Get started
Work

Linguistic Evolution of Education Research Writing: A Comparative Analysis Across Technological Eras and AI-Generated Abstracts

Year: 2026
Type: dissertation
Abstract: Academic writing is a dynamic process involving the creation, dissemination, and verification of knowledge, undergoing constant changes based on academic requirements, institutional expectations, and technologies.This thesis investigates how academic writing in the field of education has evolved over more than two centuries (1812-2025), focusing on how linguistic practices have changed across major technological periods.The study uses a corpus-based quantitative approach analyzes 27,048 educational research abstracts to identify diachronic trends in lexical, syntactic, stylistic, and semantic features across three technological eras: the Pre-Word Processor Era (1812-1979), the Word Processor Era , and the AI-Assisted Writing Era (2018)(2019)(2020)(2021)(2022)(2023)(2024)(2025).A labeled dataset of 4,000 abstracts (2,000 human-written and 2,000 AI-generated) was used for comparative analysis.Linguistic feature extraction and keyword-based lexical analysis were applied to examine changes in academic vocabulary and writing style over time, while supervised machine learning models were used as analytical tools to assess whether observed patterns are systematic.The results identified a clear shift toward more lexically dense, semantically coherent, and stylistically uniform academic writing, alongside reduced lexical and syntactic diversity, particularly in the period after 2018.These trends closely align with linguistic characteristics commonly observed in AI-generated academic text.The findings indicate a growing convergence between contemporary human-written abstracts and AI-generated writing styles, suggesting that AI-assisted tools are increasingly shaping academic writing norms in educational research.
Language: English
Open Access status: green
OpenAlex ID: W7215477603
Cites: 47
Cited by:
Related to: 10