Teachers designing multi-agent workflows fall into three behavioral archetypes: Systematic Optimizers, Prolific Creators, and Passive Observers¶

Claim · All claims
Evidence · 1 study (3 entries) · associational r1–r2 · q2
Currency · newest study 2025 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? K-means clustering of 61 teachers' platform actions identified three distinct behavioral profiles with different compositions of configuration, creation, browsing, and cloning activity. → Sun 2025

Evidence¶

Sun 2025¶

Sun, Y., Xin, H., Li, S., Niu, Q., Chai, C. S., Huang, L., & Chen, G. (2025). Modeling AI-TPACK in Practice: Insights from Teachers' Multi-Agent Workflow Design. https://arxiv.org/abs/2605.13906 doi:10.48550/arxiv.2605.13906

q2 · i? · associational · r2

K-means cluster analysis (k=3, average silhouette score 0.389) of clickstream logs from 61 teachers (8,718 actions) in a two-day AI professional development workshop. "Cluster 1 (n=18) was configuration-intensive" with Content Configuration at 32.50% of activity.

"Cluster 1 (n=18) was configuration-intensive: Content Configuration dominated their activity (32.50%), while Content Editing and Testing/Debugging rates were 3-4 times higher than other clusters."

Sun 2025 (2)¶

Sun, Y., Xin, H., Li, S., Niu, Q., Chai, C. S., Huang, L., & Chen, G. (2025). Modeling AI-TPACK in Practice: Insights from Teachers' Multi-Agent Workflow Design. https://arxiv.org/abs/2605.13906. doi:10.48550/arxiv.2605.13906

q2 · i? · associational · r1

Same cluster analysis of 61 teachers' logs: "Cluster 2 (n=35) balanced creation and browsing", with Object Creation at 25.20% and Browsing Own Content at 27.10%, while editing and testing remained minimal.

"Cluster 2 (n=35) balanced creation and browsing: Object Creation (25.20%) and Browsing Own Content (27.10%) comprised over half their activity, but editing and testing remained minimal."

Sun 2025 (3)¶

Sun, Y., Xin, H., Li, S., Niu, Q., Chai, C. S., Huang, L., & Chen, G. (2025). Modeling AI-TPACK in Practice: Insights from Teachers' Multi-Agent Workflow Design. https://arxiv.org/abs/2605.13906. doi:10.48550/arxiv.2605.13906

q2 · i? · associational · r1

Same cluster analysis: "Cluster 3 (n=8) was browsing-dominant", with Browsing Own Content at 29.00%, the lowest editing rate (1.70%), and the highest Template Cloning (2.45%) across clusters.

"Cluster 3 (n=8) was browsing-dominant: Browsing Own Content (29.00%) was their primary activity, with the lowest editing rate (1.70%) and highest Template Cloning (2.45%) across clusters."

Discussion¶

Related Claims¶

• AI-TPACK integration quality differs across behavioral clusters: Optimizers achieve cohesive systemic designs, Creators show pragmatic enhancement, Observers show unstable integration — related
• Behavioral patterns alone insufficiently predict AI-TPACK capacity because Cluster 3 is polarized between expert-level and surface-level integration — related
• Markov transition analysis shows the three clusters differ in workflow dynamics: iterative optimization loops, rapid prototyping flows, and browsing-anchored exploration — related
• Interview data link cluster differences in AI-TPACK integration to distinct profiles of AI-TK, technical self-efficacy, pedagogical beliefs, and support-dependence — related
• Teachers' post-workshop platform engagement fell into three behavioral profiles, with most teachers showing minimal creation and browsing — related
• Clustering yields a small number of behaviorally distinct team clusters that instructors can use for cluster-differentiated feedback — related
• Platform logs show behavioral shifts toward goal-oriented optimization between workshop Days 4 and 5 — related
• Policy packages fall into four archetypes defined by which design element is missing — related
• 87% of K-12 teachers ceased meaningful AI agent creation within three weeks of completing a two-day professional development workshop — related