Michael Laudenbach
My research focuses on teaching writing in the disciplines, particularly in statistics & data science, as well as engineering fields. I use computational linguistics to analyze large collections of both student and professional writing, transforming the results into lesson plans and learning interventions. My scholarly interests and teaching inform one another, and my theoretical grounding in rhetorical genre studies helps me to adapt and operationalize these concepts in different disciplinary contexts and at various student levels. For example, I’ve provided communication support to undergraduate- and graduate-level mechanical engineering courses in the form of in-class lessons and workshops as well as my assessment of major assignments.
Collaborating with faculty in statistics & data science, I contributed the first descriptive study of its kind to study the discipline with corpus linguistics methods. Part of the novel contribution is realized in my use of DocuScope, a dictionary-based text-tagger developed at Carnegie Mellon by David Kaufer and Suguru Ishizaki. DocuScope’s ability to explain rhetorical variation in style and register has been established in previous studies as particularly distinct from traditional linguistic perspectives. This corpus analysis led to instructors revising and editing their assignment prompts and grading rubrics to better match the type of writing that they ask students to produce. I’ve gained substantial experience adapting my research to the teaching of professional writing, technical research documentation, presentation design, and collaborative writing.
While completing my PhD, I led the pilot study of an in-house digital writing tool that provides automated writing feedback to users. This study was tailored to students in an introductory statistics course, and I led the collection and curation of a large-scale corpus of student and published writing in statistics and data science—the first of its kind. Recently, I’ve been working with the same research team to study variation in our statistics corpus and other corpora alongside text generated by Large Language Models (LLMs) like ChatGPT, GPT-4o, and LLaMa.
Selected publications:
Reinhart, A., Brown, D. W., Markey, B., Laudenbach, M., Pantusen, K., Yurko, R., & Weinberg, G. (2024). Do LLMs write like humans? Variation in grammatical and rhetorical styles. arXiv preprint arXiv:2410.16107.
Markey, B., Brown, D. W., Laudenbach, M., & Kohler, A. (2024). Dense and disconnected: Analyzing the sedimented style of ChatGPT-generated text at scale. Written Communication, 41(4), 571-600.
Laudenbach, M., Brown, D. W., Guo, Z., Ishizaki, S., Reinhart, A., & Weinberg, G. (2024). Visualizing formative feedback in statistics writing: An exploratory study of student motivation using DocuScope Write & Audit. Assessing Writing, 60, 100830.
Tuesdays & Thursdays, 12:30pm - 2:00pm