Systems 1 and 2: turning theory into practice
I've recently experienced what it feels like to rely on System 2 thinking. In a half-hearted attempt to learn the ukulele over the last year, I've faced the mental challenge of having to think very deliberately about using my hands in a new, unfamiliar way. Although I'm unlikely to master playing it any time soon, I've made some progress.
As Ive acquired basic skills, Ive found myself thinking less about what Im doing as I play, or — when things are going particularly well — barely think at all. Ive also become aware that, even if I think that I have learnt what to do with my hands, if the task and context changes then my previous ease and fluency disappears. Instead, I feel clumsy and slow. I have to re-learn those skills. This is what System 2 thinking feels like.
By comparison, when riding my bicycle to work — a familiar activity — I coordinate cycling, navigation, soak in the environment and watch out for potential threats on the road effortlessly, all while my mind wanders, thinking about the day ahead. This is what it feels like when System 1 is in charge: effortless, fluent and possible to multitask.
This two-system model is a useful way of framing and understanding decision making. In particular, it draws our attention to the fact that a lot of decision making and influence happen at an instinctive and automatic level (System 1). Indeed, most of the time our brain is multi-tasking: drawing on multiple reference points and learned thinking patterns to solve the problems it is presented with.
This model is also increasingly being used as a basis for designing real-world interventions, such as social norms on tax letters or presenting options for a subscription package to encourage the desired choice. Theres a dual benefit to leveraging System 1 thinking. Choices made automatically and instinctively are not only less effortful, they are also more likely to be habitual.
But from a practitioner perspective it can sometimes be tricky to know how to apply this framework to our work.
You cant see it
Its possible to measure the effects of the different types of thinking in the lab or to see System 2 in action on clear-cut mentally taxing tasks, like calculating 17 x 24 or learning the ukulele, but in the world of qualitative research its much harder. Thats because you cant see it as a researcher, and it is beneath conscious awareness for participants.
Its not as clear cut as the names suggest
The model makes the distinction sound very black and white — as if these are two distinct brain processes. This is theoretically alluring but this clarity can break down. We dont just use System 1 or System 2 to make decisions, we use both, often simultaneously.
The mental agility with which we wield these deliberate and automatic systems is more dynamic, a constant ebb and flow. For example, while out shopping the allure of a BOGOF offer on fruit might pull at our System 1, and we may fall for it in some instances, but equally System 2 might step in with a rational consideration of whether we really need the item, or if it really is a good deal.
We are master storytellers
Because our brains are powerful and complex, what we see and know of them from our conscious perspective is only the tip of the iceberg: a great deal of our mental processing goes on beneath. One of the brains unique features, however, is its innate desire to make sense of the world: to see cause and effect, even when it is not there.
This sensemaking instinct allows us to provide plausible stories to explain our choices, but these are often overly simplistic, and overlook/omit key decisionmaking influences. This means that we cant rely on peoples self-reported explanations for a complete picture of their decisions, habits or actions.
So how do we turn theory into practice?
System 2 is conscious, which makes it relatively easy to extract insight using traditional research methods. As we know, though, it only accounts for a portion of our decision making. System 1, in contrast, is elusive and hard to research. But there are things you can do as a qualitative researcher to explore the role of System 1.
Start with a hypothesis
One beauty of the explosion of interest in behavioural science over the last few decades is that there really aren't many topics that havent been put under a behavioural science lens. By starting out every project with some desk research, you can identify potential System 1 influences: these can be used to form hypotheses, feed into discussion guides, and what to look out for, etc.
If we were to look at littering behaviour, say, a wealth of studies exists around the role of personal, social and contextual influences on our decision making. A good example of a social/environmental factor is littering on the Tube. We know that littering per se is widely disapproved of: more than eight in ten people express anger and frustration about the issue (Populus, 2015).
A review of littering psychology, however, shows a difference between ought to social norms, e.g. littering is bad, and whats done around here social norms, e.g. leaving my old paper on the newspaper for others to read is the norm. So, our System 1 choices around littering behaviour will be strongly influenced by the behaviours seen in specific contexts rather than simply the more conscious, general littering attitudes. This means that we should research the context and behaviour of specific types of littering and avoid researching general littering or ought to attitudes.
Research in context
Our brain offloads heavily onto the physical and social environment during decision making and is also heavily swayed by emotions and social influences in the moment. This means that our research will go deeper, and we will gather more insight by being in context and ideally in the moment.
