Why Quarantine Distorts Our Sense of Time

There is nothing like a pandemic and ensuing quarantine to throw a wrench into our routines and remind us of the many things we used to take for granted. Who ever thought that we would pine for a mask-free trip to the grocery store or a lazy afternoon at the cafe? But perhaps the most jarring change since the emergence of COVID has been the fundamental shift in our experience of time.

Something odd happens to time when the slices of the day—the commute, the office, lunch with the coworkers—bleed into one another, with no change of scene to mark one moment from the next. For many, time begins to distort. It becomes difficult to recall which day of the week it is. Hours seem to snail by, but weeks disappear in the blink of an eye. It is during these experiences—when time drags or dashes—that even the least philosophically minded of us briefly ponder its nature. We wonder what exactly this thing we call “time” is, and why our experience of it seems to be in constant flux.

It turns out that thinkers have been posing these questions for millennia, and they are not exactly a piece of cake to answer. In fact, even the mere conceptualization of time presents us with its own set of challenges.

Time, space, and culture

Close your eyes and try to imagine time. You’ve likely constructed some sort of spatial reference point in your mind’s eye, such as a timeline with yesterday on one end, the present somewhere in the middle, and the future on the other end. If you write from right to left (e.g., Hebrew or Arabic) the future most likely sits somewhere to the left. If you write from left to right however (e.g., English), the timeline will reverse direction, and the future will hover somewhere off to the right. But things get stranger still. In English, we speak of “moving a meeting forward” or “looking back in time,” and consequently see time as flowing across a horizontal timeline. People who speak Mandarin, however, often imagine time as flowing vertically, and speak of earlier events as “up” and later events as “down.”1 There are even some cultures, such as the Aymara people of South America, who picture the future as behind them and the past as ahead.2 This conception of time is so foreign to an American that it seems almost nonsensical (just imagine if someone exclaimed “I’m looking backward to my birthday tomorrow!”).  

I hope I’ve convinced you so far that not only is it impossible to conceptualize time without referencing space, but that the chosen spatial reference point varies considerably by culture. Of course this does not discount the possibility of some sort of universal psychology of time underlying our mental and linguistic representations. It is quite likely that all humans have evolved similar cognitive strategies for keeping track of time. Therefore, we can assume that, just as we all fall prey to the same optical illusions, we will reliably experience the same temporal illusions. Take the example of Viktor Frankl’s experience in Nazi concentration camps: “In camp, a small time unit, a day, for example, filled with hourly tortures and fatigue, appeared endless. A larger time unit, perhaps a week, seemed to pass very quickly. My comrades agreed when I said that in camp a day lasted longer than a week.”3 

A day lasting longer than a week! It is tempting to chalk Frankl’s statement up to a poetic turn of phrase, or to the particularly miserable conditions he was caught in, but I think we’ve all experienced similar paradoxical perceptions of time. We’ve all looked back on a particularly mundane week with the sense that it slipped by in the blink of an eye, even though each day had felt interminable at the time.

To resolve this paradox, it is useful to think about colonoscopies.

Yes: colonoscopies. In a landmark study, patients were asked to rate their moment-to-moment pain throughout a colonoscopy procedure (general anesthesia was not used).4 Afterwards, they rated the total pain of the procedure. One of the surprising findings was that the patients largely ignored the length of the procedure in their retrospective evaluations. Instead, they seemed to be averaging their levels of pain at the most intense moment and at the final moment, a phenomenon known as the peak-end rule. This led to a flurry of perplexing results. For example, it was found that adding a few minutes of less severe pain to the end of the experience could actually improve the retrospective evaluation (because the procedure’s less painful ending brought down the peak-end average).

Now, imagine you are the patient, and towards the end of the procedure the doctor gives you a simple choice: would you rather the procedure be terminated immediately or an additional few minutes of moderate, but less severe pain be tacked on? You would probably promptly question the doctor’s credentials, and wonder how he had managed to pass medical school. This, however, is simply the naive response of your experiencing-self—the self that answers the question, “How does it feel right now?”5

Now pretend the doctor never presented you with this dilemma, but instead went right ahead and administered the final few minutes of less severe pain.  Your remembering-self—the self that answers the question, “How was it as a whole?”—would actually be more satisfied with this outcome. And since it is your remembering-self that you consult when making future decisions, the latter scenario would leave you more likely to partake in future, potentially life-saving, colonoscopies. You were right that the choice the doctor presented to you was ludicrous, but not for the reason you thought. It should have been a no-brainer for the doctor to go ahead and extend the procedure. The problem was that you were answering with your experiencing-self rather than your remembering-self.

Time from the vantage point of the two selves

Attention and emotion are the most pivotal players in the experiencing-self’s perception of time. Just like most things in life, we only notice the passing of time when we pay attention to it. Time flies when we are having fun precisely because we are zeroed in on our fun rather than the passage of time. The flow experience illustrates this well.6 Flow is one of the most rewarding mental states, in which we become merged with an activity and temporarily lose our sense of self and time. It can be experienced by athletes, musicians, gamers, and many more. One of the hallmarks of flow—and probably the reason it is so enjoyable—is that our attention is fully absorbed by the activity at hand. Only when we emerge from a flow state, and our attention is free to be deployed elsewhere, do we realize that a large chunk of time has passed. On the flip side, when we are constantly checking the clock or our smartphones, waiting in anticipation or trepidation, our attention refuses to be captured by anything but time. Unfortunately for us, this means that the more we will time to pass, the slower time crawls.

Things get even more complicated when emotion is factored in. Of course, emotion can steer our attention towards time or divert it elsewhere. But its role seems to be much more elaborate than that of mere attention-guiding. The unique, and potentially debilitating effects of emotion on time-perception are most apparent in traumatic experiences.  For example, survivors of motor vehicle accidents report the speeding up or slowing down of time more frequently than any other symptom.7 It is safe to assume that time did not slow down because the survivors were examining their watches while their cars somersaulted down the freeway. Rather, the sheer shock and terror seem to have warped their experience of the world in some fundamental way.

Unsurprisingly, different emotions can distort time-perception in opposite directions. In one study of first-time skydivers aptly titled How Time Flies, those who reported overwhelming dread leading up to the dive overestimated its duration, whereas those who reported less dread and more excitement actually underestimated its duration.8 Numerous theories attempt to unravel the tempestuous relationship between time, emotion, and attention, but for our purposes it is enough to know that it is these factors that most reliably skew the experiencing-self’s perception of time.

If we saw the world solely from the experiencing-self’s point of view, our perception of time would be distorted, but it would not be paradoxical. It is the conflicting reports proffered by the experiencing- and remembering-selves that give rise to these contradictions.

To illustrate, imagine that you are the novice sky-diver who vastly overestimated the duration of your first dive. Since you faced your fears and successfully evaded death the first time around, you decide to take a follow up vacation to Switzerland—one in which you’ll test fate on a daily basis with a week of parachuting off the tallest peaks of the Alps. You get permission from your boss, hop on the flight to Geneva, and, once again, manage to make it through with your bones intact. After that thrill-ride of a week, you return to your routine as an accountant, and slowly you acclimate to the dull rhythm of normal, adult life.

The next weekend your friend asks you how your week was, and you recount your experiences plunging down a cliff-face like a bald eagle and floating serenely into the gentle hills dotted with farm animals and quaint, pastel villages. Your friend is impressed by your stories and newfound passion. It is only on the way back from the restaurant that, slightly embarrassed, you remember that the vacation had in fact taken place two weeks ago. This past week has been a humdrum affair of burnt Starbucks coffee, typing-induced carpal tunnel, and much-cherished bathroom breaks. You realize that, because each day in Switzerland soared by in exhilaration while each day at work inched by in exasperation, you feel like your week in Switzerland lasted a lifetime, and you hardly remember that your subsequent workweek even took place.

