In a recent study published in the journal Nature, researchers looked at four separate studies across 11 countries to systematically evaluate the impacts of motivational bias on historical narratives regarding the COVID-19 pandemic. Their results show that the strength of personal bias, whether towards vaccination or from media discourse, can significantly alter memory associated with the pandemic. They further discuss how this might impact historical narratives about the pandemic, which in turn would affect future pandemic policy and preparedness. They recommend that future pandemic measures focus on long-term effects on societal trust and cohesion, not just addressing immediate public health implications.

Study: Historical narratives about the COVID-19 pandemic are motivationally biased. Image Credit: DisobeyArt / Shutterstock

What do we remember about the COVID-19 pandemic?

Large-scale surveillance data collated by the World Health Organization (WHO) reveals the coronavirus disease 2019 (COVID-19) pandemic to be one of the worst in human history. Since the beginning of the outbreak in late 2019, the pandemic has infected more than 771 million people and claimed almost 7 million lives globally. The development of anti-COVID-19 vaccines and large-scale global vaccination drives resulted in a substantial decline in infection and mortality, courtesy of which most pandemic restrictions have been lifted in 2023.

This ‘post-pandemic phase’ has been characterized by reviews of COVID-19-related policies and efforts to prepare for future disease outbreaks. Despite the availability of quantitative data from surveillance and surveys, these reviews and efforts are influenced by public and media opinion, both susceptible to personal bias.

“Because memory formation is a constructive process, retrospective narratives about historical events such as the pandemic are at risk of significant distortion. Beyond simple forgetting, recall and ex-post evaluation are prone to various forms of bias, reflecting differences in motivation and purpose (for example, a wish to conform with one’s own or the prevailing opinion).”

In the present study, the authors propose that evaluations of the recent pandemic are skewed by individual bias and that most, if not all, of this bias is negative, given the high costs universally incurred from COVID-19. They use views towards vaccination as an example – regardless of personal beliefs, most individuals were forced to comply with government policy that required vaccination. However, polarizing animosity towards individuals who opposed their beliefs may have led one faction to discriminate against the other, thereby altering an unbiased recall of the pandemic as a whole.

About the study

The researchers report four separate empirical studies examining the type and severity of personal bias in narratives about the COVID-19 pandemic. Study 1 assessed recall and evaluation bias, tinted by evaluations of political restrictions and policy and overshadowed by the prevalence of opinion-based cohorts, especially those pertaining to vaccination. Studies 2 and 3 assessed techniques and strategies by which recall bias may be attenuated. Finally, Study 4 aimed to investigate nation-specific bias in evaluations of the pandemic and examine if these evaluations spilled over to post-pandemic preparations.

The role of perceptions

Study 1 was conducted on a cohort of 1,644 German adults surveyed in the summer of 2020 or winter of 2020-21 and then repeat surveyed in late 2022. Of these, 1,216 (74%) had received at least one vaccination dose against the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus. The survey comprised questions on their perceptions and fears regarding the pandemic, affinity to government and scientific recommendations, and their frequency of adherence to social distancing and anti-COVID-government policy. They were further queried about their current life satisfaction and to what degree they felt the pandemic was exaggerated.

Linear regression analyses of their perceptions across the two time points revealed that their recall was significantly affected by variables regarding perceived risk, compliance behavior, and trust were strongly associated with current perceptions of the pandemic’s influence on their well-being. These factors, in turn, were influenced by their identification of vaccination status (a proxy for whether they supported or opposed vaccination). Vaccinated individuals recalled the pandemic and risk of infections as more severe than those who remained unvaccinated.

Affinity with government policy was observed to have similar effects, with individuals who readily complied with government restrictions more likely to be vaccinated and, in turn, have a higher infection risk perception than those who found government policy exaggerated and inappropriate.

“This indicates that greater bias when recalling the past was associated with a more extreme evaluation of political action—in either direction.”

Can we reduce the effects of bias on future pandemic policy?

