Factors Influencing AI-Assisted Thesis Writing in University: A Pull-Push-Mooring Theory Narrative Inquiry Study

Data & Metadata, Год журнала: 2025, Номер 4, С. 203 - 203

Опубликована: Фев. 10, 2025

This study aims to examine the factors that motivate, attract, and anchor students adopt AI tools during writing process in context of push-pull-mooring (PPM) theory. Utilizing a narrative inquiry research approach, this employed observation, in-depth interviews, document analysis for data collection. The identified key through reflexive thematic methods. Key pull include generation credit authorship contributions integration into academic writing. encompass topic selection, dynamic literature review, questions, proposal conceptualization, designing methods, analysis, revising drafts, managing references. incorporates active learning, self-regulated learning (SRL), inquiry-based overcoming linguistic challenges. push reference inaccuracies, confidentiality research, overreliance on AI. Three anchoring principles guide ethical incorporation thesis writing: institutional policies, augmentation, comprehensive contextual approach. But study's limitations small sample size ten from single university, which affects generalizability results.

Язык: Английский

---

The impact of guilt on student interactions with generative AI technology

Ethics & Behavior, Год журнала: 2025, Номер unknown, С. 1 - 27

Опубликована: Фев. 23, 2025

Язык: Английский

---

Students’ Perceptions of Artificial Intelligence Integration in Higher Education

European Journal of Educational Research, Год журнала: 2025, Номер 14(2), С. 471 - 484

Опубликована: Март 6, 2025

This study explores the impact of artificial intelligence (AI) integration on students' educational experiences. It investigates student perceptions AI across various academic aspects, such as module outlines, learning outcomes, curriculum design, instructional activities, assessments, and feedback mechanisms. evaluates experiences, critical thinking, self-assessment, cognitive development, integrity. research used a structured survey distributed to 300 students through Microsoft Forms 365, yet response rate was 29.67%. A thematic analysis were employed gather insights from 89 students. Thematic is qualitative method for identifying analysing patterns or themes within data, providing into key ideas trends. The limited may be attributed learners' cultural backgrounds, not all are interested in familiar with tools. questions about different areas. identify data. Benefits enhanced timely feedback, personalised experiences prevalent. tools like Turnitin supported integrity, platforms ChatGPT Grammarly particularly valued their utility tasks. acknowledges limitations linked small sample size focus undergraduate learners only. findings suggest that can significantly improve provides tailored support promotes ethical practices. recommends continued expanded use technologies education while addressing potential implementation challenges.

Язык: Английский

---

Evaluating the efficacy of ChatGPT in environmental education: findings from heuristic and usability assessments

On the Horizon The International Journal of Learning Futures, Год журнала: 2025, Номер unknown

Опубликована: Март 8, 2025

Purpose This study aims to investigate ChatGPT’s potential in environmental education concerning sustainable development goals. Heuristic evaluation and usability testing identify critical issues, including inadequate multimedia support, language barriers insufficient fact-checking capabilities. Design/methodology/approach The uses heuristic assess efficacy at a Chinese higher institution. identifies essential limitations, reliance on text-only resources, absence of assets, technical deficiencies, barriers, lack tools, context-related delayed information, inconsistency limited expertise. Data was collected through quantitative qualitative analysis, with input from experts students. Findings suggest that while ChatGPT offers opportunities for interactive learning, its limitations hinder comprehensive educational outcomes. A proposed hybrid model combining generative AI human elements enhance engagement learning effectiveness. research foundation integrating tools into education, addressing gaps fostering environments. Originality/value contributes deeper understanding the role artificial intelligence underscores importance incorporating intervention. approach framework creating more meaningful environments by leveraging unique strengths alongside technology.

Язык: Английский

---

Exploring the Scientific Validity of ChatGPT’s Responses in Elementary Science for Sustainable Education

Sustainability, Год журнала: 2025, Номер 17(7), С. 2962 - 2962

Опубликована: Март 27, 2025

As AI integration in education increases, it is crucial to evaluate its effectiveness elementary science learning, particularly promoting sustainable through equitable access knowledge. This study aims assess the validity and applicability of ChatGPT3.5 (free version) responses Earth Space science. A document analysis 1200 AI-generated was conducted scientific validity, explanatory clarity, pedagogical relevance. The employed quantitative methods accuracy alignment with curricula, while qualitative insights identified linguistic conceptual challenges. findings indicate that 94.2% were scientifically valid, 70.6% clear, but only 12.8% aligned curricula. While ChatGPT provides accurate information, many included complex terminology unsuitable for young learners. Additionally, 87.2% lacked posing challenges effective classroom integration. Despite these limitations, shows potential simplifying concepts expanding educational resources. Refining content curriculum-based filtering, adaptive language processing, teacher mediation necessary. Strengthening AI-driven strategies a sustainability focus can ensure long-term improvements learning. highlights need further research on optimizing tools education.