While novelty and change fuel the remembering-self, routine and repetitiveness elude its grasp.9 Your memory latches on to the hundreds of new and exciting experiences in the Swiss Alps, so psychological time expands, and a week feels like months. But when it comes to your bland workweek, memory struggles to find a foothold, and psychological time contracts. 

We now have all the ingredients to explain Frankl’s bewildering statement that a day felt longer than a week in the concentration camps. The hourly tortures and humiliations, and the constant fatigue and starvation, were all endured by Frankl’s experiencing-self. Those moments must have felt as if they would never end. Anxiety, sadness, terror, and mortification must have been constant companions. A day, then, felt far longer than 24 hours. Frankl’s remembering-self, however, was the one that evaluated a week in the camps. There were few novel or exciting moments for Frankl’s memory to cling to, rather each day blended into the next, a uniform mass of routine suffering.

Only the good die young

By now it should be clear that attention, emotion, and memory all distort our perception of time. The natural next question is, how can we use this knowledge to our advantage? How can we extend our psychological lives as long as possible—or is that even a goal we should strive for? Is it possible to live a long life but die young, or the converse, live a short life but die old?

In the brilliant book Moonwalking with Einstein, Joshua Foer puts it like this: “Monotony collapses time; novelty unfolds it. You can exercise daily, and eat healthily and live a long life, while experiencing a short one. If you spend your life sitting in a cubicle and passing papers, one day is bound to blend unmemorably into the next—and disappear. That’s why it’s important to change routines regularly, and take vacations to exotic locales, and have as many new experiences as possible that can serve to anchor our memories. Creating new memories stretches out psychological time, and lengthens our perception of our lives.”10

Foer makes a firm case that we can and should schedule our lives in a way that maximizes psychological time. In short, this entails slicing time into smaller segments and accumulating new and memorable experiences.

Before COVID, Foer’s advice would have seemed fairly straightforward to apply. Even if our days had been routine and unmemorable, we could have at least counted on cafes, restaurants, bars, in-person hangouts, and trips to Paris to break up the drudgery. Under quarantine however, any attempt at extending psychological time can seem futile.

Of course, this isn’t really the case—we just have to get a bit more creative. We can still change up our routines, even if the options are more limited. Many have already found new avenues to explore: in the UK, sales of board games and jigsaw puzzled rose 240% during the first official week of lockdown.11 There is always room to learn a new skill, meet a new person, discover a new artist, or glimpse a new spectacle. And even if a mask must be worn, a change of scene is always around the block.  

While Foer’s advice has its merit, especially when it comes to breaking through a rut, I am not so convinced that novel experience-collecting should be our overarching goal. Routine is valuable and psychologically healthy and over time you can easily imagine novelty-chasing losing its, well, novelty. In my next article, I’ll investigate these questions further and try to find a middle ground.

TDL Brief: The Real Drivers of Voter Behavior

As I’m writing this on October 30, 2020, all eyes are on the US election next week.

This is already shaping up to be a historic election, at least in terms of voter turnout. Even five days out from the election, Texas’ early voting surpassed 2016’s total turnout, and nationally, more than 80 million early votes have been cast. This record-breaking early voting pace, accounting for more than 58 percent of the total 2016 turnout, points to the substantial success of numerous “get out the vote” initiatives.

And this is something that should be celebrated. Historically, the US hasn’t been very impressive when it comes to voter turnout. In 2016, only 55.7% of the voting-age population voted. In fact, of all developed countries, US voter turnout ranks 31 out of 35.

But, behind any seemingly successful policy or intervention lurks potentially unforeseen and unintended consequences. As the high turnout of this election is rightly celebrated, it is our job as behavioral scientists to dig deeper in an effort to better understand how these “get out the vote” campaigns may unintentionally bias voter behavior.

For example, in his article “Is a Biased Vote Better Than No Vote?” TDL contributor Sanketh Andhavarapu highlighted that only encouraging citizens to vote, without also educating them on how to become more politically aware, may lead to voters to rely too heavily on cognitive shortcuts such as the status quo bias or the bandwagon effect.

Understanding the real drivers of voter behavior is clearly critical. So, with the 2020 US presidential election around the corner, TDL gathers some outside perspectives on what drives voter beliefs, turnout, and behavior at the ballot box.

Tom Spiegler, TDL Managing Director

1. What influences voters changes as the election nears.

By: Association for Psychological Science, Who Influences Your Vote? It May Depend on How Soon the Election Is (August 2012)

Political campaigns try to influence public opinion in a multitude of ways—TV ads, lawn signs, word of mouth, to name a few methods. And these efforts to persuade have varying rates of success. But research shows that how successful a given method of persuasion is—say, lawn signs—actually depends on how far away the election is.

When the election is next year, rather than next week, voters are more likely to be persuaded by what the majority of their group thinks about a particular candidate. In other words, when the decision of who to vote for is distant and abstract, opinions of peer groups carry a lot of weight. This means that polls declaring, for example, “70% of citizens trust X candidate on the economy” can carry a lot of weight.

Closer to the election, though, it’s a different story. At this point in the election cycle, a voter is more likely to be influenced by individual opinions than by group consensus. For example, seeing a neighbor’s lawn sign or having a conversation with a trusted friend now has an outsized impact when compared with a TV ad highlighting state-wide polling.

2. Are you someone who votes?

By: Scientific American, How Science Can Help Get Out the Vote (September 2016)

One way to increase voter turnout is to capitalize on people’s desire to shape or conform to a worthy self-identity—the identity of “someone who votes.”

A 2011 study presented potential voters with two separate questions. One of the questions was framed in terms of identity (“How important is it to you to be a voter?”) and the other was framed in terms of an activity (“How important is it to you to vote?”).

Would-be voters that were presented with the self-identity question turned out at significantly higher rates in real-world elections: the 2008 presidential election in California and the 2009 gubernatorial election.

Getting citizens to embrace “being a voter” as part of their identity can therefore be an effective tool for governments seeking to boost voter participation.

3. Want to know who will win the election? Focus on citizen unhappiness.

By: MIT Sloan School of Management, Why do people vote the way they do? Unhappiness, says MIT Sloan behavioral scientist, outweighs all other factors (September 2020) 

Predicting election results has turned into a multi-billion dollar industry. Analysts look at everything from income levels to unemployment rates, population age, and racial makeup. However, research shows that subjective well-being might be the most robust predictor of electoral success.

For example, in the 2016 U.S. election, counties where voters expressed low levels of overall life satisfaction were strongly associated with an increase in Trump’s vote share, above and beyond what a Republican candidate in a given jurisdiction would ordinarily receive.

The takeaway, which is deceptively simple, is that unhappy people vote against the status quo.

4. Moral beliefs or political affiliation—what comes first?

By: The Atlantic, What Your Politics Do to Your Morals (September 2019)

Every year, a growing number of people instinctively lunge toward one side of the ballot when election time comes around. One of the factors thought to influence a voter’s political leanings is their fundamental moral beliefs. A voter who values patriotism and respect for the troops might look at the party platforms, see that the conservatives are more pro-military, and therefore be more likely to identify as conservative.

But what if it’s the other way around? What if voters purport to care so much about loyalty and the troops because they first identified as conservative?

There is some evidence to suggest this might be the case. And if this is true, it could suggest something a little terrifying for society: that people often make their judgments of right and wrong fit with whatever party they already support. Decades of research on cognitive dissonance shows us that this is hardly beyond the pale.

