Rebel Ideas- the Power of Diverse Thinking

by Matthew Syed

  When it comes to personalised nutrition, there is a long way to go. More studies are needed with longer-term follow-ups that measure health outcomes directly rather than via indirect indicators, such as blood sugar. Further research is required to understand the microbiome and more. The start, however, has been highly promising, and gives researchers a chance to overcome the contradictions that have bedevilled the field. Above all, it articulates a vital truth that science itself can dispose us to forget – diversity matters.

  IV

  In the spring of 2010, Michael Housman, a labour economist, was working on a project to figure out why some call centre workers perform better than others. No matter how hard he looked, he couldn’t find an answer. Nothing seemed to compute. He told me:

  I was working as Chief Analytics Officer for a firm that sells software to employers to help them recruit and retain staff. We had data on 50,000 people who had taken a 45-minute online job assessment and who were subsequently hired. We examined every aspect of the assessment to see if it held clues about longevity and performance. But we kept drawing a blank.

  Housman’s team had anticipated that those with a history of jumping around employers might, on average, leave quicker. They didn’t. Staff could have had five jobs in previous years, or just one, but it didn’t predict longevity at all. The team thought that certain aspects of personality revealed by the assessment might correlate with performance, too. That didn’t stack up either.

  But then one of Housman’s research assistants had a flash of insight. The team had data on the web browsers that had been used by the applicants to fill out the assessment forms. Some of the candidates had used Safari, others had used Firefox; some had used Internet Explorer, others Chrome. Might the choice of web browser predict performance? To Housman, it seemed unlikely. Surely, this was just a matter of personal preference.

  Yet the results were startling. Those who had filled out their assessments on Firefox or Chrome stayed in their jobs 15 per cent longer than those who used Safari or Internet Explorer. They then checked the number of absences from work. Again, they found the same gap. Those who used Firefox or Chrome had 19 per cent fewer absences from work than those who had used Internet Explorer or Safari.

  If this wasn’t puzzling enough, the numbers related to performance were even more striking. Those who used Firefox and Chrome had higher productivity, higher sales, happier customers and shorter call times. ‘It was one of the most emphatic sets of results we had found,’ Housman said. ‘These were big differences, and they were consistent.’

  What was going on? Housman said:

  It took us a while to figure it out. The key is that Internet Explorer and Safari are pre-installed. PCs come with Explorer as part of the package, and Macs come with Safari. These are the defaults. To use them, you just need to turn on the computer. Chrome and Firefox are different. To use these pieces of software, you have to be curious enough to check if there are better options out there. Then, you have to download and install them.

  It wasn’t the software itself that was driving these differences in performance, it was what the choices revealed about differences in psychology. Some people have a tendency to accept the world as it is. They stick with the status quo. Others see the world as changeable. They wonder if there are better ways of doing things and, if so, act upon them. A seemingly inconsequential decision on which web browser to use revealed different positions on a psychological spectrum. Translated into the jobs they were doing, this meant many things. Remember, these were professionals working in call centres in retail and hospitality. Such jobs often have a set of scripts that are used to deal with consumer enquiries. It is easy to stick to a script. It represents the default. But every now and again, you meet a situation that isn’t covered by the script, or where a fresh approach might work better. Do you just stick to what you have always done? Or do you find a new way of solving a problem, or selling an idea, or pleasing the customer?

  Those call centre workers who could step outside convention performed significantly better. When the status quo wasn’t good enough, they came up with something original. This mindset also helps to explain why users of Chrome and Firefox stayed in their jobs longer and had fewer absences. Workers capable of altering the script are more likely to take action to fix problems, and make changes to their jobs that make them happier and more productive. Those who see the status quo as immutable are less likely to fix problems at work. They just put up with the default. Until they quit in frustration. ‘We were initially shocked by the size of the results,’ Housman said. ‘But we came to realise that the web browser decision shone a light on a crucial trait. The ability to question defaults makes a huge difference in a changing world.’13

  The Housman experiment is rightly taken as evidence for the power of mental agility. These were people with an outsider mindset, capable of stepping outside the paradigm. This made them more productive and more fulfilled. They could solve problems, rather than simply endure them.

  But there is another lesson, too, which relates to this chapter. Think of the concept of best practice. This is one of the most familiar ideas in business and hinges on a simple assertion: if there is a way of doing things that is proven to be superior, it makes sense for everyone to adopt it. In healthcare, for example, studies show that doctors often perform procedures in different ways, with the result that patients are not getting the best treatment. Best practice is often crucial. Another aspect of best practice is also well understood. It is ‘best’ not in an absolute sense, but in a comparative sense. It is the ‘best so far’. If one can show that there is an alternative way of doing things that is superior to the status quo, then best practice should be amended. In this conception, best practice evolves through time, in a rational, data-led way.