Язык: Английский

---

CTRL+ Ethics

Oxford University Press eBooks, Год журнала: 2025, Номер unknown

Опубликована: Апрель 3, 2025

Abstract When crafting argumentative essays, increasingly, students use large language models (LLMs) to help with writing, and we notice this happening taking ethics classes as part of their professional education. This chapter argues that relying too much on LLMs can lead moral deskilling because writing texts shapes skills, including sensitivity imagination future professionals. The development sharpening skills rely effort, which undermine by design. Furthermore, the struggle develop an argument find right words, structure, tone contributes relevant like resilience grit. It is itself leads a better understanding how difficult it make ethical arguments, contribute more complex perspective-taking in reflection. upshot replacing written assignments education other formats, oral assessments or practical assignments, risks losing valuable format for skill development. ends advice educators dealing student’s tendency delegate without giving up formative assignments.

Язык: Английский

---

Ethical and Privacy Considerations in AI-Driven Language Learning

LatIA, Год журнала: 2025, Номер 3, С. 328 - 328

Опубликована: Март 29, 2025

Artificial intelligence (AI) has revolutionized language learning by enabling personalized and adaptive education; however, these advancements also raise ethical privacy concerns, including algorithmic bias, data security risks, a lack of transparency in AI-driven decision-making. This study examines challenges, focusing on fairness, linguistic diversity, the balance between automated human instruction, with goal proposing guidelines for responsible adoption AI education. Through literature review comparative analysis, risks AI-powered tools were explored, assessing bias detection algorithms, frameworks, privacy-preserving techniques to identify best practices. The findings indicate that tend exhibit biases disadvantage underrepresented groups, raising concerns about fairness while exposing due inadequate measures. Implementing frameworks incorporate fairness-aware explainable models, robust protection mechanisms enhances user trust security. Therefore, addressing issues is essential ensuring integration education, where hybrid approach combining instruction emerges as most solution. Lastly, future research should focus regulatory compliance models strengthen ethics

Язык: Английский

---

ChatGPT in Education: A Review of Ethical Challenges and Approaches to Enhancing Transparency and Privacy

Procedia Computer Science, Год журнала: 2025, Номер 254, С. 181 - 190

Опубликована: Янв. 1, 2025

Язык: Английский

---

Analysing nontraditional students' ChatGPT interaction, engagement, self‐efficacy and performance: A mixed‐methods approach

British Journal of Educational Technology, Год журнала: 2025, Номер unknown

Опубликована: Апрель 9, 2025

Generative artificial intelligence brings opportunities and unique challenges to nontraditional higher education students, stemming, in part, from the experience of digital divide. Providing access practice is critical bridge this divide equip students with needed competencies. This mixed‐methods study investigated how interact ChatGPT multiple courses examined relationships between interactions, engagement, self‐efficacy performance. Data were collected 73 undergraduate graduate through chat logs, course reflections artefacts, surveys interviews. interactions analysed using four metrics: prompt number, depth knowledge (DoK), relevance originality. Results showed that numbers ( β = 0.256, p < 0.03) engagement 0.267, 0.05) significantly predicted performance, while did not. Students' DoK r 0.40, 0.01) 0.42, positively correlated Text mining analysis identified distinct interaction patterns, ‘strategic inquirers’ demonstrating performance than ‘exploratory more sophisticated follow‐up questioning. Qualitative findings revealed most first‐time users who initially resistance, they developed growing acceptance. Still, tended use sparingly and, even then, as only a starting point for assignments. The highlights need targeted guidance engineering AI literacy training help leverage effectively higher‐order thinking tasks. Practitioner notes What already known about topic Nontraditional face education, such limited technological access. emergence generative tools presents both addressing educational disparities. Existing studies on implementation predominantly focus traditional students. paper adds Empirical evidence metrics (prompt DoK, originality). Distinct patterns their relationship outcomes. among Implications and/or policy Need explicit instruction skill thinking. Importance providing technology self‐paced learning resources Value developing comprehensive addresses tool capabilities limitations.

Язык: Английский

---

AI-Powered Tools for Linguistic and Sentiment Analysis in Language Learning

Advances in computational intelligence and robotics book series, Год журнала: 2025, Номер unknown, С. 67 - 112

Опубликована: Янв. 31, 2025

This chapter introduces two AI-powered tools that enhance Systemic Textual Analysis (STA) for language learning and research. Built with Google Apps Scripts OpenAI's GPT API, the automate detailed analysis, offering feedback on schematic structure, clause complexity, thematic progression. The first tool provides discursive analysis—examining deixis, speech acts, coherence—benefiting junior researchers linguists. second offers multilingual sentiment analysis in English Arabic, enabling cross-linguistic comparisons of linguistic emotional tones. Together, these integrate structural insights, fostering nuanced text comprehension. Rigorous coding validation ensure consistency accuracy, highlighting tools' potential as scalable educational technologies support STA, improve skills, make advanced widely accessible.

Язык: Английский

---