Public trust in science remains strong during pandemic, but study suggests some decrease in late 2020

Throughout the coronavirus pandemic, scientific research has been widely used to communicate about the disease to the public. New research to understand public views of coronavirus science has found that confidence in science among people in England from March to November 2020 was good overall although declined over this time. People who had been shielding had greater trust in November 2020 compared with their description of the views that they held in at the start of the pandemic in March 2020.

The research team at the University of Bristol commissioned YouGov to carry out a survey in November 2020 and their findings are published on Wellcome Open Research. The survey, led by the University’s Elizabeth Blackwell Institute, looked at: levels of public trust in scientists and scientific information; changes in trust between March and November 2020; views about communication of scientific uncertainty; confidence in the accuracy of scientific findings; and views about whether public information is an accurate representation of coronavirus science.

YouGov surveyed 2,025 participants in England; 40.5 per cent were over 55 years old, 51.1 per cent were female, 87.7 per cent identified as white and 12.3 per cent identified as members of an ethnic minority or of mixed ethnicity. Personal characteristics, such as age, gender, ethnicity, keyworker status, carer status, shielding, and coronavirus exposure, were used to examine how different groups of people felt about science, scientists and the communication of science relating to the coronavirus pandemic. The data revealed some patterns in views of science which could help improve communication of public health measures used to manage the pandemic.

The study found trust was highest among older respondents and among those who identified as of white ethnicity. This supports the findings of the Wellcome Monitor Survey 2020, which found that people of an ethnic minority background had lower levels of trust in information from health sector and government sources than people who identified as white.

The levels of reported trust in scientific information about coronavirus were lower in November 2020 than at the start of the pandemic in March 2020, with a decline in the proportion of people responding that they trusted coronavirus science ‘a lot’ and an increase in those saying that they ‘didn’t trust much/at all’.

Similarly, trust in scientists to conduct accurate and reliable research decreased slightly between March and November 2020. Despite the potential for recollection biases, people were asked to recall their feelings from months previously. The reported changes indicate trust in science fell over the first eight and a half months of the pandemic. The greatest variations in trust were found in people of different ages and ethnicities. People who were older expressed higher levels of trust than other age groups, as did those who were white.

Some of the most interesting changes in trust occurred in people of different health status. The highest levels of ‘trust a lot’ were in respondents who thought they had had coronavirus, but had tested negative. When reflecting on their views about coronavirus science in March 2020, those who were not shielding expressed the highest levels of ‘trust a lot’ or ‘trust a little’, while those who had been shielding, or who had someone in their household shielding, expressed the highest levels of ‘didn’t trust much/at all’. By November 2020, this pattern had reversed and people who were shielding were slightly more likely to trust coronavirus science ‘a lot’ or ‘a little’. However, levels of ‘trust a lot’ declined in both groups between March and November.

Just under half of respondents thought that uncertainty was communicated ‘not very much/not at all’. At the start of the pandemic, for example during government briefings, presentations did not usually include information about uncertainty around the predicted numbers of cases, hospital admissions, or deaths.

Around 60 per cent of respondents were ‘fairly’ or ‘very’ confident in the accuracy of coronavirus science, suggesting confidence in the remaining population needs to be improved. Finally, around 64 per cent of respondents strongly agreed or tended to agree that public information is a true representation of coronavirus science, and those who did so were more likely to be older, male, and of white ethnicity.

The findings show the need to improve the reporting of science, which could increase public trust, such as confidence in reports of science generally. The study also highlights the need to develop strategies of communication with younger people and with people of an ethnic minority background. The study team say that there is more research and analysis needed to understand the complex patterns that they saw in their work. Further work is also needed to understand if views are linked to concerns about the trustworthiness of the information sources, individuals and/or organisations, or in the science that underpins public information.

Rachael Gooberman-Hill, Professor of Health and Anthropology and Director of Bristol’s Elizabeth Blackwell Institute, said: “Our study indicates that although trust in coronavirus science is good, there is still room to improve trust and communication about science. Improved communication could increase public confidence in science. As well as detailed analyses to account for inter-relationships, further research could examine reasons behind change in trust over time including patterns by age, ethnicity, and shielding status.”

Paper

Public views of coronavirus science and scientists: findings from a cross-sectional survey‘ by Rachael Gooberman-Hill, Michelle L. Taylor, Ulrika Maude, Lucy Yardley, Richard Huxtable, Jo Stubbs, Tim J. Peters in Wellcome Open Research

/Public Release. This material comes from the originating organization/author(s)and may be of a point-in-time nature, edited for clarity, style and length. The views and opinions expressed are those of the author(s).View in full here.