Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
120 tokens/sec
GPT-4o
10 tokens/sec
Gemini 2.5 Pro Pro
44 tokens/sec
o3 Pro
5 tokens/sec
GPT-4.1 Pro
3 tokens/sec
DeepSeek R1 via Azure Pro
55 tokens/sec
2000 character limit reached

Enhancing Critical Thinking in Education by means of a Socratic Chatbot (2409.05511v1)

Published 9 Sep 2024 in cs.HC

Abstract: While LLMs are increasingly playing a pivotal role in education by providing instantaneous, adaptive responses, their potential to promote critical thinking remains understudied. In this paper, we fill such a gap and present an innovative educational chatbot designed to foster critical thinking through Socratic questioning. Unlike traditional intelligent tutoring systems, including educational chatbots, that tend to offer direct answers, the proposed Socratic tutor encourages students to explore various perspectives and engage in self-reflection by posing structured, thought-provoking questions. Our Socratic questioning is implemented by fine and prompt-tuning the open-source pretrained LLM with a specialized dataset that stimulates critical thinking and offers multiple viewpoints. In an effort to democratize access and to protect the students' privacy, the proposed tutor is based on small LLMs (Llama2 7B and 13B-parameter models) that are able to run locally on off-the-shelf hardware. We validate our approach in a battery of experiments consisting of interactions between a simulated student and the chatbot to evaluate its effectiveness in enhancing critical thinking skills. Results indicate that the Socratic tutor supports the development of reflection and critical thinking significantly better than standard chatbots. Our approach opens the door for improving educational outcomes by cultivating active learning and encouraging intellectual autonomy.

Summary

  • The paper presents a novel Socratic chatbot that fosters critical thinking through targeted, reflective questioning techniques.
  • The paper utilizes parameter-efficient tuning methods like LoRA and QLoRA on Llama2 models, achieving superior performance on metrics such as BLEU and ROUGE-L.
  • The paper demonstrates that AI-powered tools can democratize education by promoting deeper cognitive engagement and self-regulated learning.

Enhancing Critical Thinking in Education through a Socratic Chatbot

The paper "Enhancing Critical Thinking in Education by means of a Socratic Chatbot" addresses a critical gap in the intersection of artificial intelligence and education. It explores the development of a chatbot based on LLMs to foster critical thinking via Socratic questioning. This approach contrasts traditional educational chatbots, which typically provide straightforward answers, by encouraging intellectual exploration and reflection through structured, provocative questions.

Problem Exploration and Methodology

Educational technology, while proliferating with numerous applications, often falls short in promoting deeper cognitive skills such as critical thinking and self-regulated learning. To address this deficiency, the authors propose a Socratic tutor built upon the foundation of LLMs, specifically fine- and prompt-tuned models like Llama2, with parameters ranging from 7B to 13B.

The methodology involves using parameter-efficient techniques such as Low-Rank Adaptation (LoRA) and Quantized LoRA (QLoRA), which enhance the pretrained models specifically for the role of Socratic questioning. The fine-tuning leverages the SocratiQ dataset, a compendium of question pairs tailored to stimulate deep reflection and analysis. These models are then prompt-tuned to elicit targeted Socratic questions which guide learners through critical examination without imposing new information.

Evaluation and Results

The chatbot's efficacy was quantitatively assessed using LLM-based simulated interactions with a focus on Theory of Knowledge (ToK) questions. Key performance metrics included BLEU, ROUGE-L, METEOR, and BERTScore, supplemented by an innovative LLM-based scoring system directly evaluating critical thinking. The Socratic tutor outperformed baseline methods significantly, achieving superior engagement and depth in generating student responses consistent with critical thinking principles. Notably, performance gains were observed across both small (7B) and larger (13B) parameter models, validating the approach regardless of scale.

Implications and Future Directions

From a practical perspective, developing a Socratic chatbot signifies a progressive step towards democratizing education, offering high-quality critical thinking exercises across varied learning environments with minimal hardware requirements. Running these LLMs locally ensures student privacy while making the technology accessibly inclusive.

Theoretically, this work underscores the potential of AI, particularly LLMs, beyond rote learning enhancements, by facilitating educational environments conducive to exploratory learning and intellectual autonomy. It provides a framework for educators seeking to embed AI-driven Socratic methods into curriculum design, promoting an educational culture of questioning and reflection.

Future research could expand this model's deployment in real-world educational settings, evaluating long-term impacts on student cognitive development. Furthermore, extending this architecture with adaptive learning algorithms could refine its capacity to cater to diverse cognitive profiles and learning speeds, enriching the personalized educational experience.

In conclusion, this paper effectively bridges AI technology with critical educational outcomes, proving that when designed thoughtfully, chatbots can transcend basic informational roles and contribute meaningfully to the development of essential cognitive skills in learners.

Dice Question Streamline Icon: https://streamlinehq.com

Follow-up Questions

We haven't generated follow-up questions for this paper yet.

Youtube Logo Streamline Icon: https://streamlinehq.com