• Barry Singleton

Scaled Insights COVID-19 General Population Research Published in Frontiers

Updated: Apr 9


As the UK Government imposed strict and unprecedented restrictions to the population in order to tackle the coronavirus (COVID-19) outbreak, Scaled Insights & University of Leeds co-created a general population survey.


You can read the peer reviewed academic paper here:

https://journals.sagepub.com/doi/pdf/10.1177/1757913920979332


The survey aimed to understand adults’ thoughts and behaviours relating to the coronavirus (COVID-19) outbreak, particularly due to the extraordinary and rapidly changing impact that it is having on the way people live. Designed to provide an understanding of how adults have responded and been affected by the pandemic the findings provide vital insights for local government, national government and future messaging as the situation continues to evolve.



After ethical approval was granted by the School of Psychology Research Ethics Committee at University of Leeds (REC number PSYC-20), 1126 participants including 495 adults residing in the West Yorkshire Combined Authority jurisdiction, completed a survey to explore adults’ thoughts and behaviours about the COVID-19 outbreak. This survey specifically asked adults about:

a. COVID-19 thoughts and behaviours including knowledge of symptoms, actions to reduce risk, concern infection, spread, impact on the economy, NHS and employment opportunities, and government response b. the impact on employment such as approach to work (furlough, work from home, unpaid leave); productivity, ability to perform role c. the impact of home schooling on work and health d. health and lifestyle behaviours such as sleep, alcohol, diet, physical activity e. wellbeing f. the sources that adults use to receive information about COVID-19.


By collecting a combination of quantitative and qualitative data, we obtained sentiment and personality scores from language samples using Scaled Insights’ proprietary software, natural language processing, and machine learning methods. Survey participants were categorised using a clustering algorithm based on the sentiment and personality scores and we were able to identify two distinct clusters:

Positive Cluster (positive sentiment, more trusting, dutiful, happy)

Negative Cluster (negative sentiment, more neurotic, insecure, stressed)

In general scores for stress and depression are quite high, but Negative Cluster has significantly higher scores than Positive Cluster. The two clusters reported significantly different attitudes and behaviours:

Negative Cluster was:


(1) more concerned about becoming infected with COVID-19 and having more severe illness, (2) self-isolating more often but shopping online less, (3) reported greater impact on sleep, diet, and physical activity, (4) reported lower wellbeing score.

This is useful to know because if we can categorise someone as Negative Cluster from their language then we can offer them additional support and information.