  But we should now see that this analysis, while useful, is also incomplete. Why? We can go back to the research of Eran Segal to expose the flaw. Suppose you ranked diets according to the blood sugar response for a population. You could conduct the study with rigour, and determine that one diet was superior to all possible alternatives. And yet this wouldn’t be the optimum diet. It would merely be the best standardised diet. A different approach would be to adopt personalisation. This doesn’t consist of comparing different diets across the whole population, but adopts flexibility at the level of the individual. And, as Segal showed, this approach can lower the blood sugar response, leaving an individual with a significantly healthier diet.

  Now think of the call centre workers. Many organisations test different scripts, compare the results, perform statistical tests, and come to what seems like a scientific conclusion about which script is the best overall. But this often misses the benefits of flexibility. The Housman experiment shows that when workers sensibly deviate from a script, they often perform better. This is partly because they are adapting to a new situation, as noted above. But it is also because it gives them a chance to play to their strengths, to bring their personalities into the conversation. The script is varying according to the individuality of the worker. To put it another way, best practice cannot be established by comparing standardised solutions; it also requires the comparison of different kinds of flexibility. And, given what we have learned about diversity, it is the more flexible systems that often win, whether one is talking about cockpits, diets, scripts or anything else.

  The question of how to build more flexibility into the world of work is now a major topic. It often centres around the scope to work from home, holidays, hours and the like. Work is often more rewarding when professionals have the latitude to tailor schedules to personal commitments, and offering such flexibility means that organisations can access new talent (such as people who might not wish to work a standard 9 a.m.–5 p.m.). This is particularly significant for Generation Y. According to one study, work–life balance was the single most important factor for young people in choosing an employer.

  And yet this represents only one facet of the power of flexibility.
The deeper significance is expressed in diversity science, and has not yet scratched the surface of the way institutions and societies function. This will prove to be a key source of innovation in the design of systems, allowing individuals to play to their strengths, and to bring their distinctiveness into the workplace.

  Flexibility has dangers, of course. When we have the latitude to make changes, we also have the freedom to get things wrong. There is always a balance. But while institutions often focus on control and error, we rarely acknowledge the dangers of rigidity that often lurk, undetected, at the heart of modern systems. In short, we need to become altogether more scientific about diversity.

  V

  Standardisation is baked into our lives. Education reformers in the early twentieth century designed schools with ‘standardised curricula, standardised textbooks, standardised grades, standardised holidays, and standardised diplomas’.14 The idea was that education should not flex to the needs of individual learners, but that individuals should be crowbarred into the needs of the system.

  The paradigm was mass-production: schools would churn out students the way factories churn out widgets. They should be taught in the same way, at the same pace, with the same tools, the same textbooks, measured with the same tests. As Ellwood Cubberley wrote in his influential 1916 guide: ‘Our schools are, in a sense, factories, in which the raw products (children) are to be shaped and fashioned into products to meet the various demands of life.’15

  This approach had advantages over the disjointed system it preceded, but it also had limitations. After all, if youngsters are different from each other in important ways, flexibility should be designed-in at the level of the student (a point that wise teachers have always known). Indeed, there is good evidence to suggest that flexibility offers better results for students and schools. The 2015 Pisa Tables showed that ‘adaptive instruction’ was the second most powerful predictor of high levels of educational outcome, rating above discipline, classroom size and more. (The only thing more powerfully correlated with performance was wealth.) Adaptive instruction is what you might expect – teachers who adapt to the needs of individual students, rather than getting everyone to do the same thing, at the same pace, at the same time.

  A recent article by the author and teacher Maria Muuri summarises the key tenets of Finnish education, often considered to be among the best in the world. Five of these factors dovetail with the themes of this book, including transversal skills, which seeks to equip children with flexible thinking, and multidisciplinary learning, which helps students to see how subjects are not separate silos but domains that can be bridged to forge new insights (recombination).16 Muuri also writes about why it is important for the system itself to be flexible. ‘Students are all individuals, so we can’t teach them all in the same way,’ she writes. ‘[In Finnish schools], there are usually at least five different levels of assignment in the same class at the same time. It also means that every student has their own specific goals.’ This is called differentiation.

  Another key factor is diversity in students’ assessment. She writes:

  The new Finnish curriculum emphasizes diversity in assessment methods as well as assessment that guides and promotes learning. Information on each student’s academic progress must be given to the student and guardians on a sufficiently frequent basis . . . We set goals and discuss the learning process, and the evaluation is always based on the students’ strengths.

  Muuri also, rather wonderfully, writes about the way that students can benefit from cognitive diversity within study groups: ‘We make a point of having students from different backgrounds work together. I believe that there’s always something that you can learn from someone who is different than you.’

