Research Presentations

This site contains presentations on research examining how Large Language Models (LLMs) generate text, with a particular focus on the implications for writing instruction and academic writing.

2026 Presentations

Writing Analytics Conference

Beyond the “Feeling” of AI: A Multidimensional Sentiment Analysis Approach to Understanding and Teaching About LLM-Generated Academic Writing
February 19, 2026

Addresses the pedagogical challenge of helping instructors articulate what feels “wrong” about LLM-generated academic writing. Using multivariate sentiment analysis applied to the HAP-E-2 corpus, we demonstrate that GPT-4o uses 4x more approval language than human academic writers and show how these findings can inform classroom practice, particularly for L2 and multilingual students.


2024 Presentations

SHEL Conference

Studies in the History of the English Language
October 19, 2024

Explores the historical context of language modeling technologies and demonstrates how current LLMs produce text that differs systematically from human writing in linguistic features, register, and rhetorical effectiveness.

RSA Conference

Rhetoric Society of America (Three Rivers)
October 31, 2024

Examines the relationship between writing and technology through computational analysis of LLM outputs, showing distinct patterns in how different models generate academic prose compared to expert and novice human writers.

Writing Program Reorientation

Writing & Communication Program Reorientation
August 21, 2024

A practical introduction to LLM technologies for writing instructors, covering the history of NLP, how these models work, and the implications for student writing and pedagogy.