AI-supported argumentative writing yields three distinct uptake clusters (text reproducers, text integrators, text reconstructors) based on recurring n-gram overlap¶

Claim · All claims
Evidence · 1 study · associational r2 · q3
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i? Preregistered study of 131 ChatGPT-supported writers found a three-cluster solution of uptake patterns, with clusters differing in n-gram overlap length and proportion of copied content. → Teresa Luther · Kate Belcher · Joachim Kimmerle · Detmar Meurers · Ulrike Cress 2026

Evidence¶

Teresa Luther · Kate Belcher · Joachim Kimmerle · Detmar Meurers · Ulrike Cress 2026¶

Teresa Luther · Kate Belcher · Joachim Kimmerle · Detmar Meurers · Ulrike Cress (2026). Collaborative knowledge construction with generative AI: Exploring argumentative co-writing processes through n-gram and cluster analysis. Intern. J. Comput.-Support. Collab. Learn. https://doi.org/10.1007/s11412-026-09470-1

q3 · i? · associational · r2

In the preregistered study (n = 131 participants who interacted with ChatGPT), recurring n-grams between AI responses and final texts were extracted, binned, and clustered via PCA and k-means. The authors "empirically identified three distinct clusters of participants on the basis of their textual overlap patterns." A scree plot showed the elbow at k = 3.

"Using an argumentative writing task, in which individuals could use ChatGPT, we empirically identified three distinct clusters of participants on the basis of their textual overlap patterns."

Discussion¶

Related Claims¶

• Affinity for technology interaction differed across clusters, with text reproducers highest and integrators scoring higher than reconstructors (d = 0.59) — related
• Current ChatGPT users were overrepresented among text reconstructors and underrepresented among text integrators (Cramér's V = 0.26) — a narrower finding that bears on this claim
• Text reproducers copied nearly all of their final text from ChatGPT with long verbatim sequences, while text integrators showed fragment-level micro-uptake — a narrower finding that bears on this claim
• Clusters did not differ significantly in the number of prompts sent to ChatGPT — a narrower finding that bears on this claim
• Cluster membership showed a strong multivariate effect on standard text-similarity metrics (Pillai's trace = 0.85, F(12, 248) = 15.31, p < 0.001) — a narrower finding that bears on this claim
• Affinity propagation clustering of gameplay attempts yielded an interpretable six-cluster solution (part.6)
• Clustering n-gram flip sequences per task yields three distinct problem-solving strategy groups across the binary challenges
• Affordance resonance can be documented at scale through AI-generated logs identifying seed contributions and cascading uptake — related
• In reviewed studies, individual data types were analyzed mainly with descriptive analysis, content analysis, network analysis, and clustering — a broader claim this one bears on