ and ArchUnit 2 2 2 ArchUnit, available at https://www.archunit.org/ to detect code smells and architectural anti-patterns, two indicators of structural degradation and design compromise often referred to as social debt when accumulated in collaborative environments. By identifying recurring quality problems in student code, this work provides concrete evidence to support iterative course improvements aligned with real-world development standards [ 33 ] , as shown in Figure 1 . The insights gained from this analysis can guide educators in refining both introductory programming courses and more advanced Software Engineering modules, ultimately fostering a culture of quality awareness and reflective development practices in future software engineers. The remainder of the paper is structured as follows. Section 2 presents background information on software quality and related work. Section 3 describes the study we conducted, including detailed information about the OOP courses and the topics covered, allowing readers to compare them with their own teaching contexts. Section 4 presents the results, highlighting the most relevant findings, which are then discussed in Section 5 , with a focus on their implications for educators. Last, Section 6 outlines the threats to the validity of our findings, while Section 7 presents final remarks and directions for future work. Figure 1 : Cyclical evaluation process: the analysis of the quality of OOP students projects brings feedback useful to improve this course and also Software Engineering courses. 2 Background and Related Works In the following, we introduce key concepts related to software quality, focusing on both code-level and architectural aspects. We also review existing tools used for quality assessment and summarize prior research on software quality issues in student-developed code. 2.1 Software Quality Software quality captures how effectively a system satisfies its specified requirements and fulfills user expectations [ 12 ] . The ISO 25010 standard [ 16 ] provides a widely adopted model for assessing software quality, defining key attributes such as modularity, reusability, and testability [ 14 ] . These quality attributes are primarily shaped by two fundamental aspects: the clarity and maintainability of the source code, and the soundness of the software’s architectural design. Both play a critical role in ensuring the system remains robust, adaptable, and easy to evolve over time. To ensure and maintain these properties, various Software Quality Management (SQM) practices are employed throughout the development process. These include code reviews, automated testing, static code analysis, and architecture conformance checks [ 4 , 18 , 29 ] , often supported by tools such as SonarQube , CheckStyle , and similar analyzers. In recent years, SQM has also extended to the social dimension of software development. Community detection techniques have been used on software dependency graphs to identify architectural issues, and on developer collaboration networks to uncover coordination problems, knowledge silos, or uneven code ownership [ 28 , 27 , 10 , 5 ] . These analyses provide complementary insights to technical evaluations, enabling a more comprehensive understanding of software quality. Source Code Quality The notion of source code quality lacks a single, universally accepted definition and is often interpreted in various ways. One of the most well-known and widely used interpretations is the notion of Clean Code , as introduced by Martin et al. [ 23 ] . According to this perspective, clean code is characterized by its readability, clarity of intent, ease of writing, and maintainability. It is designed not only to be understood by compilers, but also to be easily interpreted by other developers. To support the development of such code, a variety of static analysis tools have been introduced and are commonly adopted in both industry and education. These tools are intended to automatically detect violations of clean code principles. A comparative study by Lenarduzzi et al. [ 21 ] evaluated several popular tools, including Better Code Hub , CheckStyle , Coverity Scan , FindBugs , PMD , and SonarQube . Their findings highlighted SonarQube as the most effective tool for detecting a broad range of code quality issues. Software Architectural Design The architecture of a software system, defined as its structural organization into components, their relationships, and the rules governing their interactions, plays a critical role in determining overall system quality [ 2 , 24 ] . A well-structured architecture contributes to building systems that are robust, scalable, and maintainable, while inadequate architectural decisions can lead to technical debt, reduced performance, and decreased reliability. To address recurring challenges in system design, software architectural patterns have been

Read the complete text