Scientists and policy makers have been concerned about the emergence of an influenza pandemic for which there is neither a strain-specific vaccine nor sufficient antiviral medications [1
]. Social distancing is effective in reducing infection rates within communities [2
]. Measures of social distancing undertaken in Italy, Spain, India, and most of the United States are drastic [6
]. Local governments across the world impose four practices of social distancing including isolation of people with probable influenza; voluntary home quarantine for people with confirmed or probable influenza; preventing the congregation of people in community or employment settings; and social distancing of adults [4
The implementation of social distancing requires significant public cooperation [4
]. Previously, compliance was high in the early stages of the pandemic but decreased significantly over the course of several months [9
]. Further, the compliance of people belonging to racial and ethnic minorities and to low socio-economic groups has been low [10
]. Practical, logistical and ethical limitations of social distancing have also affected compliance [5
]. Compliance with the policy of social distancing may maintain the resilience of communities. Designing communication interventions may reduce the burden of the psychological and social consequences of social distancing and reduce trauma that may transcend generations, constraining coping capacities [12
]. In previous pandemics, people with low tolerability to policies fled from infected cities, participated in riots, attacked government officials and harmed those suspected of plaguing [1
]. Compliance with social distancing may also be influenced by one’s level of perceived risk, one’s perception of individual and family disruption, and the perceived effectiveness of agencies dealing with the crisis [13
]. The resilience of communities may depend on the extent of compliance with social distancing maintained in those communities. The maintenance of compliance will depend on appropriate communication [16
]. Thus, communication-/messaging regarding social distancing, must consider reactions to the policy, and avoid the creation of resistance, which can endanger compliance [17
]. The purpose of this study is to evaluate the specific communication messages that have the potential to act as drivers of compliance to social distancing.
Most studies on public behaviors in a pandemic tested the extent of communication that the public needs for compliance [16
]. Research on the impact of specific communication messages on compliance is scant [17
]. Previous studies called to close the gap and examine how to persuade the public to comply with social distancing [15
]. Studies also suggested using controlled research designs rather than observational studies [20
]. This study responds to previous calls seeking to discover communication messages that are likely to resonate positively with the individual, thus enhancing compliance. Higher compliance with social distancing will allow hospitals to manage their overflow, until a vaccination is available [1
]. Research also acknowledged that a strategy using various communication messages to optimize compliance with social distancing may be more effective than using one message for all [21
]. Furthermore, during pandemics, the communication resources of governments become scarce. Effective messaging enables policy makers to allocate resources based on real, immediate, and relevant data.
Effective health communication encompasses narrative messaging [22
]. Narrative messages may be effective in the context of pandemic behavior as well, because they are the basic mode of interaction that people use to influence others [24
]. The impact of messages may be explored by asking questions about whom, under what circumstances, how, and when does each message achieve an optimum effect [28
]. The power of specific messages may vary across individuals, possibly as a consequence of their perceived risk. Furthermore, the effect of narrative messaging on compliance to social distancing may depend, in part, on the extent to which people “identify” with the message [28
]. People may, therefore, be defined by the similarity in their pattern of responses to narrative persuasion messages (i.e., mindset segments) regarding social distancing. Therefore, we hypothesized that various messages will be strong drivers of compliance, while others will be weak drivers. We further hypothesize that the power of the messages as drivers of social distancing will differ based on the similarity in patterns of response to the messages.
To simplify the analysis, we present only messages with positive coefficients, driving agreement with compliance. Negative coefficients mean either that the element is neutral (irrelevant for compliance) or counterproductive, driving non-compliance. Table 3
shows messages that drive compliance among Canadians. Table 4
shows messages that drive compliance among Americans. Strong drivers of compliance with social distancing for the total sample were “To practice social distancing, everyone should stay six feet apart from others” and “The media or central government should communicate the social distancing policy”. Respondents perceived the government as reactive rather than proactive, failing to understand the significance of COVID-19.
The additive constant shows the estimated, baseline likelihood that a respondent ‘agrees’ with the message. The data show that the readiness to comply is higher among Canadians than Americans, as shown in Table 3
and Table 4
3.2. Mindset Segments
Three mindsets emerged from mathematical K-clustering based upon similarity in the patterns of responses to the individual messages [41
]. The mindsets transcend the typical patterns used to divide respondents, namely WHO they are, summarizing how people THINK, or the nature of what people feel about social distancing.
