AI evidence extraction

AI Affordances, Teacher Support, and Research-Writing Competence: The Mediating Roles of Self-Efficacy and Disposition

John Manuel C. Buniel, Myriflor A. Miranda, Elvie Lyka L. Duran, Francis Isidore B. Ambray · International Journal of Learning Teaching and Educational Research · 2026

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

9/10

Relevance

1/4

Quality (LMQS)

Evidence

Citations

0.00

FWCI

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.26803/ijlter.25.4.44

Methodology & findings

Study design

Cross-sectional survey study using Structural Equation Modeling (SEM).

Sample

N = 526, 2 groups

Primary method

SPSS version 26 for descriptive statistics; AMOS 26.0 for Confirmatory Factor Analysis (CFA) using maximum likelihood estimation and Structural Equation Modeling (SEM). Goodness-of-Fit Indices evaluated: CFI, TLI, RMSEA, SRMR, Chi-square/df ratio. Internal consistency assessed via composite reliability (CR) and average variance extracted (AVE). Discriminant validity tested using Fornell-Larcker criterion. Hypothesis testing at p < 0.01 and p < 0.001 significance levels.

Main result

The study found that "AI affordances positively and significantly influenced research self-efficacy (βH1 = 0.436, t = 9.968, p = 0.000 < 0.01) and research disposition (βH2 = 0.434, t = 7.433, p = 0.000 < 0.01)" while teacher support had a positive and significant influence on both self-efficacy and disposition. Importantly, "AI affordances (βH7 = 0.069, t = 1.536, p = 0.125 > 0.01) failed to exhibit a significant impact on research competence," suggesting that "the mere availability or perceived usefulness of AI tools does not automatically translate into improved research-writing performance."

Reports effect sizes and confidence intervals.

Research paradigm

Positivist/quantitative

Author conclusions

"Overall, AI serves as a supportive tool that boosts engagement and confidence, while teacher support remains the foundation for developing deeper, more transferable research-writing skills." The authors further conclude: "The current study clarified that research-writing competence is best cultivated when AI assistance operates alongside strong pedagogical guidance. The findings also suggested that combining AI integration with robust teacher support enhances learning experiences."

Risk of bias

Cross-sectional design limits causal inference; Self-report measures introduce response bias and social desirability bias; Single institutional context limits generalizability; No qualitative triangulation for validation; Absence of longitudinal follow-up; Self-report measures introduce response bias; Lack of qualitative triangulation; No control group or comparison condition; Potential common method variance from same-source questionnaire; Selection bias: Single institution sample may not represent broader student population; Self-report bias: All measures based on participant perceptions; Cross-sectional design: Cannot establish causality; Common method bias: Single survey method for all variables; Temporal confounding: No baseline or temporal sequence established

Limitations

This study used "a cross-sectional design and self-report measures, which may have limited causal inference and introduced response bias
The sample was drawn from one higher-education context
thus, generalizability should be interpreted cautiously
In addition, qualitative triangulation was not feasible because data collection had already been completed
future mixed-method research is recommended to corroborate and deepen the interpretation of the SEM pathways."

Open questions raised

Limited empirical evidence directly focusing on research-writing competence among undergraduate students with AI affordances
Lack of mechanistic explanations for how AI-related support and instructional support translate into competence
Few studies using structural equation modeling to test direct and indirect pathways simultaneously
Limited evidence in Philippine higher education contexts
Need for qualitative exploration of students' lived experiences with AI tools, including perceptions of overreliance and ethical concerns
Opportunity to investigate emerging AI features such as multimodal feedback and AI-supported peer review

Extracted from: pdfAgreement 57%

Explore related topics

AI Manuscript Writing Academic Writing Survey