Politicians fight to convince voters that their vision for America is the correct—and often more moral—one, but what if voters don’t actually care? Perhaps people just pick their team, and force the rest to fall into place.

Combatting Digital Addiction


At TDL, our role is to translate science. This article is part of a series on cutting-edge research that has the potential to create positive social impact. While the research is inherently specific, we believe that the insights gleaned from each piece in this series are relevant to behavioral science practitioners in many different fields. As a socially conscious applied research firm, we are always looking for ways to translate science into impact. If you would like to chat with us about a potential collaboration, feel free to contact us.


When was the last time you didn’t have your phone within reach? How many minutes does it take after you wake up to glance at your nearest screen? Technological addiction is worrisome, especially given the constraints on our lifestyles during the COVID pandemic, and the resulting dependence on digital communication that has ensued. To learn about how behavioral science might tackle this looming societal challenge, The Decision Lab reached out to Aditya Kumar Purohit, a behavioral scientist who studies digital nudges and marketing strategies. 

Aditya is a doctoral assistant in Information Systems at the University of Neuchâtel, Switzerland. He is also the founder of Nudges for Change, a platform for behavioral scientists to find nudge-related research. In his work as a behavioral change researcher, Aditya designs and tests interventions to investigate their impact on human choices and behaviors online. More specifically, he is interested in designing interventions to improve digital education, digital wellbeing, and online decision making. Previously, Aditya was an analyst at Amazon. He has supported many startups in planning and executing their digital marketing strategies by employing the principles of behavioral science. He holds an MSc in Marketing Research from Université Grenoble Alpes, France, and a Bachelor of Engineering in Computer Science from VTU, India.

A full version of the paper discussed in this piece is available at https://dl.acm.org/doi/abs/10.1145/3334480.3382810


Nathan: How would you describe the focus of your research in simple terms? 

Aditya: My research focuses on designing digital interventions to mitigate digital addiction and improve online decision making — understanding design principles that social media companies employ to make users addicted to their products and how to mitigate the dependence the users develop using digital nudges. I collaborate with startups that build tools to improve digital well-being and apply the insights of design principles to address real-world challenges. The challenges include combating problematic smartphone and social media use leading to anxiety, depression, loneliness, and Attention Deficit Hyperactivity Disorder (ADHD).

Nathan: What sort of challenge did you set out to tackle? 

Aditya: The genesis of social media was to ease messaging and sharing of content between people. To yield profit from their services, social media companies rely on users’ continued attention for their revenue generation. To achieve this continued attention for profits, they design their products to be intrinsically persuasive to attract and hold people’s attention. This approach of making users hooked has consequences such as impairing psychological health and well-being and interpersonal relationships.

Nathan: How did that broad problem coalesce into a research project? 

Aditya: In this research project we followed the design science research methodology i.e., we first identified the problem, which was the persuasive design of social media platforms. Then, we situated the objectives we wanted to achieve with our solution, which was to make individuals mindful of their social media use, and finally, we presented the design of the solution and a real-world demonstrator (i.e., NUDGE) developed by our co-author Louis Barclay.

Aditya: We believe that the same design principles that create addiction could be leveraged to mitigate it. More precisely, digital nudges, in this case, design features that steer people’s decisions without banishing freedom of choice, can be used in welfare-promoting directions. In particular, we used six digital nudges to mitigate the addictive features of social media. They are as follows:

  1. Default Nudge
  2. Hiding Nudge
  3. Friction Nudge
  4. Pause-reminders
  5. Feedback Nudge
  6. Unfollow nudge

Nathan: What did you end up finding out?

Aditya: Though we used a counter-intuitive design approach to decrease social media platforms’ usability, users appreciated digital nudge interventions. They believed that digital interventions reduced their Facebook usage and, at the same time, made Facebook more enjoyable. Our results show that digital nudge interventions helped users become reflective of their social media usage, possibly decreased their time spent, and made the experience of using social media more pleasant.

Nathan: How do you think this is relevant to an applied setting (i.e. in business or public policy)? 

Aditya: Digital addiction is a significant problem. An estimated 210 million people are suffering from social media addiction worldwide. More than ever before, it is especially important now that we find ways to mitigate and balance our internet and social media use. Most designers believe that counter-intuitive designs are bad for user experience. However, our study proved that by employing digital nudges, designers can build useful, likable, and practical apps that will assist users in maintaining healthy internet/social media use and not impair their psychological health and well-being.

Nathan: Do you see future research stemming from your study? In what directions?

Aditya: We’re currently planning and setting up an experiment involving roughly 80% more participants from different geographic locations. We aim to grasp an in-depth understanding of the described nudges individually and their impact on different personality types by conducting RCTs with a combination of usage log data. By studying the user behavior from their usage log data, we will also explore the effect of timing of digital nudge interventions on users’ behaviors. 

Aditya: We welcome new collaborations. If you have a project in digital health, online decision making, or any other field with which you could use some help, don’t hesitate to reach out by emailing me at aditya.purohit@unine.ch

TDL Brief: People Want to Be Outside Cities

People are leaving cities for the suburbs at an increasingly rapid rate. In America, USPS data shows that the rates of people moving spiked during the first several months of the pandemic. Big cities experienced the largest net losses, with New York City, Chicago, and San Fransisco being the biggest losers. 


The reasons behind this accelerated shift from the urban to the suburban are two-fold. First, the pandemic has caused residents of many cities—especially densely populated ones like New York City and San Francisco—to seek out more space. Most often, that means (at least temporarily) leaving the city. 

Second, the pandemic accelerated a shift toward working from home, leaving city-dwellers to question whether living in a city is still worth it. Prior to the work-from-home revolution, the economic and social mobility benefits that cities offered outweighed costs such as overburdened infrastructure, long commutes, and excessive noise. 

Now that many residents can get those same benefits while living outside of cities, however, cities are facing a fiscal crisis as the tax base moves away. Before the pandemic, Phoenix, Arizona was projecting a $28 million surplus for the upcoming fiscal year. Now, it’s anticipating a $26 million deficit. New York City is projecting a deficit of at least $6 billion. San Francisco was running a $420 million deficit before the pandemic—now, the city is expected to run a deficit north of $1.5 billion. 

So how can cities keep their residents, or win them back? Behavioral science offers some low-cost, high impact solutions.

Tom Spiegler, TDL Managing Director

1. Re-thinking public transit.

By: Mass Transit Mag, What Will Transit and Mobility Look Like After the COVID-19 Crisis? (April 2020)

Public transport ridership has dropped in cities around the world, and signs point to the stigma around public transport sticking around even after the worst public health effects of the COVID-19 pandemic are over. 

This crisis offers cities a chance to reimagine public transit, by embracing new technologies and new approaches to improve customer experience.

Effective data management offers a promising way forward for transit companies looking to regain the trust of their customers. By leveraging new tools for the collection and analysis of real-time data, companies can provide riders with much richer information that they can use to plan their journey to be as safe as possible. From updates about how crowded a particular bus currently is, to data about which bus routes are less likely to be busy, the possibilities for optimizing the daily commute are endless. Such initiatives will be essential for controlling the spread of COVID-19, but could also translate into public transit systems that are generally more efficient and more comfortable for their customers.

2. Reducing noise pollution.

By: BBC, How Cities Trick You Into Better Behavior (August 2018)

Let’s face it, cities are loud—and it’s well-documented that noise pollution has adverse effects on health. It has been proven to increase hypertension, blood pressure, stress hormones, risk of heart attacks, and is a major culprit behind disturbed sleep patterns. 

A major source of noise pollution in cities is drivers honking their horns. This problem was particularly bad in Mumbai, India, where behavioral scientists attempted a number of interventions in an effort to reduce horn honking.