  Some argue that there are parts of the educational system where personalisation has gone too far. Others, such as Todd Rose of Harvard, want it to go much further. This is a healthy debate, which should be guided by evidence. What is universally agreed is that providing for flexibility in otherwise rigid systems can help all students to flourish.

  Of course, the tyranny of average is about far more than education; it has infiltrated science more generally. A classic fallacy is to take averages based on male subjects and assume they apply to women, too. Think back to the cockpits. If these were ill-designed for different sizes of men, think how much worse they would be for women, who are, on average, smaller. In her book Invisible Women, Caroline Criado Perez points out that piano keys were designed for the hand size of the average man, along with such things as police body armour and military equipment.17

  Yet these physical design flaws are really a metaphor for a broader array of institutional arrangements designed for the average male, but which invisibly make things harder for women. As Perez says: ‘one of the most important things to say about the gender data gap is that it is not generally malicious, or even deliberate. Quite the opposite. It is simply the product of a way of thinking that has been around for millennia and is therefore a kind of not thinking.’

  These conceptual confusions occur with the so-called ‘hard’ sciences, too. A few years ago, Michael Miller, a neuroscientist at the University of California in Santa Barbara, conducted an experiment with sixteen subjects, who were placed in an FMRI machine. They were presented with a list of words, given a break, then presented with a list of new words. They were then asked to press a button every time they saw a word that had been on the original list.18 The brain scans of the subjects were then analysed. The objective was to determine the neural circuits implicated in verbal memory. These are typically presented in a brain map, showing which part of the brain lights up, and will be familiar to anyone who has read a paper on neuroscience. What is perhaps less appreciated is that this map is calculated by averaging across subjects.

  For some reason, Miller decided to look not at the average response but at the maps that detailed the individual responses. ‘It was pretty startling,’ he said in an interview with Todd Rose. ‘Most didn’t look like the average map at all . . . What was most surprising was that these differences in patterns were not subtle; they were extensive.’19

  Think about that for a moment. Neuroscience, one of the most exciting branches of modern research, can provide misleading conclusions because the average brain map conceals the diversity of individual responses. Rose comments: ‘The extensive differences that Miller found in people’s brains aren’t limited to verbal memory. They’ve also been found in studies of everything from face perception and mental imagery to procedural learning and emotion.’ None of this means that neuroscience is flawed. Sometimes using averages makes sense. All too often, however, scientists use averages while scarcely conscious of doing so.FN1 It is one step away from treating people not as diverse individuals, but as clones.

  *

  One of my favourite experiments in diversity science was conducted by Craig Knight, a psychologist at Exeter University. Before becoming an academic, Knight was a salesman, travelling up and down the country. It was while standing in a large office in the West Midlands that he was struck by the dangers of standardisation. He was looking down a vast row of melamine desks, all identical, stretching into the distance. The idea in vogue at the time was that all work areas should look alike, and that workers should operate in standardised spaces. For Knight, it felt more than a little depressing. He tells me:

  It was called the lean office concept and it was all the rage at the turn of the millennium. The idea was that there should be no personal items. No photos. No paintings or plants. Such things were considered a distraction. If it could be scientifically proved that a particular type of work area was the most efficient, managers believed that everyone should conform to that.20

  As a travelling salesman, Knight noticed the lean concept in office after office, managers looking proudly at lines of standard spaces and workers beavering away in structured uniformity. Managers thought that they had hit upon an empirical way of boosting productivity. Knight, who had a background in p
sychology, wasn’t so sure:

  My hunch was that this had unintended consequences. If you put a gorilla or a lion in a lean enclosure, they are really miserable beasts. They get stressed, they fight, they become impotent, they die early. My suspicion was that humans are even more alienated by standardised spaces. People have their own personalities and characters, interests and ideas. I thought that people would want to create spaces for themselves.

  It wasn’t until Knight moved into academia a few years later that he had an opportunity to put his hypothesis to the test. His experiment, conducted with fellow researcher Alex Haslam, was ingenious. They took two groups of people and gave them tasks that would be typical in an office. Subjects had to check documents, process information, make judgements and more.21

  The first group were put in the lean condition. They were all required to work in the same, minimalist, superficially efficient space. In fact, almost precisely the conditions that Knight had noted while standing in the office in the West Midlands. The second group were also put into standardised workspaces, but with a difference. This time Knight put prints on the walls and plants by the desks. Knight called this the enriched condition.

  What happened? Performance improved by 15 per cent. Perhaps this isn’t too much of a surprise. People work better in environments, even standardised ones, which are more human. ‘This confirmed what I had suspected for years,’ Knight told me. ‘Most people prefer an enriched environment. People said that the pictures and plants really cheered up the place. Lean spaces may work on an assembly line with regimented tasks, but not for cognitive or creative tasks.’