The Shape of Agency: Fostering Agency in Qualitative Research through Data Visualization
Loading...
Date
2023-09-08
Authors
Ugaya Mazza, Larissa
Journal Title
Journal ISSN
Volume Title
Publisher
University of Waterloo
Abstract
Qualitative data analysis is important to the field of healthcare since it allows researchers to understand the lived experience of patients, practitioners, and everyone in between. However, qualitative data requires time and effort, which is not always available. A potential way to overcome this barrier is to use artificial intelligence as a tool to help researchers with data analysis. However, many qualitative researchers do not have the programming skills to use AI and are reluctant to lose their sense of agency when conducting research. As a potential way to bridge this gap, we explored the use of data visualizations to foster researcher agency and make using AI more accessible.
We used Design Science Research and developed a datavis tool prototype to map out how researchers perceive agency. A user centered design approach was used to design a non-functional data visualization tool with the assistance of 5 qualitative heath researchers. Two semi-structured interviews were used to facilitate the user centered design, the first to provide guidelines for the prototype and the second for testing the tool and altering any features considered confusing or lacking.
The results showed that qualitative researchers have a wide range of cognitive needs when conducting data analysis and for that, need a variety of visualizations to best accommodate their needs. Additionally, they place high importance upon choices and freedom, wanting to feel autonomy over their own research and not be replaced or hindered by AI. Despite this, participants were open to the idea of delegating tasks, so long as they could maintain the final choice on results.
Seven barriers were identified for the fostering of agency when conducting research with AI: full AI delegation, lack of transparency with results, no choice in how results are reached, excessive freedom with no guidance, lack of ability to make edits, no guidance on how a tool works, and restricted movements.
As potential solutions for these issues, five facilitators were found during the interviews. Those being: providing choices for different kinds of data visualization, explaining the AI process in simple language, the addition of co-creation tools, addition of guidance in navigation, and the ability to enable free movement.
Description
Keywords
data visualization, artificial intelligence, qualitative research, design