Dr Sara Bru Garcia

Sarah Wong

Hi there! It’s Sara and Sarah here 👋 (we know…) We are both members of the Habit Weekly PRO team. Nice to meet you! 🙂

Welcome to this case study where you will go on a deep dive into behavioral change techniques with us!

We hope that by the end of this case study, you would have a good understanding of the do’s and don’ts around designing with social norms!

If you are new to behavioral design, we'll start with an introduction for you! If you are already familiar with social norms, feel free to skip through to the more advanced bits!

But first, let’s start with this recent summer of heat waves in Europe…

As we’re writing this, it was recently 40C in the UK. Clearly not what we are used to, and made worse by lack of air conditioners here…

So ‘fans’ have become a popular purchase in the past weeks.

There are so many fans on Amazon with similar ratings and prices, how do people even begin to decide? 😵

Personally, I’ve become increasingly aware of how easily influenced I am by labels like ‘Best Seller’ and ‘Amazon’s Choice’.

👉

👉

Hmm… others liked this, there must be a good reason… 🤔

A study on Amazon’s conversion has shown that these badges have significant impacts on viewership and conversion!

Customer reviews can also provide a powerful signal and can help us understand whether people liked the product after buying it.

These are all examples of ‘social proof’ we see in our daily life. In other words, using 'social norms' to influence our actions.

Social Proof

Social Norm

One type of conformity we just saw in the Amazon example is ‘informational conformity’

A preference to conform to the views or behaviors of others (descriptive norms), on the assumption that they have better knowledge on the situation or topic.

Social Norm

Social Norm

This tendency is useful especially when collecting this information could be hard, costly, time consuming, or if the consequence of making the wrong choice is minimal.

Social Norm

There is also another type of conformity - ‘normative conformity’ - which is our tendency to conform to the expectations of others (injunctive norms).

Social Norm

Norm conformity could be driven by many reasons...

For example, our need to feel accepted or liked, to feel consistent with our internal values, or to avoid social sanctions.

Reasons for compliance

Source: UNICEF’s guide - Everybody wants to belong

They help us navigate social situations in certain contexts or cultures. For example, being quiet in cinemas or libraries.

Reasons for compliance

Source: UNICEF’s guide - Everybody wants to belong

These norms encourage us to behave in a prosocial manner which may not necessarily align with our self-interest.

However, there is also a dark side to our tendency to conform to social norms.

✅

In this case study, we would like to dive deeper into how we can design with social norms to drive positive behavior change.

Remember when we talked about descriptive norms and injunctive norms?

These two concepts are the fundamental building blocks of social norms interventions.

Knowing the prevalence of certain behaviors can signal the ‘wisdom of the crowd’, for example, the benefits of positive behaviors or lack of consequences of negative behaviors.

Descriptive norms can be communicated in many ways. For example, using percentages e.g. ‘70% of people eat vegetables in every meal’...

👉

... or descriptors like ‘Most / The majority of people brush their teeth’, or absolute amounts like in this newsletter example.

Reference: Anik L, Norton MI, Ariely D. Contingent Match Incentives Increase Donations. Journal of Marketing Research. 2014;51(6):790-801. doi:10.1509/jmr.13.0432

Interestingly, messaging that are contingent / conditional on % of the population performing certain behaviours can also imply descriptive norm.

Reference: Anik L, Norton MI, Ariely D. Contingent Match Incentives Increase Donations. Journal of Marketing Research. 2014;51(6):790-801. doi:10.1509/jmr.13.0432

Injunctive norms messaging, on the other hand, signals what we should or should not do in order to be accepted or to avoid negative consequences.

This tends to be communicated through wordings such as ‘should, ought, approve, would, support, believe, expect’.

For example, ‘Most people expect others to queue’.

Or in more subtle messages like 'Well done for doing X', 'Thank you for doing Y', or 'Doing X will help achieve Y', etc.

Reference: De Groot JIM, Abrahamse W, Jones K. Persuasive Normative Messages: The Influence of Injunctive and Personal Norms on Using Free Plastic Bags. Sustainability. 2013; 5(5):1829-1844.

One study observed significantly fewer plastic bag usage in supermarkets when an injunctive norm message is displayed alongside supermarket’s original environmental messaging.

A third and important building block is the concept of reference network.

Reference: Noah J. Goldstein, Robert B. Cialdini, Vladas Griskevicius, A Room with a Viewpoint: Using Social Norms to Motivate Environmental Conservation in Hotels, Journal of Consumer Research, Volume 35, Issue 3, 1 October 2008, Pages 472–482.

A well know study (Goldstein et al.) showed that displaying descriptive normative messages in hotel rooms led to a higher rate of towel reuse compared to when standard environmental messages were displayed.

