Literature Review Example: Exploring L2 Master’s Students’ and Teachers’ Perceptions of Generative AI in Thesis Writing in Algeria

Exploring L2 Master’s Students’ and Teachers’ Perceptions of Generative AI in Thesis Writing in Algeria

1. Introduction

1.1 Context and significance of generative AI in thesis writing

Generative AI tools, such as advanced language models, have revolutionized the academic writing landscape by offering capabilities that support drafting, editing, and idea generation. Within the context of thesis writing, these technologies can streamline complex processes such as literature integration, linguistic refinement, and structural organization. For L2 (second language) learners undertaking master’s‐level theses, generative AI holds particular promise in mitigating linguistic barriers, accelerating research workflows, and enhancing the overall quality of written output. In Algeria, where students often confront dual challenges of mastering academic discourse in English or French and navigating rigorous research requirements, the advent of generative AI introduces both opportunities for pedagogical innovation and questions regarding academic integrity.

1.2 Purpose, scope, and research questions

This literature review aims to synthesize existing insights on L2 master’s students’ and teachers’ perceptions of generative AI deployment in thesis writing within the Algerian higher education context. The scope encompasses attitudinal dimensions such as perceived usefulness, potential drawbacks, and pedagogical implications. The central research questions are: (1) How do L2 master’s students perceive the benefits and challenges of using generative AI in thesis composition? (2) What are teachers’ attitudes toward the integration of these tools in supervisory and instructional practices? (3) Where do students’ and teachers’ views converge or diverge regarding the ethical and educational implications of generative AI adoption?

1.3 Definitions of key terms (L2 learners, generative AI, thesis)

L2 learners refers to individuals undertaking academic study in a language that is not their native tongue, typically involving specialized linguistic and cognitive demands. Generative AI denotes algorithmic systems, often powered by deep learning architectures, capable of autonomously producing coherent text based on user prompts. A thesis is an extended, original research document submitted as a capstone requirement for a master’s degree, demonstrating the student’s capacity for scholarly inquiry, methodological rigor, and academic writing proficiency.

Note: This section includes information based on general knowledge, as specific supporting data was not available.

2. Theoretical Background

2.1 Overview of generative AI technologies in education

Generative AI technologies encompass a range of natural language generation (NLG) systems that facilitate content creation through predictive text algorithms. Educationally, these tools have been leveraged for automated feedback, writing prompts, and personalized learning experiences. Their integration into academic writing environments can afford real‐time stylistic and grammatical suggestions, thereby potentially reducing cognitive load and reinforcing writing conventions. However, the rapid evolution of these systems also presents challenges related to overreliance, diminished critical thinking, and uneven access across diverse institutional settings.

2.2 L2 writing theories and thesis composition challenges

L2 writing research emphasizes the multifaceted process of acquiring target‐language proficiency, integrating factors such as interlanguage development, transfer from the first language, and sociocultural influences. Master’s thesis composition extends these theoretical concerns by demanding advanced rhetorical competence, discipline‐specific conventions, and methodological precision. Common challenges include lexical limitations, syntactic complexity, and the organization of scholarly arguments, all of which can be compounded by the high stakes associated with graduate‐level research output.

2.3 Algerian higher education context and policy

Algerian higher education has undergone policy reforms aimed at aligning with global standards and promoting research capacity, including the introduction of the Bologna Process-inspired license-master-doctorat system. Despite these reforms, disparities in resource allocation, language support infrastructure, and digital literacy training persist across universities. Generative AI adoption remains nascent, with limited formal guidelines governing its use in academic writing and assessment, thereby leaving stakeholders to negotiate ethical and pedagogical norms independently.

Note: This section includes information based on general knowledge, as specific supporting data was not available.

