Skip to content

Systematic Qualitative Information Diagrammer (SQuID): Automating Inductive Coding with LLMs to Assist Affinity Diagramming​

Affinity diagramming is slow and labor-intensive. SQuID demonstrates how interactive AI support can accelerate early analysis while preserving researcher control and interpretive rigor.
SQuID is an AI-assisted system that supports researchers in organizing, clustering, and interpreting qualitative data through mixed-initiative affinity diagramming.

Qualitative researchers often rely on affinity diagramming to synthesize large volumes of interview, observation, and open-ended survey data. However, this process can be time-intensive, cognitively demanding, and difficult to scale—especially in collaborative or interdisciplinary settings.

Built on our prior work (QuAD) SQuID (Systematic Qualitative Information Diagrammer) addresses these challenges by combining the interpretive power of human reasoning with the pattern-recognition capabilities of artificial intelligence. The system assists users in clustering related excerpts, generating potential thematic labels, and supporting justification and reflection through features like “Generate a Reason” and “Regroup Controls.” Rather than replacing qualitative interpretation, SQuID embodies a mixed-initiative approach that keeps users “in the loop.” Researchers can accept, modify, or reject AI-generated suggestions, creating a transparent record of how codes and themes evolve over time.

Through iterative user studies with experienced qualitative researchers and students, SQuID has been shown to accelerate early-stage analysis, improve traceability of analytic decisions, and prompt deeper reflection on emergent themes. This project advances the understanding of how interactive AI systems can augment reflexive thematic analysis while maintaining rigor, transparency, and researcher agency. SQuID aims to make qualitative synthesis not only faster—but more systematic, reflective, and collaborative.

Demonstration Video

Coming soon

Publications

Ariel Goldman, Cindy Espinosa, Shivani Patel, Francesca Cavuoti, Jade Chen, Alexandra Cheng, Sabrina Meng, Aditi Patil, Lydia B Chilton, and Sarah Morrison-Smith. 2022. QuAD: Deep-Learning Assisted Qualitative Data Analysis with Affinity Diagrams. In Extended Abstracts of the 2022 CHI Conference on Human Factors in Computing Systems (CHI EA ’22). Association for Computing Machinery, New York, NY, USA, Article 419, 1–7. https://doi.org/10.1145/3491101.3519863

Presentations

S. Favela, M. Maillet, Y. Wu, S. Morrison-Smith. 2023. SQuID – Systematic Qualitative Information
Diagrammer. Hamilton College Summer Research Poster

Team

Faculty

Previous Undergraduate Researchers

  • Francesca Cavuoti
  • Jade Chen
  • Alexandra Cheng
  • Sebastian Favela
  • Cindy Espinosa
  • Ariel Goldman
  • Luiza Leschziner
  • Matthew Maillete
  • Sabrina Meng
  • Shivani Patel
  • Aditi Patil
  • Yifan (Ivan) Wu