In one experiment, researchers installed a red button on the dashboard of cars, which would light up if the driver honked the horn. The light would have to be turned off manually. Installing the “bleep” button on the dashboard reduced honking by 61%. 

In another experiment, the city of Mumbai tested traffic lights that stay red if drivers honked their horns at the light. If the meters registered noise levels of 85 decibels or over, the lights were reset and stayed red for longer.

As these examples illustrate, even low-cost adjustments to urban choice architecture can have large impacts on the wellbeing of a city’s residents.

3. Making cities more walkable.

By: Vox, How Barcelona is taking city street back from cars (April 2017)

The unfortunate truth about most of our cities is that they aren’t built for humans: they’re built for cars. Not only is this less than ideal for many city-goers, who value green space and walkability; it also fuels (pun intended) the epidemic of air pollution that plagues urban environments, contributing to countless health problems and increasing the chances of premature death.

Barcelona’s solution to this problem: “superblocks,” or groups of nine city blocks where the inside is closed off to through traffic—that is, vehicles trying to get from one part of town to the next. Inside superblocks, the speed limit is kept at a nice, calm 6mph. The idea is to make cities more pedestrian-friendly, freeing up more street space for pedestrians.

Similar initiatives have started to take root in cities like NYC and Toronto, which have experimented with closing certain streets to allow for activities like expanded outdoor dining.

4. Adding more green space.

By: The Conversation, How cities can add accessible green space in a post-coronavirus world (June 2020)

The COVID-19 pandemic has highlighted a couple of facts about urban green space: First, that opportunities to enjoy nature are hugely important for the upkeep of both physical and mental health (particularly when we are otherwise being encouraged to isolate ourselves), and second, that access to these areas is massively unequal. As our cities currently stand, the amount of green space is inadequate to meet the public’s needs, and the people who suffer the most are those who are already disadvantaged.

Even after social distancing measures are eventually lifted, green space will be an essential part of the urban toolkit for weathering climate change, helping cities to better manage stormwater, heat stress, and air quality.

By adopting a nature-based approach to urban design, cities can be made more livable and more resilient. The most successful approaches to green-blue infrastructure design start from the understanding that cities are ecosystems in and of themselves, and need to be managed as such. By centering ecosystem services—the benefits that humans get from ecosystems—in urban planning, we can move towards a more sustainable, healthier future for the billions of people who live in cities globally.

How Biases Can Color Entrepreneurial Decision-Making

Innovation and entrepreneurship can be synonymous with uncertainty, and an environment of uncertainty is often the perfect atmosphere for biases to come into play. One of the most pivotal decisions that entrepreneurs have to face is whether or not to pivot (pun intended). Succumbing to certain cognitive biases during such decisions can be costly.

In the startup world, pivoting means shifting to a new strategy. During pivots, the ultimate goal remains the same: to solve a specific problem in society. The methodology used to ultimately address this problem, however, changes. Pivots can take a variety forms. Here are some examples:

  • Changing the platform of your product. For example, moving from mobile technology to a web-based platform.
  • Utilizing a different revenue model and/or moving to a different target market. For example, shifting from a B2C (Business-to-Customer) model to a B2B (Business-to-Business) model. 
  • Adding (or removing) a significant feature to (or from) your product.

Pivoting can help increase a business’s growth, evade large challenges, diversify market reach, and improve customer reach. Embracing the pivot is key to finding success in the world of entrepreneurship. After all, a product idea will rarely have the perfect business model or product market fit from the beginning. The flexibility to pivot is essential for any true entrepreneur. 

At the same time, the decision to pivot should be backed by extensive planning. This article won’t necessarily help you determine if it is the appropriate time for your startup to pivot. Instead, it will discuss the different biases that may affect your ability to make a sound decision. Knowing these biases can help you avoid common decision-making pitfalls, and potentially protect your startup. 

Biases that can influence the decision to pivot

Survivorship Bias

Survivorship bias is the heuristic of focusing on a successful subgroup and believing it to be representative of the entire group, often due to the failed subgroup’s lack of visibility.1 Survivorship bias is especially prominent in the startup world. It is widely known that nearly 90% of startups fail.2 Nevertheless, a majority of entrepreneurs are confident that their venture and entrepreneurial journey will be successful. Oftentimes, these bold statements are backed by references to stories of successful startup founders such as Mark Zuckerberg or Bill Gates (a.k.a. the survivors). What does not cross people’s minds so easily, however, is the number of failed founders. In a sense, survivorship bias is a form of the availability heuristic, which describes our tendency to rely more heavily on information that comes to mind quickly and easily when making decisions about the future.3

Survivorship bias affects decisions to pivot in a similar way. There are several startups became successful as the result of a major pivot: some notable examples include Slack, Twitter, and Instagram. Such stories can misrepresent the success rate of radical pivots. As a result, founders may be inclined to pivot, even when doing so is more of a last-ditch effort rather than a sound business strategy.

Hard pivots are often only a means of stalling failure for many companies.4 Therefore, it is important to recognize when the urge to pivot may only be a reflection of deeper underlying problems with your startup. Understanding survivorship bias can help you navigate this decision with less unrealistic optimism and more sound reasoning, as focusing a bit more on the failures can paint a more holistic picture. And that picture is sobering: One study found that approximately ~10% of startups fail due to an unsuccessful pivot.5 

Of course, successful pivots are indeed inspiring. Just remember: pivots are not always successful. 

Sunk Cost Fallacy

The sunk cost fallacy describes our tendency to consider past investments of resources when making decisions, often resulting in the decision to continue the behavior at hand.6 While virtually everyone succumbs to the sunk cost fallacy, it can be especially common among entrepreneurs, who invest great amounts of time, energy, and capital into their projects with hopes of seeing their vision materialize. This fallacy can be costly, as it can bias entrepreneurs towards executing a strategy or operating their business well beyond the point where they should have given up. Why? They don’t want their “sunk costs” to go to waste. As a result, they continue to waste more resources. 

Sunk costs can make it very difficult to decide to pivot. This relates to the concept of “psychological ownership” in entrepreneurship, which describes entrepreneurs’ tendencies to become one with their ideas, both internally and externally.7 Prior investment can cause entrepreneurs to develop a personal attachment to their product or strategy, which can contribute to the risk aversion associated with pivoting.8 It is important that entrepreneurs do not get too attached to their original or present business model. By avoiding doing so, they can be more open-minded when approaching alternative strategies that could indeed put them on the path to success. It can also help to frame projects in the context of future utility, by thinking and talking about your business in the future tense. 

Overconfidence Bias

Entrepreneurs are passionate about the business they are building. Their commitment and unwavering belief in their startup is often what helps them transform their idea into actuality. However, this can be a double-edged sword, as entrepreneurs’ faith in themselves can affect their ability to notice that their idea may not actually be viable. This is often referred to as overconfidence bias, which describes people’s tendency to overestimate their abilities to perform a task or the probability of success in comparison to an objective evaluation. 

Overconfidence bias can affect the decision to pivot in both ways. For example, an entrepreneur may have great faith in their current strategy, and thus avoid a pivot. On the other hand, they may be inclined to execute a hard pivot, and may do so carelessly due to their overestimation of the likelihood of success. 

Confirmation Bias

The biases described above all contribute to confirmation bias, which is our tendency to more frequently notice and absorb information that supports our underlying beliefs. Biases such as survivorship bias, sunk cost fallacy, confirmation bias, and even overconfidence bias can exacerbate confirmation bias. Entrepreneurs may find themselves selectively focusing on feedback that supports their original idea, and ignoring feedback that suggests otherwise. This, of course, applies to decisions regarding pivots as well.9 

For example, an entrepreneur who just developed their minimum viable product (MVP) may begin to conduct pilot testing. It is likely that the entrepreneur has some hypotheses going into the testing phase. As a result, the entrepreneur may interpret the results of the pilot testing in ways that validate their original hypotheses. As a result, he or she may miss crucial signs that indicate a pivot in strategy is necessary. 