49.3% reused towel

👆

Whatsmore, when the social norm messaging is made as local and relevant to the target group as possible, towel reuse rate increased further.

This shows how we care about in-group norms in particular.

To design effective social norms interventions, it is critical to identify the relevant reference network to your target group.

The more similar we are (or feel we are) to the reference group, the more likely we are to adhere to the group’s social norms.

This could be personal similarities (e.g. gender, ethnicity, age, values, beliefs…) or situational similarities (e.g. being in the same place, like the same hotel room)...

...even when these similarities are artificially created by, for example, randomly assigning people to specific groups.

Furthermore, not everyone is equally important in a reference network.

For example, imagine a culture of working late in the office.

If your boss is staying late, this may have more influence on your behavior than if a peer is staying late...

Identifying who these key influencers are in a reference network could help you better target your interventions.

👈

Reference: Ellin Lede, Rose Meleady, Charles R. Seger,
Optimizing the influence of social norms interventions: Applying social identity insights to motivate residential water conservation, Journal of Environmental Psychology, Volume 62,
2019, Pages 105-114, ISSN 0272-4944

For example, by making in-group identity salient such as in this study by Ellin et al.

Identity priming

Identity priming

Reference: https://www.bi.team/wp-content/uploads/2016/10/Legacy-Giving-Report-final.pdf

Messenger effect

Role modelling

Knowing who the key influencers are in the reference network also allows us to identify suitable messengers
and role models in social norms interventions.

Messenger effect

Role modelling

Reference: https://www.bi.team/wp-content/uploads/2015/07/MINDSPACE.pdf

Messenger effect

Role modelling

Reference: Paluck EL. Reducing intergroup prejudice and conflict using the media: a field experiment in Rwanda. J Pers Soc Psychol. 2009;96(3):574-587. doi:10.1037/a0011989

Another example demonstrating the efficacy of role modelling is the use of radio soap operas to promote reconciliation in post-conflict Rwanda.

Messenger effect

Role modelling

Paluck observe that the soap opera - a romantic story of forbidden love - whilst did not change individual beliefs, successfully shifted individuals’ perception of (injunctive) social norms on issues like intergroup marriage, and led to behavior change such as tolerating different views in discussions.

Social comparison

Finally, part of why social norms interventions work is because they enable social comparison.

Social comparison

Social comparison

Reference: De Dominicis, S., Sokoloski, R., Jaeger, C.M. et al. Making the smart meter social promotes long-term energy conservation. Palgrave Commun 5, 51 (2019).

For example, using smart meters to display real time energy consumption data and framing feedback as social comparison is shown to be effective in driving long term energy conservation.

Social comparison

Strava is another good example where social comparison is rampant. If you are not familiar, Strava is an exercise tracker with social network features primarily for cyclists and runners .

Social comparison

🥶

💪🏼

My problem with Strava is everyone in my feed are marathon runners and long distance cyclists. This only induces anxiety and makes me hide my activities, forsaking the opportunity to get social support (‘kudos’) altogether. 🤦🏻‍♀️

Social comparison

🥶

💪🏼

One thing Strava could do - more proactively recommend users to join Clubs that are local and suitable to their levels. An experiment Strava could run is to test whether 'personal social network' or 'local runners of similar levels' is a more influential reference network to drive behaviour change.

✅

✅

Yay! We have now learnt 3 building blocks and 4 related tactics you could use in social norms intervention design! 🎉

Social comparison

Messenger effect

Role modelling

Identity priming

Let’s have a cup of tea and take a little break...

One question you may have on your mind is, what influences the degree to which people conform to social norms?

For example, what makes some more likely to wear masks during Covid than others?

Being aware of these factors allows you to more accurately diagnose the strength of existing norms you are tackling, as well as provide ideas on potential intervention tactics.

For example, building self-efficacy for people to overcome harmful norms

Such as teaching people who are trying to quit smoking how to navigate social contexts that encourage smoking.

✅

✅

👉

In the final part of this case study, let’s take a look at some common pitfalls when designing social norms interventions.

Despite 'social proof' being one of the most commonly used intervention technique for teams around the world.
This doesn’t necessarily always translate to greater efficacy.

Failure in identifying the right reference network

Messengers that are not trusted or credible

Failure in identifying the right reference network (⚠️Pitfall #1), and having messengers that are not trusted or credible (⚠️Pitfall #2) are two common reasons why interventions could be less effective.

Besides these, we must also be mindful of the following pitfalls...

Boomerang effect

In a study (Cialdini et al.), visitors to Arizona's Petrified Forest National Park were exposed to messages that discourage the theft of petrified wood. When a descriptive norm message is shown, theft increased in the park! (⚠️Pitfall #3)

Many visitors have removed petrified wood from the park, destroying the natural state of the Petrified Forest.

