Automating pedagogical evaluation of LLM-based conversational agents

Manolis Mavrikis

Keywords: automated evaluation, Socratic dialogue, AI tutor

TL;DR: Automating AIED dialogue evaluation using LLM-as-a-judge approach finding they match human judgment on surface features but struggle with subtle teaching behaviours.

Abstract:

With the growing adoption of large language models (LLMs) in educational settings, there is an urgent need
for systematic and scalable evaluation methods. Traditional natural language generation metrics such as BLEU,
ROUGE and METEOR excel at measuring surface‐level linguistic quality but fall short in evaluating the interactive,
adaptive nature of dialogue alignment of conversational agents, particularly in relation to their intended design.
To address these gaps, we propose an evaluation strategy that extends beyond technical evaluation (linguistic
coherence and semantic relevance). In this pilot study we compare human and LLM-based evaluation of a
conversational agent, with a focus on Socratic dialogue as a specific instantiation. Early results indicate that
our LLM-as-a-Judge aligns closely to human evaluators for clear, surface‐level qualities like encouragement and
actionable guidance, but less on subtle pedagogical behaviours such as recognising errors and maintaining natural
dialogue flow. These early results underscore the promise of LLM-based evaluators for scalable assessment of
tutoring behaviours while highlighting the need for targeted fine-tuning and hybrid approaches to improve
nuanced error detection and dialogue coherence.

Submission Number: 142