Final takeaways

To pivot, or to persevere—it can be the most difficult decision for entrepreneurs to make, and a crucial one at that. Upon reviewing the various biases that can affect this decision, we also notice a third option quietly lurking: to pull the plug. While pivots are quite common among successful startups, it is important to realize when you are pivoting only to put off admitting failure. Realizing this saves time, effort, and resources down the road. At the same time, it is essential that entrepreneurs are able to recognize when their present strategy is not working, and a pivot is necessary. Understanding the biases described in this article can help entrepreneurs approach the decision to pivot more rationally. 

One final recommendation comes from Eric Ries, entrepreneur and author of The Lean Startup: to seek the “perspectives of outside advisors who can help us see past our preconceptions and interpret data in new ways.”10 While being aware of the aforementioned biases can be helpful, it is often not enough to escape them. Third-parties have the ability to more objectively assess the state of your startup, as they personally haven’t invested their own resources.

Why Does Gender Bias Persist?

When my mother decided she wanted to work outside of the home, she faced incredible challenges. She was told that her place was at home, that she would never match up to a man, and that she would bring disrepute to her family. Employers asked her if she planned to have children, what she would do if her husband got transferred, and even held interviews in hotel rooms where men on the panel sipped whiskeys. Despite the challenges, my mother, and many women like her, battled sexism, negative stereotyping, and discrimination at work, society, and at home to scale incredible heights.

It has been forty years since my mother chose to work. While incredible strides have been made in reducing the gender gap since then, it persists stubbornly in varying degrees across the world.

Evidence shows that global progress on gender parity has been mixed when assessed across several dimensions, including: economic participation and opportunity; educational attainment; health and survival; and political empowerment. While individual countries have made incredible progress, on average, we have a long way to go. According to the World Economic Forum, it will take us 95 years to close the gender gap in political representation, with women holding only a quarter of parliamentary seats and of ministerial positions today. Worse still, it will take us a staggering 257 years to achieve parity in economic participation and opportunity.1

This forces us to ask questions about how gender biases came to be in the first place, and why they have been so stubborn across millennia despite major changes in how humans think or organize themselves. For example, despite the enormity of objective evidence and philosophical arguments that the scientific revolution has produced favoring gender parity, we are still far from being unbiased. Behavioral science could help us understand why we hold these biases that are so hard to get rid of, and more importantly, give us some insights on how to change mindsets and ensure gender parity in society, at work, and at home.

The anatomy of gender bias

There seems to be broad agreement that women have mostly had lesser status and power compared to men across all societies, and that phases of equality have only been brief.2 Hypotheses about why son preferences and general pro-male biases develop range from the relative importance of men and women amongst hunter-gatherers to role relevance in agrarian societies. Identifying the causes of these biases may not be easy, and it may be more worthwhile to identify both specific and broad sets of biases and work to eliminate them.3

One important distinction to draw is between explicit and implicit bias. We are explicitly biased when we are aware of the fact that we are prejudiced against women. In other words, we are acting deliberately—something Nobel Prize winning Economist Daniel Kahneman calls System 2 or “slow” thinking.4 For example, in many parts of the world, people knowingly engage in sex-selective abortions, spend less on educating their daughter compared to their sons, pay women less than men for similar jobs, and caricature women in positions of fame and power.5

However, people who are not explicitly biased, who may even think of themselves as supporting gender parity, can unknowingly be biased in their actions or use negative stereotypes, and potentially even hold the same underlying beliefs as those with explicit biases. Acting on an implicit bias would be akin to System 1, or “fast” thinking without deliberation.6 In other words, one may never realize that one is being biased.7

Central to why explicit and implicit biases develop, and are so hard to get rid of, is the fact that they reflect our core beliefs. Beliefs are essentially energy-saving shortcuts our brains use to make sense of the world around us: they are representations of the world we live in, and are efficient templates to hold as a group and transfer to others.8 Further, processes in the brain involved in abstract belief formation evolved from simpler processes related to interpreting sensory perception. Since we experience the external world entirely through the filter of our senses, we find it hard to see that these perceptions are not objectively real, and may not be accurate.9

Add to all of that the neurological need for the steady state of internal conditions known as “homeostasis,” a dynamic physiological state of equilibrium that leads to a natural resistance to change, and you’re left with something akin to cognitive dissonance.10 We don’t like to have our beliefs challenged, and are unlikely to change what we believe even in the face of contradictory evidence.

Another factor at play is that, more often than not, we tend to yield significant power to society when we evaluate our beliefs, adhering to social norms. Since our beliefs are closely tied to our self-concept, we look for consistency and therefore tend to “pick and choose” evidence (confirmation bias), believe things which seem familiar (mere exposure effect), evaluate evidence and generate evidence, and test hypotheses in a manner biased toward our prior opinions and attitudes.11,12,13,14

Changing gender beliefs and behavior

Beliefs about gender have changed across the world, and will continue to change as people’s philosophies evolve and they adjust to new evidence. However, many argue that people need some sort of intervention to change their views on gender. Some studies have found success in this approach: A study in India found that gender quotas across village councils weakened stereotypes about gender roles in public and domestic spheres, and eliminated the negative bias against women with respect to their ability govern amongst male villagers.15 Research in the US has also shown that legislators who have daughters vote differently on women-related issues, suggesting it might be beneficial to support candidates with female children.16

However, not all interventions have worked. Recent research on whether legal interventions can help change beliefs led to a surprising conclusion: the results showed that having a sexual harassment policy in place at an organization may not alter explicit gender beliefs, and may actually have the unintended effect of activating negative gender biases, which run contrary to the policy’s equalizing aims. Financial incentives have also fallen flat as a means of changing behavior.17 For example, a state government in India, in order to combat son preferences and a rising population, offered its citizens a financial incentive for having a daughter, and a lower incentive for having two girls or one boy. This led to fall in the number of children families had, but did not change the sex ratio. In other words, families still preferred boys over girls.18

It is because of the slow-burn nature and mixed results of interventions focused on directly changing beliefs that Harvard Professor Iris Bohnet suggests “shoving” rather than “nudging,” using behavioral design to de-bias organizations instead of individuals. According to Bohnet, practice must be the first to change, with beliefs following suit later.19 For example, in the hiring process, she suggests the following methods for eliminating bias:

  • Crafting job descriptions to remove bias by making them equally appealing to men and women.20
  • Anonymizing resumes to remove bias against specific groups of people.21,22
  • Using an “evaluation nudge” in which candidates are evaluated jointly rather than separately, stopping evaluators from relying on cognitive shortcuts such as gender stereotypes.23
  • Using structured interviews and tests to ensure objectivity and fairness—they are far more effective in ensuring equal treatment.24

The success of these interventions is not contingent on a change in beliefs—only a change in the choice architecture. By making minor adjustments to the context in which decisions are made, the effects of unconscious bias can be sharply reduced. The question that remains is whether forcing a change in behavior will eventually have an impact on people’s underlying prejudice.

