arXiv subjects

Chun-Yen Chang

Publications and source records attributed to Chun-Yen Chang.

3 recordsLinked to original sources

arXiv2025

ChatGPT and Gemini participated in the Korean College Scholastic Ability Test -- Earth Science I

The rapid development of Generative AI is bringing innovative changes to education and assessment. As the prevalence of students utilizing AI for assignments increases, concerns regarding academic integrity and the validity of assessments are growing. This study utilizes the Earth Science I section of the 2025 Korean College Scholastic Ability Test (CSAT) to deeply analyze the multimodal scientific reasoning capabilities and cognitive limitations of state-of-the-art Large Language Models (LLMs), including GPT-4o, Gemini 2.5 Flash, and Gemini 2.5 Pro. Three experimental conditions (full-page input, individual item input, and optimized multimodal input) were designed to evaluate model performance across different data structures. Quantitative results indicated that unstructured inputs led to significant performance degradation due to segmentation and Optical Character Recognition (OCR) failures. Even under optimized conditions, models exhibited fundamental reasoning flaws. Qualitative analysis revealed that "Perception Errors" were dominant, highlighting a "Perception-Cognition Gap" where models failed to interpret symbolic meanings in schematic diagrams despite recognizing visual data. Furthermore, models demonstrated a "Calculation-Conceptualization Discrepancy," successfully performing calculations while failing to apply the underlying scientific concepts, and "Process Hallucination," where models skipped visual verification in favor of plausible but unfounded background knowledge. Addressing the challenge of unauthorized AI use in coursework, this study provides actionable cues for designing "AI-resistant questions" that target these specific cognitive vulnerabilities. By exploiting AI's weaknesses, such as the gap between perception and cognition, educators can distinguish genuine student competency from AI-generated responses, thereby ensuring assessment fairness.

cs.AI↗

arXiv2025

Students' Acceptance of Arduino Technology Integration in Student-Led Science Inquiry: Insights from the Technology Acceptance Model

This study examines high school students' acceptance of Arduino technology in a student-led, inquiry-based science class, using the extended Technology Acceptance Model (TAM2) as a guiding framework. Through qualitative analysis of interviews and classroom observations, we explored how students perceived Arduino's usefulness and ease of use. Going beyond traditional quantitative TAM studies, this qualitative TAM research provides a nuanced, in-depth understanding of the contextual factors shaping technology acceptance. Key findings reveal that acceptance was driven not only by instrumental factors like job relevance and output quality but also by the unique sociocultural context of the Korean education system, where technology use was perceived as valuable for university admissions (subjective norm and image). Critically, unlike earlier research that emphasized programming challenges, participants in this study found Arduino accessible and intuitive, thanks to integrated visual block-coding tools. These findings highlight the importance of both technological design and pedagogical support in shaping students' experiences. Implications for science curriculum design, teacher preparation, and equitable technology integration in secondary education are discussed.

cs.HC↗

arXiv2023

Investigating Effects of Perceived Technology-enhanced Environment on Self-regulated Learning: Beyond P-values

This study examined the effects of a technology-enhanced intervention on the self-regulation of 262 eighth-grade students, employing information and communication technology (ICT) and web-based self-assessment tools set against science learning. The data were analyzed using both maximum likelihood and Bayesian structural equation modeling to unravel the intricate relationships between self-regulation, self-efficacy, perceptions of ICT, and self-assessment tools. Our research findings underscored the direct and indirect impacts of self-efficacy, perceived ease of use, and perceived use of technology on self-regulation. The results revealed the predictive power of self-assessment tools in determining self-regulation outcomes, underlining the potential of technology-enhanced self-regulated learning environments. The study posited the necessity to transcend mere technology incorporation and to emphasize the inclusion of monitoring strategies explicitly designed to augment self-regulation. Interestingly, self-efficacy appeared to indirectly influence self-regulation outcomes through perceived the use of technology rather than direct influence. Analytically, this research indicated that Bayesian estimation could offer a more comprehensive insight into structural equation modeling by more accurately assessing our estimates' uncertainty. This research substantially contributes to comprehending the influence of technology-enhanced environments on students' self-regulated learning, stressing the importance of constructing practical tools explicitly designed to cultivate self-regulation.

cs.HC↗