On a project exploring risky sexual health behaviours of young people abroad, for example, we went to Magaluf to conduct observational research and in-the-moment interviews. We observed first hand how in the evenings the influence of the emotionally-charged mood, the effect of alcohol consumption and the norms of behaviour changed. The morning-after was quite different. After the event, different stories were told by the young men we met — who were all having unprotected sex — and the young women we met — none of whom were having risky sex.
This understanding of impulsive and instinctive decision making in the environment, as well as the postrationalisations of behaviour, helped the client design solutions that worked in-themoment to nudge System 1 thinking. This approach was also taken in the Design Councils work on violence and aggression in A&E where extensive observation and understanding of the behaviours in the moment exposed a very different set of perpetrators and triggers to the ones that had been expected.
Observe first, ask later
People are innately programmed to look for causes of events/cause and effect in the world. What we perceive, however, isnt always accurate. Life as a researcher would be much easier if you could ask people for an explanation of their behaviour and they were able to provide the right answer. But people dont always know what causes them to think/act in the way they do. Rather than asking people to do this, focus your research on taking an observational stance:
- What is the behaviour/what are people doing?
- In what context?
- What are the possible influences on their behaviour?
This observational approach can take a little getting used to: it feels natural to just ask, yet observing delivers really important findings. Researching recycling behaviour, for example, results in people recounting their recycling habits and how they recycle everything. But a quick tour around many participants homes showed their recycling behaviour was limited to the kitchen, while elsewhere home recyclables and nonrecyclables were muddled together. As researchers we need to make sure we are observing and looking for evidence that may confirm or disconfirm peoples personal accounts. When you do ask questions, ask people about the last time they did something rather than about their attitudes or opinions in an effort to focus on behaviours.
Spot the rules
A lot of our System 1 and even System 2 thinking is rule based; we use decision rules/heuristics to help simplify our choices. For example, brands are often used as a shortcut to quality. Rules are very important for our brains: they allow us to make choices quickly and easily. By using your initial desk research and real-life observation to identify the rules consumers are using, you can help clients align their behaviour or brand in a way that exploits the rule.
In the previous recycling instance, for example, there was one rule governing recycling behaviour in the kitchen (where litter should be sorted by category), but a different one relating to elsewhere in the home (dictating that all types of litter should go in the same bin). We are following a rule habitually in each instance, but the context and design of our home environment shapes our behaviour differently in one room to another. Comparing behaviours in different contexts and environments can often be a useful way of bringing these rules to the surface.
Its all relative
A lot of our System 1 thinking is done by comparison: it is quick to calculate the relative advantage of one choice over another. In our research its important to look at choices against comparison points rather than getting drawn into focusing on one particular product or behaviour.
Take a visit to the supermarket, where we feel that around 3 pence per tea bag is a reasonable price, whereas in Starbucks well happily pay £1.80 for a tea. We dont compare these costs, we consider them within their choice set: tea bags in the supermarket against other tea bags, while considering the price of a cup of tea in Starbucks within the context of other drinks.
When we focus in on the choice and its relative context we can start to look at the rules and hypothesise how a new competitor may affect this.
The relatively low cost of Uber, for example, has tipped it closer to the cost ofpublic transport than mini-cabs and taxis were previously. This, combined with an incredibly easy booking and paying mechanism, means that Uber is more frequently considered an everyday choice as part of the transport mix.
Disrupt System 1
System 1 is a set of habits and instincts we have learned over time to make decision making quicker and easier. The challenge for us as researchers is that, because these are so entrenched, we no longer rely on conscious thinking or our working memory to enact them. While they are hard to research directly we can learn a bit about them by disruption. By asking participants to document their habits for a few days (e.g. their travel choices), followed by a period of documenting a change (e.g. depriving themselves of their preferred mode of transport), we force ourselves into System 2 thinking. In this mode, we become more aware of the role of System 1 and the load it was bearing. This was very much the case in my ukulele experience: when I stepped into a new behaviour, I became more aware of the learning process my brain was going through.
What at first glance looks like a simple and easy way to dissect decision making, in practice starts to look a lot more complex and tricky to apply and unpick. The more you learn about Systems 1 and 2, the less surprising that becomes: beneath the alluring simplicity of the two-system model sits a web of complex and often conflicting psychological influences. At the last count the list of named biases and heuristics was well over a hundred and fifty!
As researchers theres a balance to be struck between the simplicity of considering the logical and conscious (System 2) vs automatic (System 1) and the complexity and nuance of more specific influences and effects. Often, once we start our research, we uncover a lot more nuance than can be accounted for by the simplified Systems 1 and 2 model. What it does offer, however, is an everyday reminder to look beyond the influences that participants can tell us about and dig deeper into the hidden forces that shape our decisions.
Copyright © Association for Qualitative Research, 2018