Closing comments

While social movements, spirited individuals, academics, and even whole countries will drive the gender parity agenda ahead, evidence so far suggests it will be long, hard, and drawn-out affair. Reflecting on female and economic empowerment, Nobel Prize-winning economist Esther Duflo observes, “… in one direction, development alone can play a major role in driving down inequality between men and women; in the other direction, empowering women may benefit development… [but] the interrelationships are probably too weak to be self-sustaining, and that continuous policy commitment to equality for its own sake may be needed to bring about equality between men and women.” Indeed, even if we cannot see the results all of the time, we must spirit on because it is the right thing to do.25

Nudging Through Active Socialization


At TDL, our role is to translate science. This article is part of a series on cutting edge research that has the potential to create positive social impact. While the research is inherently specific, we believe that the insights gleaned from each piece in this series are relevant to behavioral science practitioners in many different fields. As a socially conscious applied research firm, we are always looking for ways to translate science into impact. If you would like to chat with us about a potential collaboration, feel free to contact us.


Navigating the topic of helping people make the best decisions for themselves, without infringing on their rights, or missing opportunities to improve their choices, is a difficult balance. To learn more about how this problem is tackled in the academic world, we reached out to Katie Mehr. 

She is a third-year Ph.D. student in the Decision Processes track at The Wharton School. Previously, she worked in Professor Gretchen Chapman’s Medical Decision Making Lab. Before starting at Wharton, she earned an undergraduate degree in economics from Rutgers – New Brunswick. Most of Katie’s research falls into two broad areas. She studies behavior change, or how to nudge people towards completing desirable yet difficult behaviors, like exercising, saving for retirement, or eating healthy food. She also investigates consumer behavior, which includes questions related to advertisements, recommendations, and product reviews. 

A full version of the study discussed in this conversation is available here: https://www.journals.uchicago.edu/doi/full/10.1086/708880


Nathan: How would you explain your research question to the general public?

Katie: We wanted to test whether prompting people to seek out and mimic a life hack used by a peer to achieve a goal of shared interest could be helpful. This is why we call it the “copy-paste nudge” Because you’re “copying” what works for a friend and “pasting” that strategy into your own life.

Nathan: What did you think you’d find, and why?

Katie: There has been a lot of research to suggest that people naturally emulate their peers. But, people may fail to absorb all the useful information that they could pick up from others. For instance, you may have a friend who uses a particularly clever strategy to motivate herself to exercise regularly or to study harder, but you may not notice it or think to try it out yourself. So we prompted people who were pursuing a particular goal, in our case, the goal to exercise more, to find a hack used by a friend to achieve that very same goal and then “copy and paste” it. This would help them learn from and socialize with peer role models, boosting their likelihood of goal achievement. Which meant, in this instance, increasing the amount of time they spent exercising.

Nathan: What rough process did you follow?

Katie: We ran a large experiment where participants were randomly assigned to one of three groups. In our copy-paste prompt group, participants were prompted to find and emulate a friend’s strategy to help them exercise more often. In our quasi-yoked group, participants passively received an exercise strategy to mimic that someone else in a previous study had “copy-pasted.” In the simple control group, participants were not given any exercise strategies. Everybody was, however, prompted to make exercise plans.

Nathan: What did you end up finding out?

Katie: We found that participants who found and mimicked an exercise strategy, in other words, those who were in the copy-paste prompt group, reported spending 55.8 more minutes exercising over the course of one week than those in the simple control group. That meant an average of 32.5 minutes of additional exercise over the quasi-yoked groups as well, which showed that our intervention was even more successful than more passive recommendations.

Nathan: How do you think this is relevant to an applied setting (i.e. in business or public policy)?

Katie: Copy-paste prompts have several advantages that make them easily applicable to a wide range of settings. In addition to being virtually costless and easy to implement, they can be applied to a wide range of problems, like learning how to study more effectively or saving more money for retirement. More broadly, our work suggests that people do not spontaneously absorb all helpful social information from their peers. Thus, prompting people, ranging from employees at a business to consumers looking to schedule their flu shot, to learn and adopt life hacks from others can add value.

Nathan: Do you see future research stemming from your study? In what directions?

Katie: While we’re excited by these initial studies, we’re interested in expanding this work further. For instance, our original study relied primarily on self-reported measures of exercise, so observing and measuring objective behavior would be useful. We also only examined copy-paste prompts’ effectiveness over the course of ten days, so exploring the long-term effects of copy-paste prompts could be important. Finally, we want to know when copy-paste prompts work best and gain more insight into why they work. 

How Might Behavioral Science Transform Education?


Jayden Rae and Marielle Montenegro, two of our senior consultants at The Decision Lab, tell Nathan Collett about their work tackling some of the biggest challenges in education reform. Some of the topics discussed include: 

  • Low-cost, easily scalable solutions in education
  • Effective interventions in higher education, specifically surrounding retention rates for low-income students
  • High-impact interventions using behavioral science
  • Bad behavioral science
  • The value of “reform” vs working within the system
  • The various powers that clash over which textbooks make their way into the classroom
  • Online learning: pitfalls and opportunities

Discussion: Education x Behavioral Science

Nathan: To start off, could you tell me a little bit about where education is currently at with respect to behavioral science? 

Marielle: Behavioural science applied to education is lagging behind in comparison to other industries like finance and insurance. This is likely attributable to the fact that these are for-profit industries, which are focused on efficiency and a bottom line, whereas education is mostly public—so they’re not really focused on the ROI as much. 

Marielle: We’re only now starting to see the impact of potentially applying behavioral science and an understanding of how people make decisions into the education journey in the United States. Right now, education is not necessarily what people think about when they think about behavioral science. The typical nudges that have been applied successfully in the past have been in tax collection or in the health space but education is still kind of a new field being linked to behavioral science. We’re only now starting to see the impact of potentially applying behavioral science to better understand how stakeholders in the education ecosystem (students, educators, policy makers, school administrators) make decisions along different points in their journey. 

Nathan: So what are some areas where it has been applied in effective ways?

Jayden: There are interventions used in education that have attracted the interest of behavioral science, because more typical structural adjustments or measures haven’t been as successful. One issue is called the summertime melt, which is essentially in the summer after first year, a large percentage of students drop out of school. Some interventions have been focused around supporting students throughout the process using behavioral science techniques to ensure that dropout rates are not as high as they might be otherwise.

Jayden: Another successful intervention in higher education has been increasing access to financial aid, by using nudges to increase students’ access to federal loans. Getting federal loans, for example, is a very bureaucratic process, with a lot of steps. There’s essentially a lot of friction in accessing them, which, even if it is not by design, is not absolutely necessary. By creating nudges and integrating behavioral science into messaging, loans have become more accessible, particularly to vulnerable student populations.

Marielle: Yes, so although  the application of behavioral science to education is still in the early stages, we’ve really been able to see through these examples that behavioral science provides an opportunity to apply low cost and scalable interventions. These opportunities can be used to promote equity, because it allows a wider diversity of students to access high quality education. This will ultimately improve student outcomes and success down the line, which we hope will help us progress towards a more equitable society. 

Solutions Unique to Behavioral Science

Nathan: What is the value of using behavioral science in education?

Jayden: I think one of the main values that behavioral science can provide here is being integrated within existing systems. So when we think about changing or shifting education systems, we often think of the word reform. This type of change requires a high degree of political will and usually quite a bit of economic backing, as education is a huge system. These can both be really difficult to organize and mobilize, while behavior change in contrast usually can happen within existing systems. On the flip side, it doesn’t require the same level of will of individuals or economic backing to actually see that change when using our methods. So that’s one of the big values, is that it is a bit more feasible, or a little more nimble.

Jayden: The second main value that I see is that behavioral science can really bring implicit blind spots, biases, and other social, environmental, and structural barriers to the surface, which might be impeding positive outcomes, but aren’t extremely obvious. Behavioral research can help understand some of the underlying reasons why big challenges or big barriers might persist over time. In particular, these can be used to identify specific barriers that implicate disadvantaged or minority populations. There’s a lot of data suggesting, for example, that students of color are less likely to access higher education. There’s higher attrition and they are less likely to advance into more lucrative jobs after graduation. So there are a lot of very deep embedded barriers that behavioral science can bring to light and give us clues on how we might be able to address them.

