Skip to main content Skip to navigation

Changing people’s behaviour in times of Covid-19 - further expert comment

beachAs news reports of police struggling to persuade social gatherings to break up and go home begin to emerge, Katharina Lefringhausen, Helen Spencer-Oatey and Perry Hinton of the University
of Warwick Centre for Applied Linguistics and Dan Spencer from the University of Oxford look at how norm theory can be applied to tailor appeals to different groups, highlighting the role of shared personal values, shared social groups, and the 'provincial norm.'

Changing Peoples Behaviour in times of Covid-19. Part II:  Getting the Messaging Right for Different Audiences 

"Despite the recently installed lockdown in the UK due to the Covid-19 crisis, there are still various groups of the public who are not following the government’s plea to exhibit social distancing and self-isolation.

Although police enforcement will surely influence some people to behave as they have now been instructed, clearer messaging is still required to support this instruction. For example, only hours after the British Prime Minister’s announcement of lockdown on the 23rd of March, some Twitter users questioned how the police were going to keep track of people’s time spent outside or establish whether the people accompanying them were part of their own household.

Often these comments were paired with the idea that members of one’s own group are likely to have similar reservations about the newly required behaviours (e.g., “many people in my neighbourhood/people who live in small flats and with no garden will find it impossible to stay at home”).

In our previous piece we suggested ways in which Robert Cialdini’s (e.g. Cialdini, 2012) norm focus theory can offer some very helpful insights on this matter. Now we provide a more nuanced picture, focusing on the importance of audience, in the hopes of providing some practical advice.

To remind ourselves, Cialdini argues that there are two types of norm: a descriptive norm, which refers to a behaviour that is common or widespread, and an injunctive norm, which refers to a behaviour that is approved or disapproved of within a social group. Prescriptive injunctive norms identify what people should do, whereas proscriptive injunctive norms identify what people should not do.

Cialdini points out that people are motivated or influenced by both types of norms primarily when the norms are brought into focus and their attention is directed to them.

To what extent are we influenced by the person who is advising/telling us what to do?

For example, over the past few days, it has not only been PM Boris Johnson sending out messages to the public, including both proscriptive injunctive norms (e.g. don’t panic buy) and prescriptive norms (e.g. practise social distancing).

Others, from Hollywood stars like Arnold Schwarzenegger sitting in a hot tub and smoking a cigar to a team of Leeds nurses at their clinic, have all shared similar norm messages about how to behave during the outbreak.

arnie in hot tubFrom a norm focus theory perspective, Schwarzenegger and the nurses did well in drawing attention to their messages, but how may their messages be differentially received by different audiences?

Social psychologists suggest that when making decisions during times of uncertainty, people are most likely to follow the norms of those who appear most similar to themselves. This similarity can take different forms:

  • Social group membership (Cialdini, 2012): If we perceive ourselves to belong to the same social group as the messenger and have a key trait in common (e.g., gender or nationality), we are more likely to follow what they say. Thus, if members of the public strongly identify with the UK, they may be more likely to follow the injunctive norms expressed by the PM than by Arnold Schwarzenegger. If residents in Leeds have a strong local identity, they will be more likely to follow the nurses’ advice than that of the PM.
  • Endorsed personal values (Wan, 2015): When someone’s personal values align with those perceived to be endorsed by the messenger, that person is more likely follow their advice/instruction. This has been tested with data from the USA during general elections, where high compatibility between one’s personal values and those collectively perceived to be endorsed by the politician predicted people’s vote for this particular candidate (Wan, Tam, & Chiu, 2010). In other words, people who value self-indulgence may be more likely to follow Schwarzenegger’s message, given he may be collectively perceived to favour hedonism, too (e.g., sitting in a hot tub whilst smoking a cigar).
  • Shared environment/provincial norms (Goldstein, Cialdini, & Griskevicius, 2008): A rather unexpected form of perceived similarity is individuals following the norms of those who share or have shared an environment relevant to a particular behaviour (known as provincial norms). In a hotel study, Goldstein et al. (2008, as cited in Cialdini, 2012, pp. 308-309) tested which of three guest groups would show more towel reusage: one group received a standard environmental appeal, another group received a descriptive norm message (“the majority of other guests who previously stayed at this hotel reused their towels at least once during their respective stays”) and the third group received a provincial descriptive norm message (“75% of the guests who stayed in this room [room #xxx] participated in our new resource savings program by using their towels more than once” . Interestingly, the second group yielded a 18.3% higher towel re-use rate relative to the standard environmental appeal, while the third group produced the highest towel reuse rate (32.5%) relative to the standard environmental appeal.

