AI is everywhere these days. It’s in your pocket,  driving cars, selecting what show we should binge on, and which search results we should see first. So, what about qualitative research? 

I was curious to know how social science and humanities researchers are utilizing AI and how they are dealing with the more complicated issues that it raises. While AI offers some amazing tools to streamline and enhance the way we approach qualitative research, it also brings up a host of questions about how much we can—or should—trust machines to do the heavy lifting in fields that thrive on human insight and interpretation.As a geographer, I’m concerned about AI’s environmental impacts, often overlooked.

As a geographer, I am also concerned with the environmental impacts of AI development, which are often overlooked in discussions about its technological promise. Kate Crawford, in “Atlas of AI”, sheds light on the massive ecological footprint AI leaves behind—from the extraction of rare earth metals to the immense energy consumption required to train AI models. These processes contribute to resource depletion, environmental degradation, and significant carbon emissions, raising ethical questions about the sustainability of AI systems. The hidden costs behind AI’s seemingly virtual world challenge us to rethink its true environmental and social impact.

So, I thought it could be helpful to review some research on these issues, and it could help me think about the potentials and pitfalls of AI in qualitative research in this series. For the first part I mainly looked for ways that AI is making things faster and more accessible for researchers and in the second part, I review the ethical concerns and challenges that come with it. Also, in keeping with the spirit of our work at the Digital Initiatives, I’ll introduce some of the best open-source tools that make it easier for everyone to harness AI’s power in qualitative analysis.

Potentials:

Efficient Data Processing and Thematic Analysis.

Let’s face it—manually coding qualitative data can be a time-consuming task, and large language models (LLMs) like ChatGPT can seriously speed up this process. AI allows researchers to work more efficiently and focus on interpretation by automating repetitive tasks like coding and thematic analysis. Studies show that AI can handle massive datasets, identify key themes, and streamline analysis in a fraction of the time (Zhang et al., 2024).

Take Taguette as an example. This open-source platform enables researchers to easily tag and code data, export results, and navigate qualitative data. By handling the heavy lifting, AI frees up time for researchers to engage with deeper, more complex aspects of the data. Similarly, platforms like AcademiaOS leverage AI to assist in grounded theory development, supporting human researchers in managing large amounts of information (Übellacker, 2023).

This expansion of AI utilization in research has also helped democratize the field to a certain extent. Historically, advanced tools for qualitative analysis were locked behind expensive paywalls of high-end, proprietary software like NVivo, or Atlas Ti.

Open-Source Platforms for AI-Driven Qualitative Research

–   Taguette

Taguette is an open-source qualitative research tool that allows researchers to code and tag qualitative data efficiently. Its user-friendly interface makes it accessible to both novice and experienced researchers. Since it’s free and runs on local machines, it’s a great solution for those who want to ensure their data is secure.

–   Weft QDA

For researchers on a budget, Weft QDA offers basic qualitative coding functionalities without the cost of commercial software. Its simplicity makes it a great choice for those new to qualitative research, while its open-source nature ensures that users can adapt the tool to their specific needs.

–   AcademiaOS

AcademiaOS is designed to assist researchers with grounded theory development, using AI to automate much of the coding and analysis. While it handles the heavy lifting, it still requires human oversight, ensuring that the theoretical framework remains sound.

 

References:

Crawford, K. (2021). The Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. Yale University Press.

Übellacker, T. (n.d.). AcademiaOS: Automating Grounded Theory Development in Qualitative Research with Large Language Models. Ar5iv. Retrieved October 5, 2024, from https://ar5iv.labs.arxiv.org/html/2403.08844

Zhang, H., Wu, C., Xie, J., Iyu, Y., & Cai, J. (n.d.). Redefining Qualitative Analysis in the AI Era: Utilizing ChatGPT for Efficient Thematic Analysis. Ar5iv. Retrieved October 5, 2024, from https://ar5iv.labs.arxiv.org/html/2309.10771