3. Key Findings from Literature

3.1 L2 master’s students’ perceptions of generative AI

Scholarly discourse indicates that L2 master’s students often view generative AI as a valuable scaffold for linguistic expression and structural articulation in thesis writing. Many report increased confidence in drafting complex sections, particularly in literature reviews and methodological descriptions. Conversely, some express concerns that reliance on AI‐generated text may impede the development of independent writing skills and engender superficial engagement with scholarly sources.

3.2 Teachers’ attitudes and experiences

Educators supervising L2 students display a nuanced stance toward generative AI: they appreciate its potential to enhance students’ draft quality and to streamline revision cycles, yet they voice apprehensions regarding academic integrity, originality, and the erosion of traditional mentorship roles. Teachers also note the absence of institutional frameworks to guide acceptable use, leading to ad hoc decision-making and varied supervisory expectations across departments.

3.3 Points of convergence and divergence

Students and teachers commonly acknowledge the pragmatic benefits of generative AI for overcoming language barriers and expediting writing tasks. However, divergence arises over questions of ownership and authenticity: students may perceive AI assistance as akin to consulting a tutor or reference work, while teachers may frame it as a form of unauthorized aid. This tension underscores the need for shared guidelines that reconcile enhancement of language support with preservation of scholarly integrity.

Note: This section includes information based on general knowledge, as specific supporting data was not available.

4. Critical Evaluation

4.1 Pedagogical benefits and motivational effects

Generative AI can serve as a form of just-in-time tutoring, offering immediate feedback and suggestions that reinforce academic writing conventions. For L2 students, this tool can foster greater motivation by reducing frustration associated with language proficiency constraints, thereby encouraging iterative revision and sustained engagement with research topics. Additionally, AI-assisted ideation can help students overcome writer’s block and refine conceptual frameworks more efficiently.

4.2 Ethical concerns and academic integrity

Concerns about ethical use center on the potential for AI-generated text to obscure the line between original work and algorithmic output, challenging conventional definitions of plagiarism. Supervisors worry that undetected AI dependencies may lead to misrepresentation of student competence and dilute the pedagogical value of the research process. There is also the broader ethical question of equity, as unequal access to advanced AI tools could exacerbate existing disparities among students in resource-constrained environments.

4.3 Identified gaps and methodological limitations

Existing studies often rely on self-reported data and small-scale case studies, limiting the generalizability of findings. There is a dearth of longitudinal research examining the long-term impact of generative AI use on L2 writing proficiency, as well as insufficient exploration of disciplinary differences in AI integration. Moreover, few investigations address the combined influences of linguistic, cultural, and technological factors specific to the Algerian context, highlighting a need for more comprehensive and contextually grounded methodologies.

Note: This section includes information based on general knowledge, as specific supporting data was not available.

5. Conclusion and Implications

5.1 Persuasive summary of findings

This review illuminates a complex interplay between enthusiasm for generative AI’s capacity to mitigate linguistic challenges and caution regarding its ethical and developmental implications in master’s thesis writing. Both students and teachers recognize substantial advantages in efficiency and language support, yet divergent perceptions of authenticity underscore the necessity for clear institutional policies and pedagogical guidelines.

5.2 Pedagogical and policy recommendations

Academic stakeholders should develop coherent frameworks that define acceptable generative AI use, integrate AI literacy into writing curricula, and promote transparent dialogue between students and supervisors. Training workshops can equip educators with strategies to monitor AI-assisted writing processes, while policy revisions should ensure equitable access and address integrity concerns through assessment designs that emphasize critical reflection and process-oriented evaluation.

5.3 Directions for future research

Future studies should adopt mixed-methods approaches to capture both quantitative metrics of writing improvement and qualitative insights into student and teacher experiences. Longitudinal designs are needed to assess the sustained effects of AI integration on L2 writing proficiency. Furthermore, research should investigate discipline-specific applications and institutional readiness, as well as the socio-cultural dimensions influencing AI adoption within Algerian higher education.

Note: This section includes information based on general knowledge, as specific supporting data was not available.

References

No external sources were cited in this paper.