Document length alone carries some signal (word-count classifier ROC-AUC 0.68), but the fixed-word sliding-window design largely prevents stylometric features from capturing length differences¶
Claim · All claims
Evidence · 1 study · designr2·q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person · recheck by 2029: tool and model accuracy changes quickly
Subclaims¶
Reading the evidence codes
q is the design tier, i the impact, n the sample, and the kind with its r
the rigour judged within that kind: what each code means.
This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.
q2 i? A classifier using only total document word count achieved a cross-validation ROC-AUC of 0.68, showing length contains some information, while the 250-word windowing design largely prevents the stylometric features from directly capturing document-length differences. → Rajesh Kumar 2026
Evidence¶
Rajesh Kumar 2026¶
Rajesh Kumar, Nabeel Siddiqui, Alexander Fuchsberger. (2026). Detecting GPT-Assisted Writing Using Interpretable Stylometric Features. https://arxiv.org/abs/2609.26687 doi:10.48550/arxiv.2609.26687
q2 · i? · design · r2
Post-hoc control analysis comparing a word-count-only classifier with the main windowed model. Independently authored documents were longer (paired t(89) = 6.46, p < .001), and replacing TTR and Hapax Ratio with moving-average TTR produced an ROC-AUC close to the full model.
"a classifier using only the total document word count achieved a cross-validation ROC-AUC of0.68, showing that document length itself contains some information about the writing condition."
Discussion¶
Related Claims¶
- SHAP analysis identifies Hapax Ratio as the most influential stylometric feature in Random Forest predictions, followed by NSR, Noun Ratio, Adverb Ratio, and TTR — related
- Random Forest using nine interpretable stylometric features detects GPT-assisted student writing with held-out ROC-AUC of 0.870 and F1 of 0.842 — related
- In the random forest, word count was the most important readability feature while traditional readability formulas ranked near the bottom — related
- The stylometric detector misclassified 4 of 18 independently authored documents as GPT-assisted (FPR 22.2%) and 2 of 18 GPT-assisted documents (FNR 11.1%), with wide confidence intervals — related
- Academic word list count, word count, and Flesch-Kincaid grade level are the most important features for predicting cognitive engagement in discussion posts — related
Page metadata
| Field | Value |
|---|---|
type |
claim |
id |
sliding-window-controls-document-length-effects |
title |
Document length alone carries some signal (word-count classifier ROC-AUC 0.68), but the fixed-word sliding-window design largely prevents stylometric features from capturing length differences |
description |
Document length alone carries some signal (word-count classifier ROC-AUC 0.68), but the fixed-word sliding-window design largely prevents stylometric features from capturing length differences |
status |
review |
evidence_strength |
weak |
newest_evidence |
2026 |
recheck_by |
2029 |
generated |
by process:wiki-ingest on 2026-10-10 |
trust |
unverified — no one has confirmed this page's content |
sources |
1 — rajesh-kumar-2026 |
Frontmatter is the page's machine-readable record — see the Schema & Guide.