A qualitative study exploring researchers' experience of involving and engaging seldom-heard communities in big data research 2023 Teodorowski et al

Abstract

Background
Big data research requires public support. It has been argued that this can be achieved by public involvement and engagement to ensure that public views are at the centre of research projects. Researchers should aim to include diverse communities, including seldom-heard voices, to ensure that a range of voices are heard and that research is meaningful to them.

Objective
We explored how researchers involve and engage seldom-heard communities around big data research.

Methods
This is a qualitative study. Researchers who had experience of involving or engaging seldom-heard communities in big data research were recruited. They were based in England (n = 5), Scotland (n = 4), Belgium (n = 2) and Canada (n = 1). Twelve semistructured interviews were conducted on Zoom. All interviews were audio-recorded and transcribed, and we used reflexive thematic analysis to analyse participants' experiences.

Results
The analysis highlighted the complexity of involving and engaging seldom-heard communities around big data research. Four themes were developed to represent participants' experiences: (1) abstraction and complexity of big data, (2) one size does not fit all, (3) working in partnership and (4) empowering the public contribution.

Conclusion
The study offers researchers a better understanding of how to involve and engage seldom-heard communities in a meaningful way around big data research. There is no one right approach, with involvement and engagement activities required to be project-specific and dependent on the public contributors, researchers' needs, resources and time available.

Patient and Public Involvement
Two public contributors are authors of the paper and they were involved in the study design, analysis and writing.

Open access, https://onlinelibrary.wiley.com/doi/10.1111/hex.13713

 

Medicine can't do Big data in the traditional sense, it usually involves working with giant databases with raw data dating back years if not decades.

All medical data is heavily filtered. Big data requires raw data, not interpretation. Trying a Big data approach without raw data is pointless, it's the equivalent of reviewing your own notes for new insights, or trying to make out details on an image that has been JPEGed 50% 10x over, the data is irretrievably lost. Medical data is selective by nature, in many cases distorted. As a choice, unfortunately healthcare hasn't managed to deal with large masses of data yet, still too much manual paperwork and traditions that make it culturally impossible.

Maybe in some edge cases like medical imaging, but it's still very selective data.