Research Q&A: Behavioural science and teacher recruitment

Can simple prompts help encourage more trainee teachers to apply for placements in rural and remote schools? In today's Q&A we speak to Dr Karen Tindall, a Senior Adviser at the Behavioural Insights Team Australia about its work with the New South Wales government, the state's education department, and local universities.

The Behavioural Insights Team (BIT) – nicknamed ‘the nudge unit – started in 2010 as part of the UK Government's Cabinet Office. It's now a ‘social purpose consultancy' partly owned by the UK government, employees and charity Nesta. To date, it has run more than 800 projects in dozens of countries, including 400 randomised control trials. The Australia office in Sydney was set up in 2015.

What was the initial aim of the BIT?

It started as a seven-person team … tasked with applying insights from the behavioural sciences to inform policy, improve public services and deliver results for citizens and society. However, the team also faced a ‘sunset clause' – they would be shut down at the two-year mark if they had not ‘transformed two major areas of policy' and ‘achieved at least a ten-fold return on the cost of the team'. By using robust evaluation (randomised controlled trials) the team was able to show clearly how a behaviourally-informed approach could make government policies more effective and efficient, and the value of the approach is now recognised globally.

What's your role and background?

Before joining BIT I was a postdoctoral researcher, looking at how social psychology theory was being applied in public policy. In my time with BIT I've worked across a number of social policy projects, aimed at improving health outcomes, increasing charitable giving, reducing domestic violence reoffending, as well as a number of projects aimed at improving education outcomes in Australia.

What sectors do you work with in addition to education?

We set up the Behavioural Insights Team Australia office in Sydney in 2015. We work in close partnership with federal, state and local government, businesses and charities to design and improve policies, programs, and services using more realistic models of human behaviour.

Our team in Australia works across the scope of social policy areas on a wide variety of challenges. For instance, we have worked with partners to encourage flexible working and more physical activity in workplaces; to improve sustainability by reducing the use of single-use plastics and reduce energy use on peak-demand days; to reduce sexual harassment on university campuses, and reduce domestic violence reoffending.

What is the theory behind the work you're doing? Is it ‘nudge' theory?

Behavioural insights is the practical application of behavioural science. This includes research in fields such as behavioural economics, cognitive psychology, social psychology, and anthropology. Behavioural science is the study of what makes people take action and why. It provides a realistic model of human behaviour. We like to think we are rational beings and our decisions are made through deliberative analysis of costs and benefits. But that is the case only some of the time. Fifty years of behavioural science research has shown that when we are making all sorts of decisions and judgements our brain uses rules of thumb and shortcuts, and is subject to biases.

The idea of a ‘nudge' was made famous by Thaler and Sunstein in their 2008 book of the same name. Thaler and Sunstein said that a nudge ‘alters people's behavior in a predictable way without forbidding any options or significantly changing their economic incentives. To count as a mere nudge, the intervention must be easy and cheap to avoid'.

However, despite being nicknamed ‘the nudge unit', the work of the Behavioural Insights Team is less concerned about whether or not we are ‘nudging' and more concerned with identifying empirical findings about human behaviour that can be used to make public policy more effective. Our focus is on changing behaviour rather than changing attitudes or beliefs, on findings that are tested and shown to be effective, and on insights that can be used to make traditional policy levers (laws and regulations, taxes and subsidies, and providing information) more effective.

Is this all about very small changes, rather than something huge and complex? Assuming that's the case, why is this important?

When the team started back in the UK the initial focus was demonstrating how small changes could have a big impact. For example, the team showed that simply adding a line to an existing reminder letter could bring forward millions of pounds in tax revenue, or that reminder SMS messages could increase attendance at court hearings and hospital appointments.

Even as we moved into more challenging policy problems we have found that an intervention doesn't need to be huge and complex to have a big impact. Even small barriers which require effort to overcome can prevent people from doing something. And even small changes to something as simple as letters and forms has had a huge effect on the lives of individuals.

For example, even when financial assistance is offered to disadvantaged young people they don't necessarily apply for it. So, a group of US academics (Bettinger, Long, Oreopoulos, & Sanbonmatsu, 2012) found a novel way to change this. With one group of young people they provided information about the financial assistance and found it didn't make any difference at all. With another group they offered to prefill the information in the forms using tax return data that was already held by the government. This automated pre-filling of forms led to a significant increase (8 percentage points) in the number of young people who actually went to university – not just an increase in completing the form, but in actually attending university. Something that is potentially life changing.

Can you tell me about BIT Australia's work with the rural teacher recruitment trial in New South Wales?

It is a constant challenge to attract teachers to rural and remote Australia. However, teachers are more likely to work in rural schools if they have already completed a rural placement during their teacher training. Alongside the New South Wales Department of Customer Service's Behavioural Insights Unit, NSW Department of Education, and three universities, we ran a set of trials to encourage trainee teachers to apply for teaching placements in rural New South Wales.

In our trial with University of Wollongong in 2017 to overcome some of the ‘friction' in the application process, we made it easier to apply for rural placements by replacing the paper application with a partially pre-filled online form and increasing the number of rural schools that could be selected for students' placements. We also tested a series of prompts, including: extra information about rural schools; encouragement to discuss the idea with family and friends; and a timely reminder before applications closed.