Jayden: One last area of value is that behavioral interventions can be quite cost-effective. Our insights can seem small but make a significant difference by changing some choice architecture, changing messaging and framing, or providing simple tools and checklists. These strategies seem extremely small and in many ways they are. They’re also effectively costless, which means that testing these interventions is more feasible but they can have quite strong impacts when they’re applied at scale.

Nathan: Cool. So you mentioned issues around feasibility and funding. For example, in the for-profit industries, it’s smooth to get to that ROI focus really quickly. Who’s funding behavioral science interventions in education? Who’s seeking you guys out and finding your value?

Marielle: That’s a tricky question because from my perspective, behavioral science is already starting to be integrated into governmental systems. Consequently, there is greater awareness about it. It has become a buzzword that’s being implemented by a lot of nonprofits and foundations. By and large, they’re the ones funding behavioral science in education at the moment. But there’s definitely been a lot of research that is starting to be done in academic institutions around the value of behavioral science in education.

Incentives and Goal Structures

Nathan: One thing I’m curious about – you mentioned how sometimes it’s easier to get behavioral science units going when profit is involved and when it’s a private organization – do you have any knowledge about whether behavioral science has been used at all in charter schools, or for-profit education systems that are generally a bit more targeted towards the bottom dollar?

Nathan: My intuition is that that drive might not be there. And their goals might be of a very different nature, right? Like their interventions might not be to get more students to graduate but maybe just attracting the kind of student that they would want, the kind that can pay tuition, for example.

Jayden: I’m not sure if there’s such a clear delineation between the ultimate goals for-profit and non-profit education systems in education, just because the outcomes we want to achieve are very different in education. They are less concrete; ideally, it’s more about student outcomes and student learning than profit. It’s harder to measure those but I do understand the point around for-profit institutions. They may act on different incentives but education is a unique beast in the sense that we’re essentially providing a public good.

Marielle: We haven’t worked with private K-12 schools but we’ve definitely interacted with for-profit university systems and we are trying to understand how we can improve access to their programs and services for a diverse range of students. We’re definitely thinking about how they can be part of this systematic reform, but it’s not, like Jayden said, focused on trying to increase their bottom dollar. It’s more just actually focused on student outcomes. Because they are not necessarily an evil. They’re a necessary part of the system.

What Sets Education Apart

Nathan: Okay. Going from there, you’ve talked about how education has different outcomes than other areas where behavioral science has been applied in the past. Education is geared towards student outcomes and increasing learning as opposed to increasing profit. What have you learned by applying a behavioral approach to better understand issues in education and how we can address those issues?

Jayden: Recently we have worked on two main areas of education. The first one is in the context of curriculum adoption. We’ll define curriculum as all of the instructional tools that teachers use to communicate subjects and concepts to students. From a super high level, one key thing we learned in instructional materials research is that there is a really bad choice environment in curriculum development. It is really hard for school districts or educational institutions to choose instructional materials that are best suited to students’ needs. Some of the reasons that this is the case is that there are tons of choices to choose from and there aren’t clear signals about which are high quality and which are not. There are a lot of incentives from publishing companies, for example, to kind of try to co-op the market. So there are a lot of issues related to power at play.

Jayden: And so through our work, we identified really specific barriers at the behavioral level made by individual decision makers within organizations. Some of these barriers included a tendency towards the status quo. For example, if you’ve bought a textbook from Pearson for the past 15 years, your next adoption cycle might result in a textbook from that same publisher. This could involve barriers related to teachers being most interested in textbooks that have really nice pictures on the front but may not actually communicate concepts in the most clear way to students. A lot of the issues were actually related to limited stakeholder engagement. Not having really clear strategies to engage stakeholders meaningfully throughout the process to ensure that the ultimate decision reflects their needs was crucial. Some of the consequences of these barriers is that there are poor student outcomes associated with having low quality instructional materials that aren’t well suited to their needs.

Jayden: Through this work and by uncovering a lot of these barriers, we began testing interventions, which we are still doing. Some of these are related to improving the choice environment for major decision-makers as well as students and teachers, and others involve providing tools to decision makers to use more evidence and be more deliberative in their decision making. Some of the other interventions are around improving group dynamics amongst decision makers, so that they can arrive at a consensus. Through this work, we’ve realized that it’s a really complex process. Education systems are not going to change very quickly but behavioral science can definitely play a role in at least reviewing some of the barriers and providing some potential solutions that we can test out, outcomes that will hopefully have ultimately positive outcomes for students.

Marielle: I think it is worth mentioning the other project that we’re most recently been engaged with, which is teacher preparation. It’s important to note that for both these projects, these are in the context of the United States. You can see how the barriers that we have identified to improving education systems are applicable in other education systems around the world. Because it’s really around how the decision environment is structured, as well as the factors around it, like your social environment, your financial context, and how that can affect your success as a student.

Marielle: With this teacher preparation work, we were looking at how we can improve the access to high quality teacher preparation programs to a diversity of candidates. Basically how we can increase the options for promising teachers and incentivize them to make the right choices. We looked across the teacher preparation continuum, from the time they decide to become a teacher, how they choose their program—including how they choose a high quality program versus a lower quality program—and then how they go through that program and then get hired in a district as a teacher and continue to teach over time. Part of the issue that this is addressing is that there’s a shortage of teachers in the United States but there’s also a shortage of a diverse range of teachers from high quality programs. So we’re specifically looking at increasing the diversity of candidates that enter the teaching profession. 

Nathan: How do you do that?

Marielle: We’re still figuring it out but for our research, we did around a thousand surveys to teaching candidates, admin faculty of teacher preparation programs, district hiring managers, and beginner teachers. We also conducted about a hundred interviews with these different stakeholders to get a sense, more qualitatively, of the barriers that they face in their specific portion of the teacher preparation pathway or continuum.

Marielle: One of the barriers that we identified that were really salient was that there’s quite poor information architecture of teacher preparation programs. It’s not really clear when you’re deciding to become a teacher which programs are high quality. There’s not a clear definition for this, what program types are available, what financial aids are available. And there’s a lot of reliance on environmental and social factors. Often, people will just choose things that are most geographically accessible to them. They will choose a teacher preparation program that’s within their city or their town, or something that’s recommended from someone in their network. So this decision is frequently reliant on the most, by default, accessible sources on information, that aren’t necessarily considered in a holistic sense to see what would be most impactful or high quality to actually enter into.

Marielle: We also found that there’s quite a limited availability of appropriate student support services to manage the teachers’ workload, especially because they often have to do a lot of co-occurring internships on the path to becoming a teacher. Prospective teachers are lacking guidance around financial support, how teachers are able to manage their finances is a bit of a black box because it’s a really expensive program to begin with. And one often has to begin with an undergraduate program before even starting teacher’s college. It’s already a lot of investment early on, with very little promise of return, which is why a lot of people end up dropping out. And it very much favors people in more privileged circumstances. So really trying to think about ways that our interventions can access a wider diversity of candidates by providing them with the necessary toolkit to enter into this program and succeed is central to our work.

Potential Limitations

Nathan: Do you think that structural issues like finance and the high costs associated with it are things that behavioral science can address? Or is that beyond our reach?

Marielle: I think at the bare minimum, it can start addressing it. Because, for example, some of the interventions that we came up with address accessibility, not availability, when it comes to finances. This is addressing barriers. It’s not necessarily that there aren’t these financial opportunities or aids available, it’s that they’re not accessible, or they’re not well known, or teacher candidates find out about them much later, when they’ve already invested a lot of money in loans or debts, et cetera. So it’s really just at the very minimum, providing the information in a comprehensible way, at the right time. 

