Unsilencing Through AI-Generated Images: Contrasting Participants’ and Researchers’ Images as Tools for Visualizing Tacit Experiences

Abstract

We explore the role of image generation through artificial intelligence (AI) in enhancing participants’ involvement in reflective interviewing, contributing to a more nuanced understanding of AI’s different roles in supporting reflection through images. We gave multicultural participants – second-generation immigrants or those who grew up in multicultural households – the option to generate their own images or assemble a collage from pre-prepared images to visually represent their experiences of growing up with plural and partial cultural influences. We found that researchers’ pre-prepared images limited the introduction of biases embedded in AI training data and reduced technological disruptions while allowing for multiple avenues of interpretation and improved accessibility. However, researchers’ images have the potential to silence participants’ unique experiences by limiting the availability of visual material. We discuss the potential and challenges of AI in extending the involvement of marginalized and silenced groups in qualitative research through artistic expression.