Weekly PhD Newsletter | SIGN UP NOW Weekly PhD Newsletter | SIGN UP NOW

Tweeting about thorny choices: using large-scale Natural Language Processing to analyse public debates about decision-making in healthcare


   Information School


About the Project

We are delighted to announce this fantastic opportunity to conduct a fully funded PhD project (by EPSRC) at The University of Sheffield (available to UK home students only). The studentship covers a PhD project in the areas of computational social media analytics for the understanding of public debates on healthcare decision making. The PhD will be hosted at the Information School (iSchool) and co-supervised by staff in the iSchool and the Department of Sociological Studies. The project will be multidisciplinary and offers excellent opportunities to develop knowledge and skills in quantitative (e.g., Natural Language Processing, Data Science) and qualitative (e.g., discourse analysis, thematic analysis) research methods. You will be supervised by a team of experts on NLP and social media analytics as well as in sociological theory, social science methodology, and healthcare decision making. The University of Sheffield offers a wide range of career development opportunities, such as the ThinkAhead programme (https://www.sheffield.ac.uk/medicine-dentistry-health/research/think-ahead). For a more detailed project definition, see below. 

Eligibility and process:

To apply, follow the link https://www.sheffield.ac.uk/postgraduate/phd/apply. When completing the application form, please mention this opportunity and reference. The successful candidate must be able to start in late March 2023. Shortlisting and interviews are likely to take place in early February 2023. 

For entry requirements, see the link above. In addition, candidates should have good knowledge in statistics and mathematics as well as good programming skills. Knowledge and experience in one or more of the following areas are desirable: research experience in any field, NLP, data mining, machine learning, statistical modelling methods, sociology, social policy, health policy.  

Project details:

As Twitter is increasingly used in political discourse, automated large-scale extraction and analysis of argumentation/debate from tweets is gaining in importance. Yet studies in this area are rare; conversations on Twitter are composed of long threads of short tweets that may cross-reference Web resources and form complex relationships, thus creating major challenges for researchers. 

One set of debates that has been intensified by the global COVID-19 pandemic concerns decision-making in healthcare. Decisions about allocating scarce health resources have attracted particularly vehement disputes. Disputants have included health professionals, disability activists, patient organisations, lawyers, journalists, and ethicists. However, there have been calls for a wider debate involving the general public to inform policy and guidance on healthcare decision-making. Indeed, while the ethical discussions and media influences on decision-making frameworks have been extensively researched, much less is known about the public’s perceptions of, and judgements and arguments regarding, healthcare decisions during the pandemic. 

The project will develop novel NLP methods of argumentation mining on Twitter. It will combine these with thematic and discourse analysis to mine and analyse the largest COVID-19 Twitter dataset (01/02/2020-31/03/2021, with approximately 80,000,000 tweets from the UK). During this period, national lockdowns reduced in-person interaction, which fuelled online activities. The project will focus on two cases from the UK that revolve around controversies about clinical guidelines and about Do Not Attempt Cardio-Pulmonary Resuscitation Orders. 

Thus the project will aim to: (i) innovate NLP methods for analysing argumentation on social media; (ii) provide vital insights into public views that can inform policy and guidelines for some of the most challenging decisions professionals in the health system must take.


Email Now


Search Suggestions
Search suggestions

Based on your current searches we recommend the following search filters.

PhD saved successfully
View saved PhDs