12,637 papers · updated 18 Sept 2026livingmeta.ai
← Browse all papers
AI evidence extraction

Augmenting dissertation mentorship through multi-modal generative AI: adapting language and visuals to diverse learning styles

Daren Scerri, Alan Gatt, Stephane Role, Gerard Said Pullicino · Artificial Intelligence in Education · 2025

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

7/10
Relevance
E
Evidence
1
Citations
1.76
FWCI
Top 10%
Impact

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.1108/aiie-05-2025-0085

Methodology & findings

Study design

Classroom experiment with a multi-modal GAI prototype evaluated at a vocational college where learners queried the AI mentor about research proposal forms, methodology assistance, and theory-context connections..

Main result

The study found that "51% of learners sitting for the pilot session preferred the kinesthetic mode, a finding in line with the college's vocational structure." Additionally, "Results for the AI-driven experiment confirm that increased accessibility, a reliable and/or verified data corpus and the AI mentor ability to explain with language adapted to each learning style are the three main features that distinguish the specialized prototype from generic chatbots like ChatGPT or Gemini."

Reports effect sizes.

Research paradigm

pragmatism

Author conclusions

The authors conclude that "while human interaction with the lecturer and/or mentor remains critical, the proposed AI solution augments mentorship provision beyond classroom hours and to a level of detail, which is difficult to achieve in traditional classroom or mentor meeting settings." They also note that "our work is distinct in applying a unified text-table-image pipeline to an educational tutor and combining this with learning style-based conditioning."

Risk of bias

Selection bias - sample from single vocational college may not generalize; Lack of control group comparison; Small pilot sample size not specified; No baseline measurements reported; No information on blinding or randomization; Single institution evaluation may limit generalizability; Potential self-selection bias in learner participation in pilot session

Open questions raised

  • The authors note that while prior multi-modal RAG systems exist, their work is distinct in applying a unified text-table-image pipeline to an educational tutor and combining this with learning style-based conditioning.
Data: not_statedCode: not_statedExtracted from: pdf

Explore related topics

AI Literacy TrainingDoctoral TrainingDesign Science

Related papers

← Browse all papers