In the intervention group – which received this set of personalised and timely prompts – the proportion of trainee teachers applying for a rural teaching placement tripled. In this group, 12.6 per cent of trainee teachers applied for rural placements, compared to the business as usual group (which also had access to the online application process but were not sent the timely prompts), where just 4.2 per cent applied.

The online application process is now the standard and the School of Education at the University of Wollongong is using the lessons learned to trial new communication tools and framing.

Is this just a case of good marketing, or just reducing the barriers and making things easier for people, rather than influencing behaviour?

Drawing on cognitive and social psychology, we can see that even good marketing is often not enough to change behaviour. How choices are framed matters, but providing information or changing attitudes is often not sufficient to change behaviour.

A lot of services, policies and programs are automatically designed with an assumption that those using the service will act as rational decision makers – that people weigh up the costs and benefits of an action and make the best choice for their current and future self. It is quite easy to think of examples of how human behaviour does not reflect this rational model. How information and choices are presented and framed to us has a disproportionate influence over our response to it.

Just being convinced of the value of exercise isn't enough to get me to the gym. Even if I have the best intentions to get a health screening, it doesn't mean it will make it to the top of today's to do list. To change behaviour we need to work with the more realistic model of human behaviour. This means taking into account and working with the many cognitive heuristics, mental shortcuts and rules of thumb we are subject to.

For example, there is a huge problem of heuristics and bias in job recruitment. Several studies have shown how hiring managers are affected by the names of applicants. Other studies have shown that we are affected to a much greater extent than we realise by the order in which we see information.

This led our team in the UK to develop an online recruitment platform to eliminate bias from hiring. This platform, Applied, has now been used by over 70 000 job applicants. People sifting applications never see a candidate's name or where they went to school. Each application is chunked up into separate work-related tasks, each of which is independently reviewed by those carrying out the sift. And the order of these reviews is randomised so as to remove any ordering effects.

Beyond savvy marketing and making things easier, we can draw on the wealth of knowledge across policy domains to develop behavioural interventions to support and help people. If you think about the challenge for students, many learners may feel uncomfortable in the learning environment due to their previous educational experiences. We found that a short but well-designed reflective writing exercise can help. An exercise where vocational students reflected on and wrote about their personal values improved achievement by 25 per cent (4.2 percentage points, from 16.7 to 20.9 per cent). By creating a sense of belonging in the classroom we were able to have a meaningful impact on educational attainment.

What about the criticism that people might feel duped or manipulated into doing something?

True of any policy approach, ideas from the behavioural sciences produce manipulative policies if they are used manipulatively. A key concept discussed by Thaler and Sunstein is ‘libertarian paternalism' (the original title for the book Nudge) – in which freedom of choice should be maintained while making it easier for individuals to succeed.

Any policy, program or service that seeks to tackle a particular policy issue should do so as effectively as possible. If there is new available evidence from the behavioural sciences that will make the policy more efficient and effective it should be applied so the government is spending its limited resources wisely.

It is possible to maintain freedom of choice and still apply behavioural insights. For example, we can have a meaningful impact on lives by making the most of existing support networks that may have otherwise gone untapped. In the UK, 16 to 19-year-olds who fail their GCSE grades in Maths or English are required to re-sit these exams. Despite this, only 27–30 per cent of those resitting walk away with a pass mark. A barrier to learning that was identified by students was social isolation – a feeling of not knowing who to turn to for help.

In partnership with the UK Department for Education and Professor Todd Rogers at Harvard University, we ran a trial in which students were asked to pick ‘study supporters' from their family, friends and employers. This study supporter received regular text messages throughout the year. These messages included key events, content updates, and conversation prompts. This simple SMS intervention had a big impact: students receiving the intervention were 27 per cent more likely to pass their exams.

By tapping into these existing social connections to foster education outside of the classroom, we were able to help students engage, and boost both attendance and results. We are currently trailing a similar intervention here in Australia and expect to have results early next year.

What else do you have in the pipeline here in Australia and internationally in the K-12 space?

We have a whole range of educational interventions in the works. Globally we will be testing a preschool curriculum for parents, delivered via SMS, to promote early reading, maths and socioemotional skills. We are testing software to diagnose the root causes of students' maths mistakes and help teachers tailor support. We are testing interventions to boost the perceived value of STEM (science, technology, engineering and mathematics) subjects and tackle barriers that young women may encounter in taking these subjects.

There is also a program of work we are really excited about in Australia. The question of how to foster the ethical development of young people has taken on fresh urgency with recent, rapid changes in technology. Young people are finding new ways to socialise, learn, and express themselves, and encountering new ethical challenges in the process. The response of many organisations, schools and governments have, to date, focused on minimising risks, rather than empowering young people as moral agents. Over the past two years, we have been designing, trialling and refining a series of interventions to encourage the ethical development of young people online.

References

Bettinger, E. P., Long, B. T., Oreopoulos, P., & Sanbonmatsu, L. (2012). The role of application assistance and information in college decisions: Results from the H&R Block FAFSA experiment. The Quarterly Journal of Economics, 127(3), 1205-1242.

Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improving decisions about health, wealth, and happiness. New Haven, CT, US: Yale University Press.