Writing Style Fingerprint

Activity Overview

At the start of the semester, each student writes an in-class assignment that's designed to provide a highlight existing writing and research skills.  The goal is to get a baseline for each student in a controlled environment free of technological assistance.

This baseline writing sample serves two purposes later in the semester:

• Evaluating Progress: As the semester progresses, each student should show greater clarity in their writing style and improved citation formatting.

• Checking for AI and Plagiarism: If a future assignment doesn't match the writing style seen in the baseline, then this could be a possible flag for AI usage.  (More on the uses and limitations below.)

Writing Styles Change with Consistency

Limitations for Plagiarism Detection

We must recognise that the fingerprint comparison cannot be used to "prove" a student has used AI — it's only one possible indicator.  Depending on how your assign your baseline versus the later assignments, the same student may exhibit very different writing styles and "quality" in their submissions:

• Different instructors have different expectations.  A student's natural writing style might easily match one instructor's expectations of adding opinion and using "I," whereas the same student might struggle to put ideas on the page if another instructor says "Your paper should avoid bias, so don't include your opinion or use 'I.'"  When your student writes an early-semester baseline, that student loosely follows the expectations they've adopted from prior instructors — later in the semester, that student attempts to meet the expectations of the current class with the added pressure of earning a good grade.

• Every writer has "good" and "bad" days.  Your student might have just pulled an all-nighter before writing their baseline, but then they were well rested before writing their research paper.

• Students may have multiple baseline styles depending on the situation.  In the past, I've submitted assignments that had different styles in the same assignment.  When I'm fully awake, I often take careful notes and write with controlled sentences that are very boring.  (Much like this textbook you're currently reading!)  When I'm sleep deprived or even half asleep, I write in a style that's creative and exciting and not always fully coherent.  When I write assignments under a deadline, the sections I write early with have that controlled style, whereas the pages I write in the final hours might sound entirely different.

When I was working on my Ph.D., I wanted to run a IRB study that would compare student progress in creative writing across the span of a semester.  Unfortunately, this ran into a couple major problems, and my advisor ruled out this idea for two key reasons:

• What's the standard of "good"?  When we evaluate writing, determining quality is partly subjective, especially in creative writing.  That said, the research writing of a typical English 101 composition course should have certain expections for structure, clarity, and citation.  Whenever we consider a student's progress, we must focus on those key elements that can be measured through grades and assessment.

• Different instructors have their own pedagogical models.  Even if I did come up with some rubric for that standard of "good" writing, other writers and researchers would disagree on whether it's valid, especially in creative writing.  So please remember that everything on this website represents my personal judgment on how to write well — and my particular pedagogy will not work for every writing situation.

•