Guest Editors: 

• Amanda Barany, University of Pennsylvania
• Ha Nguyen,  University of North Carolina at Chapel Hill
• Ryan S. Baker, University of Pennsylvania

Scope and Rationale:

The use of artificial intelligence (AI) for qualitative research has transformed traditional research methods by enhancing the quality, scalability, and accessibility of qualitative data analysis, particularly when used by researchers as part of a hybrid approach which is not fully automated but instead leverages human strengths where relevant. Qualitative data analysis can provide rich insights into students’ and educators’ perspectives and experiences, and can create data annotations which can be used in several forms of educational data mining such as correlation mining, causal mining, and prediction modeling. Historically, data scientists have leveraged classical Natural Language Processing (NLP) tools such as text classification, keyword extraction, sentiment analysis, and topic modeling to understand phenomena in qualitative data, but the limits to the ways these approaches handle context, nuance, cross-language adaptability, complex task automation, and scalability with rich data sources have been extensively documented. Contemporary AI tools such as large language models (LLMs) have potential to augment existing approaches by drawing on large datasets and deep neural network architectures to understand complex language patterns in ways that can offer richer and more accurate insights, support advanced automation, and enable more adaptable and contextual understanding. When used in conjunction with human interpretation, such tools can also serve as “co-researchers” to improve the comprehensiveness and rigor of research outcomes, and support the integration of qualitative methods with educational data mining methods.

While promising, the adoption of contemporary AI tools for qualitative research, including LLMs and Generative AI, remains relatively new, and brings with it a new set of challenges, particularly in terms of ethics, data privacy, reproducibility, and methodological transparency, as well as raising foundational questions about what qualitative research is. Best (or at least better) practices for data process and analysis procedures incorporating AI are emerging, but remain uncertain and often limited in scope.

This special section aims to map the current landscape of AI-driven qualitative research in ways  that can help researchers integrate these tools effectively and ethically. The need for a systematic understanding is critical, as researchers across fields such as education, social sciences, and health sciences increasingly rely on tools such as LLMs to streamline the analysis of data in qualitative fashions. Given the rapid evolution and application of AI, this special section will explore key trends, methodologies, and best practices, offering both a snapshot of the field and a roadmap for future work.

Relevant Topics:

• Human-AI partnership for qualitative data analysis
• Automation of codebook development
• Using AI to support insight development and inductive coding
• Automation of data coding procedures
• Ensuring validity and reliability of qualitative themes and constructs
• Improving reliability in automated coding
• Roles and responsibilities in hybrid human-AI research workflows
• Human and AI critique and feedback in qualitative data analysis
• Effects of human-AI collaboration on qualitative researchers' perspectives and practices
• Effects of human-AI collaboration on AI outputs
• Influence of AI-enhanced approaches on theory-building and research frameworks in qualitative research
• Foundational consideration of the meaning of qualitative, ethnographic, and phenomenological analysis when involving AI as a partner 

Submission Instructions:

We invite researchers, practitioners, and scholars to submit a 500-800 word abstract (including references) detailing their proposed contribution. Submissions should include a proposed title, a list of authors, and the names and contact information for two or more possible reviewers for the work. Please include a brief overview of the research context, methodology, tools used, and any key findings or challenges. Submissions should be emailed to ryanbaker.handin2@gmail.com no later than February 1, 2025 AOE (anywhere on Earth). All abstracts will undergo a blind peer-review process by the special section review committee and receive feedback by February 15, 2024. 

Full paper manuscripts should be submitted via the JEDM submission portal by July 15, 2025 AOE. Formatting, style, and manuscript length should adhere to the submission requirements for the Journal of Education Data Mining.

Proposed Timeline:

• February 1, 2025: Abstract deadline
• February 15, 2025:  Abstract feedback sent to authors
• July 15, 2025:         Full paper submission deadline
• October 15, 2025:   Review round 1 return
• January 15, 2026:   Review round 2 return
• TBD                 Special Section Publication