Mindsets 1 (MS1) and 2 (MS2) are similar in both countries. People belonging to MS1, “Tell me exactly what to do”, may be driven to higher compliance by messages regarding ways to ensure compliance. Strong messages were “To ensure compliance, people should be under military lockdown”, “Compliance may be ensured by restricting shopping for food (3 at a time) and pharmacy (1 at a time)”, and “Designated young volunteers should shop for the elderly and disabled”. People belonging to MS2, “Pandemic onlookers”, may be driven to higher compliance by messages relating to the perceived risk of the virus. Strong messages were “A dangerous virus is spreading wildly”, “All the news seems to be about the COVID-19 virus”, and “Health experts suggest…government is reactive rather than proactive to the pandemic”. People belonging to this mindset may agree with any social distancing practice.
Mindset 3 (MS3) differs dramatically between countries. Canadian respondents appear to respond to authority, whereas American respondents appear not to respond to authority. Canadians belonging to MS3, “Bow to authority”, are driven to higher compliance by the communicator of the policy and believe that either provincial government, federal government, the media or religious clergy should communicate the social distancing policies. Americans belonging to MS3, “Tell me how”, may be driven to higher compliance by messages that focus on the practices of social distancing. Strong messages were “Everyone stays 6 feet apart”, “Socialize and work only from home on the internet”, “Confined to within 300 feet from home”, and “Wear your mask everywhere”.
3.3. Personal Viewpoint Identifier
For effective knowledge translation, the emergence of these findings requires an innovative way to assign a new person or group in the population to these previously discovered mindsets. To translate the knowledge derived in this study for implementation, we developed a prediction tool called a Personal Viewpoint Identifier (PVI). The PVI tool is a method by which health authorities may assign a person or group in the population to a mindset. This is based on the summary data, converting six of the strong distinguishing messages to questions answered by a “agree or disagree”. The six messages were chosen using a Monte-Carlo simulation where random variability is added to the data, and the most discriminating messages emerge in the face of random variability [42
]. Based on the answers to the questions in the PVI, the individual or group is assigned to one of the three mindsets, and thus, the proper messages are established for that individual (Table 5
This study examined the influence of numerous communication messages on compliance with social distancing in North America. Communication messages represent relevant, topic-specific ideas. Ordinarily, one would tailor the messaging to who a person IS demographically. This study suggests tailoring messaging by the way the person THINKS about compliance to a health policy in a pandemic, particularly because messaging shaping public behavior during a pandemic is an undertested phenomenon. This study revealed messages that drive positive responses to social distancing, varying by individuals or groups that think alike. The first hypothesis was corroborated as some messages were strong drivers of compliance and some were weak drivers of compliance. Results echo previous research claiming that the extent to which messages drive social distancing may depend, in part, on the extent to which people “identify” with the different narrative messages [28
The second hypothesis was also corroborated as three mindsets emerged from the similarities in patterns of responses to narrative communication messaging regarding social distancing. Even during the midst of a pandemic, when social distancing is critical and novel, approaching each of the three different mindset segments with appropriate communication messages should enhance compliance with social distancing. Health policy makers and government leaders should take into consideration the proclivities and sensitivities of these three mindsets when suggesting how to properly implement the policy of social distancing over time.
Findings of this study should be interpreted in light of its limitations. Our study variables reflect our choice of constructs, which was based on prior literature, but does not encompass additional independent variables that are not yet acknowledged as potential drivers of compliance with social distancing. Further, given the continuing pandemic, respondents may have been previously exposed to messages on social distancing in real life before participating in this study. The exposure to messages, the frequency of the exposure and the potential prime effect of messages may have influenced the extent of agreement with messages regarding social distancing. These external cofounders, however, cannot be controlled in a cross-sectional design.
Directions for Future Studies
Future studies may track the relationship between using communication messages by mindset segment membership and compliance with social distancing as measured by cellular data. Longitudinal future studies may also test the prime effect of communication messages and the effect of exposure to messages and their frequency on the level of compliance.