AI evidence extraction

Inquiry-based learning patterns in large language model-driven learning environments: An exploratory study from Bloom’s perspective

Yiming Luo, Ting Liu, Patrick Cheong‐Iao Pang, Dana McKay, Shanton Chang, George Buchanan · Australasian Journal of Educational Technology · 2026

AI-generated evidence extraction, verified across multiple analytical personas. Not a substitute for the peer-reviewed original.

7/10

Relevance

0/4

Quality (LMQS)

Evidence

Citations

36.38

FWCI

Top 10%

Impact

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.14742/ajet.10773

Methodology & findings

Study design

Exploratory mixed-methods study combining experiment, thematic analysis, and observational data.

Sample

N = 117

Main result

The study identified that "14 interaction patterns among students at different levels of prior knowledge" emerged when examining students' behaviors with LLMs across different cognitive stages. The research found that "self-efficacy and metacognitive monitoring" significantly influence students' learning behavior in LLM-driven environments, and that students demonstrate varying reliance on AI depending on their cognitive processing stage.

Reports effect sizes.

Research paradigm

Mixed methods (qualitative-dominant with quantitative descriptive elements)

Author conclusions

The authors conclude that "Educators can improve student inquiry-based learning outcomes by designing cognitive scaffolding that targets specific higher-order thinking stages" and that "Instructional designers should develop planning frameworks that mitigate over-reliance on artificial intelligence while fostering student metacognitive monitoring." They emphasize that "Our study provides new insights for the development of IBL in the era of emerging artificial intelligence technologies."

Risk of bias

Not explicitly stated in abstract. Potential risks include: selection bias (students may self-select into the study), lack of control group comparison, and possible observer effects during screen recording observations.; Selection bias: Sample limited to students in a specific data science academic writing task; Observer bias: Thematic analysis of qualitative data may be subject to researcher interpretation; Attrition: Unknown dropout rates from initial enrollment to completion; Confounding: Prior knowledge differences among participants acknowledged but impact not fully controlled; Selection bias: Sample composition and recruitment method not specified in abstract; Observer bias: Thematic analysis conducted by researchers with potential interpretive bias; Confounding variables: Prior knowledge levels not fully controlled; Self-selection: Unclear if participation was voluntary

Open questions raised

The study identifies the need for: (1) design of guiding planning frameworks and scaffolding to address over-reliance on AI; (2) enhanced metacognitive monitoring in LLM-driven environments; (3) training programs to enhance students' critical evaluation skills within LLM-driven contexts; (4) further research on how students interact with LLMs at different cognitive levels during inquiry-based learning.
Insufficient research on how students interact with LLMs for inquiry-based learning
Need for design of guiding planning frameworks and scaffolding to address AI reliance challenges
Gap in understanding metacognitive monitoring in LLM-driven environments
Limited guidance on cognitive scaffolding design for different cognitive stages
Insufficient research on how students interact with LLMs during inquiry-based learning

Data: not_statedCode: not_statedExtracted from: pdfAgreement 55%

Explore related topics

AI Literacy Training Doctoral Training Mixed Methods