This week’s post and screencast was developed by Ph.D. candidate, Matthew Nyaaba.
Matthew is in the Department of Educational Theory and Practice at the University of Georgia, specializing in Elementary and Teacher Education. As an international and first-generation scholar from Ghana, Matthew’s research focuses on the intersections of Artificial Intelligence, culturally responsive pedagogy, and equity in science education. He serves as a Research Assistant at the AI4STEM Education Center, where he co-leads cross-cultural AI research initiatives and has organized global webinars on responsible AI use in education.
Transcript: Exploring CAQDAS
Hello, everyone. I am really excited to share a concept that has intrigued me for some time now—Computer-Assisted Qualitative Data Analysis Software, or what is often abbreviated as CAQDAS. I still sometimes wonder whether to say each letter—C-A-Q-D-A-S—or pronounce it as one word like “CAQDAS.” For simplicity and rhythm, I will go with ‘CAQDAS’ in this presentation.
My journey with CAQDAS began during one of my qualitative research courses. We touched briefly on it, explored one of the tools, and again I had the opportunity to perform qualitative analysis using one of the tools, MAXQDA. Still, I felt I had not fully grasped its potential. That curiosity stayed with me. So, I decided to dig deeper into what these tools really offer and how they’re evolving, especially now that we are seeing a wave of AI integration. And today, I am excited to walk you through what I have learned.
I structured my exploration around several key questions. First, what exactly is CAQDAS? What is its history? What are its capabilities and benefits? And more importantly, how is it being reshaped by artificial intelligence? Along the way, I’ll also touch on best practices and limitations—because every tool, no matter how powerful, has its caveats.
At its core, CAQDAS tools are designed to assist researchers in managing, coding, and analyzing qualitative data—this includes text, audio, video, and even images. What makes these tools powerful is their ability to organize themes, support theory building, and generate visual representations of patterns in the data. However, it is important to note that they only facilitate the process—they do not do the thinking for you. The human analyst is still very much at the center.
Historically, CAQDAS tools began gaining traction around the 1980s. One of the earliest tools was HyperRESEARCH, and it was the University of Surrey’s CAQDAS Networking Project that really helped the term gain prominence in the research world. The term “CAQDAS” itself was coined by Fielding and Lee in 1989. By 1994, through the University of Surrey’s project, it began to gain serious momentum among qualitative researchers.
As I explored the capabilities of these tools, I realized they can do quite a lot. They help with coding—identifying patterns and labeling them within texts. They support data retrieval, memo writing, and even querying your dataset. Tools like NVivo, MAXQDA, and ATLAS.ti also support team collaboration. Imagine doing a team-based thematic analysis where everyone is on the same page and contributing to the coding process in real time—that’s what some of these platforms make possible.
In terms of benefits, these tools shine when you’re working with large volumes of data. They enhance efficiency, ensure transparency in your analysis process, and allow for an audit trail. They also encourage deeper engagement with data, often surfacing patterns you might have missed if you were coding manually. And for collaborative research, they offer ways to compare inter-coder reliability, helping to improve consistency and reflexivity.
I explored some of the most widely used CAQDAS tools. NVivo is perhaps the most common, especially in universities, and many institutions—including the University of Georgia—offer licenses for students and faculty. Others include MAXQDA, Dedoose, QDA Miner, HyperRESEARCH, and ATLAS.ti. While many require subscriptions, the functionality they offer can significantly streamline your research process.
Now here is where things get especially interesting—CAQDAS and AI.
Many of these tools are beginning to integrate AI to assist with coding and theming. For example, in MAXQDA, you can now use plain language to generate codes. The system highlights your transcript, generates themes, and can even suggest coding frameworks. This is a huge win for those who may find traditional coding complex or overwhelming.
But even more fascinating is this shift: researchers are increasingly using stand-alone AI tools, like ChatGPT, to analyze qualitative data. With the right prompt strategies, these tools can generate codes, develop themes, and interpret findings—sometimes even identifying supporting quotes and page numbers from the original transcripts. My own ongoing research (Nyaaba et al., 2025) explores this very shift, showing how AI, when guided by thoughtful prompts, can enhance transparency and accuracy in inductive thematic analysis.
This brings up an important question: Are we moving towards AI-assisted qualitative research as the norm? Will tools like CAQDAS become obsolete as AI grows smarter and more accessible? Or will they evolve together—blending human insight with machine efficiency?
Of course, CAQDAS and AI both come with limitations. Overreliance on automated coding might lead to coding overload or even disengagement from the data. It can also limit interpretive depth if not carefully managed. These tools also have a learning curve and often require time investment and training. And for AI, hallucinations and inaccuracies are still a concern, especially when interpretations are involved.
That is why we must embrace best practices. We must remember that these are tools, not methods. The role of the researcher—your judgment, your contextual understanding, your interpretive skill—is irreplaceable. Whether you are using traditional CAQDAS tools or experimenting with AI-powered analysis, your engagement with the data must remain central. Proper planning, alignment with research goals, and regular cross-checking of outputs are essential.
To wrap up: CAQDAS empowers researchers. It helps manage data efficiently, enhances transparency, supports collaboration, and facilitates complex analysis. But it must be used with care and thoughtful judgment. As AI continues to evolve and integrate with CAQDAS, we are witnessing a transformation in how we engage with qualitative data.
So, I leave you with this final question:
What do you think the future holds for CAQDAS and AI in qualitative research?
Are we on the path to a fully AI-assisted analysis process? Or will human insight continue to lead the way?
Thank you for taking this journey with me.
References
CAQDAS Project, Surrey. (2024). Retrieved from https://www.surrey.ac.uk/computer-assisted-qualitative-data-analysis/about#:~:text=The%20CAQDAS%20Networking%20Project%20was,and%20its%20adoption%20by%20researchers
Gibbs, G. R. (2013). Using software in qualitative analysis. In U. Flick (Ed.), The SAGE Handbook of Qualitative Data Analysis (pp. 277-294). Sage.
Nyaaba, M and Min, SungEun and Abiswin Apam, Mary and Acheampong, Kwame Owoahene and Dwamena, Emmanual and Zhai, Xiaoming, Optimizing Generative AI’s Accuracy and Transparency in Inductive Thematic Analysis: A Human-AI Comparison (March 11, 2025, preprint). Available at SSRN:https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5174910
Perkins, M., & Roe, J. (2024). Academic publisher guidelines on AI usage: A ChatGPT-supported thematic analysis. F1000Research, 12, 1398
St. John, W., & Johnson, P. (2000). The pros and cons of data analysis software for qualitative research. Journal of Nursing Scholarship, 32(4), 393–397.
Vignato, J., Inman, M., Patsais, M., & Conley, V. (2022). Computer-assisted qualitative data analysis software, phenomenology, and Colaizzi’s method. Western Journal of Nursing Research, 44(12), 1117–1123.
Wolski, U. (2018). The history of the development and propagation of QDA software. The Qualitative Report, 23(13), 6–20
QualPage 