How to Analyse Large-Scale Qualitative Data without losing the plot
Insights from a Randomised Controlled Trial of LSD Microdosing for Major Depressive Disorder
This midday talk will take place online, please join our mailing list for the link. This will be sent out the week before and again on the day of the talk.
Qualitative interviews are essential in exploring the experiences of participants. However, challenges emerge as trials scale up in size. In smaller investigations ( e.g. n = 20), it is feasible for one researcher to transcribe and analyse the data. However, in larger trials—such as our randomised controlled trial of LSD microdosing in major depressive disorder (n = 80 interviews), the volume of qualitative data becomes unmanageable. Traditional thematic analysis is prohibitively time-intensive, and although saturation can, in theory, be reached, it is nearly impossible to distil so much material into a cohesive argument manually.
To address this problem, this presentation will cover the potential of using a semi-supervised machine learning pipeline to analyse qualitative data at scale. The proposed pipeline may offer a new way to extract themes from large qualitative data sets that would otherwise be unattainable using traditional approaches. The presentation will describe the pipeline, how it was applied, and the strengths and limitations of this approach. It will also present thematic results of this pipeline, highlighting differences in participant experience between the LSD and placebo groups.
Bio:
Carina Donegan is a PhD candidate in Psychological Medicine at the University of Auckland (thesis submitted May 2026). She completed her undergraduate training in Psychology and Neuroscience at the University of Otago. Her doctoral research centred on open-label and randomised controlled trials investigating LSD microdosing for major depressive disorder. She has a primarily psychological focus, examining expectancy effects, changes in depressive symptoms, and participants’ subjective experiences. Her work integrates quantitative and qualitative approaches, with a particular interest in developing rigorous methods for analysing large-scale qualitative data. She is currently seeking postdoctoral or further research opportunities.