Relating this back to our opening example, let’s assume that those Twitter users who were questioning how the police would keep track of people’s time spent outside may also be less likely to comply with the behavioural restrictions.

Some of them may assume that other members in their local environment would agree with the difficulty of adhering to social distancing and self-isolation. Thus, the messages from the PM, Schwarzenegger or local nurses may be less persuasive than if they were to observe neighbours complying with self-isolation.


In sum, we could all try applying some norm theory to support the enforcement of the nationwide lockdown, as follows:-

  • Investigate the important social groups/categories or values for people who are currently not adhering to the PM’s instructions and use those to convey the same message (social distancing, self-isolation and hand-washing).
  • Example for values: If someone expresses no care for the health of the elderly and more vulnerable members of the public, share norm messages with them that stress self-focused values instead - for example, that by staying at home and finding creative ways to stay entertained or exercise, they will foster their creativity abilities and resilience, skills that will benefit them even after crisis.
  • Example for social groups: If someone supports a particular football club, then check whether this club has made a public statement in line with the PM's instructions or, if not, maybe contact them and ask them to produce such a message. A more proximal social group constitutes one’s family - if a member does not adhere to the PM’s instructions, ask other family members and that person’s friends to share norm messages that stress compliance with social distancing, self-isolation and hand-washing.
  • If it is not possible to identify someone’s important social group/category or values, use provincial norm messages - share pictures of your street or house that show compliance with social distancing and self-isolation. Or share pictures of your local supermarket when you witness people expressing social distancing (after asking for consent to take a picture of them).

To further illustrate these points, here is a real-life example.

On the 24th of March, more than 20 people stood "shoulder to shoulder" for a barbecue in Coventry. “The crowd refused to disperse even when reminded about the need for social distancing, police said. Officers had to tip the barbecue over to put an end to the gathering.”

From a norm focus theory perspective, we can assume that:

  • Members of this BBQ party perceived that it was okay to do so, because they saw members of their meaningful social groups (e.g., friends and family members) rejecting social distancing.
  • The injunctive norm message not to stop the BBQ was expressed by a social-identity irrelevant group for the party guests (i.e., the police). Thus, they did not disperse.

Whose message would this group follow? Maybe that of a local celebrity or local pub? If so, an injunctive norm message from such a more meaningful person or group could be shared in this district to reduce the chance of further noncompliant behaviours in times of a nationwide lockdown.

25 March 2020


  • Cialdini, R. B. (2012). The focus theory of normative conduct. In P. A. M. Van Lange, A. W. Kruglanski, & E. T. Higgins (Eds.), Handbook of Theories of Social Psychology. Vol.2 (pp. 295–312). London: Sage.
  • Goldstein, N. J., Cialdini, R .B., & Griskevicius, V. A. (2008). Room with a viewpoint: Using normative appeals to motivate environmental conservation in a hotel setting. Journal of Consumer Research, 35, 472–482.
  • Wan C. (2015). Understanding cultural identification through intersubjective cultural representation. Journal of Cross-Cultural Psychology, 46, 1267-1272.
  • Wan C., Tam K., Chiu C. (2010). Intersubjective cultural representations predicting behaviour: The case of political culture and voting. Asian Journal of Social Psychology, 13, 260-273.


Sheila Kiggins

Media Relations Manager

07876 28166