Applying machine-learning to rapidly analyze large qualitative text datasets to inform the Covid-19 pandemic response

Comparing human and machine-assisted topic analysis techniques

Bibliographic Data

ID22074362
AuthorsLauren B Towler (0000-0002-6597-0927, University of Bristol, corresponding author), Lauren Towler, Paulina Bondaronek (0000-0003-0096-1234, Department of Health and Social Care), Trisevgeni Papakonstantinou (0000-0001-5116-1531, Department of Health and Social Care), Richard Amlôt (0000-0003-3481-6588, UK Health Security Agency), Tim Chadborn (0000-0001-6264-3843, Department of Health and Social Care), Ben Ainsworth (0000-0002-5098-1092, National Institute for Health and Care Research), Liz Yardley (0000-0002-3853-883X, University of Bristol), Lucy Yardley
Year2023
Volume11
Pages1268223-1268223
Publication date2023-10-31
Peer ReviewedYes
Open AccessYes
TypeARTICLE
VenueFrontiers in Public Health (JOURNAL)
Journal identifiersISSN: 2296-2565 • E-ISSN: 2296-2565
PublisherFrontiers Media SA (PUBLISHER • CH)
DOI10.3389/fpubh.2023.1268223
PMID38026376
OpenAlexW4388072077
LanguageEN
Citations received5
References cited33

Introduction: Machine-assisted topic analysis (MATA) uses artificial intelligence methods to help qualitative researchers analyze large datasets. This is useful for researchers to rapidly update healthcare interventions during changing healthcare contexts, such as a pandemic. We examined the potential to support healthcare interventions by comparing MATA with "human-only" thematic analysis techniques on the same dataset (1,472 user responses from a COVID-19 behavioral intervention). Methods: In MATA, an unsupervised topic-modeling approach identified latent topics in the text, from which researchers identified broad themes. In human-only codebook analysis, researchers developed an initial codebook based on previous research that was applied to the dataset by the team, who met regularly to discuss and refine the codes. Formal triangulation using a "convergence coding matrix" compared findings between methods, categorizing them as "agreement", "complementary", "dissonant", or "silent". Results: Human analysis took much longer than MATA (147.5 vs. 40 h). Both methods identified key themes about what users found helpful and unhelpful. Formal triangulation showed both sets of findings were highly similar. The formal triangulation showed high similarity between the findings. All MATA codes were classified as in agreement or complementary to the human themes. When findings differed slightly, this was due to human researcher interpretations or nuance from human-only analysis. Discussion: Results produced by MATA were similar to human-only thematic analysis, with substantial time savings. For simple analyses that do not require an in-depth or subtle understanding of the data, MATA is a useful tool that can support qualitative researchers to interpret and analyze large datasets quickly. This approach can support intervention development and implementation, such as enabling rapid optimization during public health emergencies

Codebook · Content analysis · Data science · Health care · Information retrieval · Linguistics · Natural language processing · Qualitative research · Sociology · Storytelling · Thematic analysis · Triangulation · Computational and Text Analysis Methods · Computer Science · Mental Health via Writing · Qualitative Research Methods and Applications · Artificial Intelligence