University students with high versus low GenAI literacy scores show significantly different ENA interaction structures during an AI-assisted abstract writing task¶

Claim · All claims
Evidence · 1 study · associational r1 · q2 · i2 medium
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i2 High- and low-literacy groups differed significantly on the primary dimension of the ENA space (U = 683.50, p < .001, r = .35). → Angxuan Chen 2026

Evidence¶

Angxuan Chen 2026¶

Angxuan Chen, Jiyou Jia. (2026). Tracing GenAI Literacy: Uncovering Student-AI Interaction Patterns in Academic Writing through Epistemic Network Analysis. First International Workshop on Advancing AI Literacy with Learning Analytics (AI-LIT) @ LAK26, April 2026. https://arxiv.org/abs/2606.00040 doi:10.48550/arxiv.2606.00040

q2 · i2 · associational · r1

Statistical comparison of projected ENA scores for 162 students split by median GenAI Literacy Test score into HG (n = 89) and LG (n = 73). The test "confirmed significant differences between the two groups" with r = .35, a medium effect. The design is associational, not causal.

"A Mann-Whitney U test confirmed significant differences between the two groups on the primary dimension (X-axis) of the ENA space (𝑈= 683.50 ,𝑝<.001 ,𝑟=.35 )."

Discussion¶

Related Claims¶

• ENA revealed high CT students had significantly more coherent SRL cognitive networks spanning planning, execution, and self-reflection than low CT students — related
• Intrinsically motivated students form reflective-iterative metacognitive co-occurrence networks, but overall ENA centroid positions do not differ significantly between motivational groups — related
• High- and low-performing students did not differ significantly on the midterm-exam ENA network dimensions despite detected network differences — related
• High-literacy students' ENA networks center on links among AI Improvement Commands, Clarification Questions, and Meta-Commands, described as a collaborative, epistemic approach — a narrower finding that bears on this claim
• Low-literacy students' ENA networks are dominated by links between Abstract Generation Commands and Factual or Challenging Questions, described as a transactional approach — a narrower finding that bears on this claim
• Learning-oriented video groups differ significantly from the productivity-oriented group on the primary ENA axis, with large effect sizes — related
• General AI literacy is not significantly related to in-service science educators' AI use or to their age, gender, or years of service — related
• The Most Improved profile showed an integrated AI-support pattern with strong ENA connections among coaching, modeling, and reflection and frequent model uptake in group discussion — related
• Epistemic network analysis shows high-learning groups integrate re-annotations with knowledge-confirming discourse more densely than low-learning groups — related
• ENA contrasting-cases analysis shows high- and low-effectiveness studies differ significantly in their design-element and intention networks — related