Study 2 had two main objectives – 1. To investigate if monetary incentives could result in more accurate recall and, 2. If providing metacognitive information that recall bias is rampant could result in more precise recall. These objectives derive from previous research, which has shown that both factors in other domains coax individuals into correcting their own judgments, thereby reducing bias.

The study cohort comprised 3,105 German and Austrian participants studied during January 2023, of whom 71% had received vaccination against the pandemic. Participants were randomly assigned to the intervention conditions (monetary or informative) or control groups, following which their perceptions of the pandemic were queried. Interventions included the chance to win 100 Euros in the monetary cohort and information regarding the extent of bias in the metacognitive cohort.

Across both intervention and control groups, linear regression analyses once again revealed inherent biases between vaccinated and unvaccinated participants. While unable to statistically alter excepted vaccination-status-specific response, both interventions did result in a non-zero change in perceptions and recall, indicated by results exceeding regions of practical equivalence (ROPE) for infection probability.

Study 3 aimed to investigate if the strength of interventions (in this case, incentives) would alter Study 2’s findings and involve a greater probability of winning the 100 Euro cash prize, given participants’ greater recall accuracy. The study comprised 906 vaccinated German adults surveyed in July 2023. Participants were divided into intervention and control cohorts following the randomization methodology of Study 2. Study findings revealed that despite bias still influencing an overestimate of pandemic-related risk assessment, recall accuracy was significantly improved over Study 2.

“…a mixed-effects regression (controlling for multiple answers from the same individual, including n = 5,360 answers, see Extended Data Table 5) revealed that offering an incentive decreased directional bias (main effect: b = −0.35, s.e.m. = 0.10, P = 0.001) and increased the influence of past ratings (interaction effect of incentive and past ratings: b = 0.08, s.e.m. = 0.02, P = 0.002), indicating a reduction of recall bias.”

Taken together, the findings from studies 2 and 3 reveal that metacognitive interventions are unable to affect cohort-specific personal bias, and while stronger incentive can reduce biased recall, it fails to eliminate it entirely.

Does it matter where you live?

Building upon previous 2020 research in 10 countries, namely Australia, Italy, Germany, Mexico, Japan, South Korea, Sweden, Spain, the United Kingdom (UK), and the United States (US), Study 3 comprised 5,121 participants from these countries, 88% of whom had received vaccination against COVID-19. Spain had the highest vaccination rate (96%), while Japan had the lowest (72%).

The methodology comprised the same questionnaires used in studies 1-3, and results were compared against the previous benchmark study that Study 4 followed up on. Most participants across nationalities were found to overestimate perceived infection probability, and with Japan and Mexico being notable exceptions, underestimated the severity of illness. Bias pertaining to government effectiveness was found to vary significantly across evaluated countries (31% in Italy and 81% in Japan).

Perceived illness severity was associated with government effectiveness evaluations – overestimating severity was correlated with higher perceived government effectiveness.

“The results suggest that although a vaccinated majority has a more positive view of the measures taken during the pandemic, as warranted by respective perceptions of the past, a small segment of society has a strong desire to take revenge on those who spoke out or took responsibility during the pandemic. In summary, we observed polarized evaluations of the pandemic and indicators of social tension in many countries and across continents.”

Conclusions

In the present report, researchers carried out four studies aimed at evaluating the nature and strength of bias in historical assessments and memory recall pertaining to the COVID-19 pandemic. Their findings revealed that individual bias plays a substantial role in affecting memory recall and perception, with people who refrained from vaccination depicting polar opposite assessments of the severity of the pandemic and the effectiveness of government interventions compared to those who were vaccinated.

“…the four studies reported here highlight the complex nexus of attitudes, memories and behaviors surrounding the COVID-19 pandemic. Motivational factors related to identity and behavior in extreme situations seem pivotal in this context, linking the past to biased memories and future behaviors. Researchers and policymakers must pursue a better understanding of these connections to develop more fruitful ways of learning from the past to improve crisis preparedness and response.”