👉

Boomerang effect

Showing an injunctive norm message mitigated this boomerang effect, and led to an overall decrease in theft behaviours.

👉

Boomerang effect

Negative framing (what should not be done) is also found to be more effective than positive framings (what should be done). One theory explaining this is that negative framing attracts more cognitive attention as people seek information that help them avoid social disapproval.

Effect of less observable norms decay over time

Reference: Ayres, I., Raseman, S., & Shih, A. (2013). Evidence from Two Large Field Experiments that Peer Comparison Feedback Can Reduce Residential Energy Usage. Journal of Law, Economics, and Organization 29 (5): 992–1022.

⚠️ Another pitfall (#4) was observed in an energy conservation study. When households receive letters telling them their energy consumption with norms messaging, a reduction in energy consumption is observed. However, this effect decayed over time.

Effect of less observable norms decay over time

👉

Upon receipt of reminder letters, positive effect is observed again, suggesting repeated reinforcement is necessary until habits are formed to sustain behavioral change.

Effect of less observable norms decay over time

👉

In this case, intervention effects persisted after households received one year of monthly energy consumption reports with social norm messaging.

⚠️ Pitfall #5 - being in the majority is not always possible.

Reference: David D. Loschelder, Henrik Siepelmeyer, Daniel Fischer, Julian A. Rubel, Dynamic norms drive sustainable consumption: Norm-based nudging helps café customers to avoid disposable to-go-cups, Journal of Economic Psychology,
Volume 75, Part A, 2019, 102146, ISSN 0167-4870.

For example, a study by Loschelder et al. aimed to increase the use of reusable cups in a university cafe by placing signs by the coffee machines.

Minority norms

👉

Being a minority norm, signs using dynamic norms messaging is observed to have the strongest effect compared to injunctive or descriptive norm messaging.

Minority norms

👉

⚠️ Pitfall #6 - when there are inconsistencies between descriptive and injunctive norms, people tend to put more weight on observable descriptive norms.

For example, observing lots of littering on the street suggests a social norm behavior, even if there are signs saying ‘Most people disapprove of littering’.

People reject information inconsistent with their beliefs or observed norms

👉

To mitigate this, practitioners should minimise contextual cues suggesting negative norms, and aim to shift descriptive norms first.

The need to coordinate norm shift is particularly important when it comes to influencing people to abandon harmful norms. For example, child marriage in some countries. 

Overlooking necessary societal-wide norm shift

Overlooking necessary societal-wide norm shift

This brings us to the last pitfall we will discuss in this case study -  overlooking the need for collective/ systemic change. (⚠️ #7)

Overlooking necessary societal-wide norm shift

Overlooking necessary societal-wide norm shift

In the case of child marriage, changing individual beliefs alone would not be sufficient to shift behaviour if the behavior is prevalent, expected, or if there would be negative sanctions from the community if individuals do not conform.

Overlooking necessary societal-wide norm shift

Overlooking necessary societal-wide norm shift

Overlooking necessary societal-wide norm shift

Even if it is made known that most disapprove of the behavior, people need to trust the source of the information and be sure that there would be no negative sanctions for deviating.

Overlooking necessary societal-wide norm shift

Overlooking necessary societal-wide norm shift

👉

In this case, a coordinated community-level shift may be necessary to remove the threat of sanctions and to create new positive norms.

Overlooking necessary societal-wide norm shift

👉

If you would like to learn more, we highly recommend Cristina Bicchieri’s lectures, books, and her work with UNICEF where this topic is covered in depth.

So, that’s all for this case study! We hope you have learnt something useful!

✅

✅

✅

Here is a summary of what we have covered!

✅

✅

✅

You can also download a cheat sheet here for future references 🙂

What did you think about this case study? Let us know so we could improve!

Last but not least, here is a list of the key references we used for this case study if you want to read more on this topic! Enjoy! 👋

References

• Nolan, J., & Wallen, K. (2021). Social Norms and Persuasion. In B. Van Rooij & D. Sokol (Eds.), The Cambridge Handbook of Compliance (Cambridge Law Handbooks, pp. 404-421). Cambridge: Cambridge University Press. doi:10.1017/9781108759458.028
• UNICEF (2019). Everybody wants to belong: a practical guide to tackling and leveraging social norms in behavior change programming.
• Cristina Bicchieri (2016). Norms in the Wild. How to Diagnose, Measure, and Change Social Norms. Oxford: Oxford University Press, 264 pages, ISBN 978-0190622053.
• Institute for Government (2015). MINDSPACE: Influencing behaviour through public policy.
• Academic papers and other references are cited throughout this case study