Marielle: So some of the interventions we came up with were opting in candidates’ default to access to information and providing them with workshops. If they’ve been identified as a low income or potentially vulnerable candidate, giving them the resources that they need early on, in the beginning, so that they have this to fall back on without having to search for it themselves, amongst other things that they probably already have to consider.

Nathan: Nice. I’m going to pivot a little bit. Have you ever seen behavioral science applied poorly in education? Are there any times where behavioral science has gone badly?

Jayden: I think the first thing to call out is that when there are examples of behavioral science being applied poorly, there are usually a couple of reasons at work. I think the first one is that any behavioral intervention should be tested first, before it’s scaled up. Whenever we design any interventions, one of the first things we do is consult with stakeholders who are on the ground and have a really clear understanding of the context, as well as an understanding of some of the potential negative effects that an intervention could have on particular groups, so that we don’t overlook that in an intervention design.

Jayden: Also, we try not to take shortcuts. Sometimes behavioral solutions can be overly simplistic, and kind of ignore complexity. So trying to strike the right balance between having interventions that can address really challenging issues with something that’s real holistic at the same time. So that’s always a really hard thing to strike.

Jayden: From one really particular example, there’s a great book called Weapons of Math Destruction, which focuses on the role of algorithms and AI when applied to social challenges. In the book, one of the key examples is the Boston public school system. This school system implemented an algorithm that was designed to kind of weed out poor quality teachers in the schools. One of the key inputs to the algorithm was the test scores of students in the elementary schools. This was used as an indicator of how high quality teachers were, because the expectation was that poor quality teachers will probably work in classrooms where children are underperforming. This actually led to a really large percentage of teachers of color being fired from their jobs as a result of their students’ test scores. What that algorithm didn’t take into account was the fact that most of these teachers were working at Title One schools, which are schools with a really high percentage of low-income students, paired with the fact that the teachers who work in these schools are more likely to be of color.

Jayden: So essentially what I’m trying to illustrate with this example is that when we try to find causation, it’s not always there. And sometimes behavioral interventions or other types of interventions don’t lead to socially positive outcomes.

Marielle: It’s definitely not perfect but I think the best thing you can do is really not to make any assumptions about what the challenges are. Our brain is wired to make assumptions, we have to if we want to easily take in all the information around us. In order to take our tendency to take these shortcuts into account, we try and avoid assumptions whenever we can. Our research questions are generally quite open, and we ask quite a broad range of questions. A lot of very quantitative questions but also a lot of open-ended questions that we can probe deeper into if we find something interesting.

Marielle: With the example that Jayden illustrated, context is really important. Through the projects that we’ve done, especially with teacher preparation, where the context was so varied, depending on which pathway to teaching they were taking, it’s really important to consider where the candidate for example is coming from, and what kind of background they have, and how this will affect their decision making.

Marielle: Getting a good understanding of these contextual factors is really important in applying behavioral science successfully, so to speak, within education.

Nathan: This reminds me of my talk with Brooke, when we discussed applying evidence without a proper framework of considering your values and your norms, which structure the choices that you make when you’re applying research. I got the sense that a lot of what you’re saying comes out of basically just applying math without fully thinking through what you’re doing ahead of time. Or maybe just kind of coming to conclusions based off of biases that are kind of prior to the evidence gathering phase of the process.

Marielle: One thing that sets us apart at The Decision Lab is that we tend to think of things through a behavioral framework but we don’t apply it early on. We apply it a lot later on. So we use it to kind of conceptualize our questions but in terms of actually designing interventions, we don’t try to shove things through a very limited framework that doesn’t really capture a lot of the complexities. So I think that kind of helps us avoid this poor application and mitigate potential unwanted outcomes.

Specific Outcomes with High ROI

Nathan: Interesting. What are the highest impact areas in education that behavioral science can support?

Jayden: One specific area of interest is financial aid. Very specific interventions like increasing the defaults for a loan package can boost student GPAs and increase the likelihood of participation in the program. Another area is in applications to colleges, where simplifying forms and creating new methods for assisting students can reduce the time it takes to apply. This makes it more likely that busy students, who may be handling jobs or pressures at home, to apply to college, since it is less of a pain to do so. 

Marielle: Another really high impact area is in academic outcomes. Strategies like personalized emails to tutoring sessions can reduce the burden of pursuing college. We increased tutor registration by 34 percent in one of our projects. This most likely helped these students adapt to university and find a place that is a good fit for them. 

Looking to the Future

Nathan: Okay. That’s very interesting. Going forward, what are some ways that behavioral science should be integrated into education? What are some things that may not have been done so far, and ways that we could better integrate behavioral science into the design of education systems and resources?

Jayden: I think one really pertinent area right now is integrating behavioral science into online learning resources. Millions of students right now have been forced into an online learning environment that isn’t necessarily well-designed for learning use. So some principles around design thinking, designing online resources could be extremely helpful in increasing student learning. So some really specific examples would be for example, creating reminders for students when they need to complete their assignments, or nudges around when they have to enter class. So just really simple visual cues to communicate student expectations would be really helpful with online tools.

Jayden: The student experience in most online courses can be meaningfully enhanced using evidence based behavioral design to eliminate or ameliorate behavioral barriers that undermine student learning. It’s such a novel environment that is just filled with information. Things like timely reminders to complete assignments, or removing hassles that make simple tasks effortful, can reduce the barriers that face over 25 million students across the United States, especially those that already face challenges, such as low-income or marginalized young people. 

Nathan: What are some examples of how you might tackle information overload? I imagine that’s a huge problem for students doing online learning right now, where things seem overwhelming. There seems to be too much information and it’s not accessible or clearly presented, just because it’s a totally new format that hasn’t been well-developed.

Marielle: Kind of breaking things down into manageable chunks is a really big thing in behavioral science, because once people see that it’s kind of parceled out into specific things that they can accomplish, it kind of taps into that feedback loop where they feel positively reinforced once they complete a task. So they complete that, and then they go onto the next one and it seems a lot more manageable that way. And it kind of improves your intrinsic motivation, because you feel like you’ve accomplished something.

Marielle: And I think that’s already starting to be applied in a lot of online learning systems like Coursera. I think when schools and universities have transitioned into online learning, they haven’t really had the time to adapt to those systems that are already really well established. And so they’re really behind. And I think a lot of people can kind of relate to the fact that their professors haven’t really adapted to this sort of best practice for online learning as well as courses and online learning systems that have already been online for a very long time.

Nathan: Do you think there’s anything we can learn from sites like Khan Academy, or even Wikipedia, or other places online that are well-recognized and manageable? People seem to be able to digest them in a way that they can’t yet with online learning.

Jayden: I mean, those resources are often tapping into principles of good teaching, versus things that are strictly behavioral in nature.

Marielle: This question is probably best answered by going back to the question of how behavioral science should be integrated into the design of education systems and resources. Online learning is obviously a very salient area right now, because a lot of the world is functioning online. But I think over time, we need to start thinking more about how behavioral science can be embedded into the design and process improvement of these education systems. So making sure that you’re asking the right questions when you’re gathering data about the success of programs or students in order to identify barriers that might be overlooked if you’re just looking at test scores, for example.

Marielle: For example, the thing with online learning is that we might be designing really optimal systems to engage students but we might be overlooking the fact that a lot of students are not necessarily in the right context to access these online tools and resources. Whether that is due to being under a lot of financial stress or having to care for children or dependents, it can make it a lot harder to solve problems by just changing defaults and simplifying systems. There’s a lot of other external factors that behavioral science allows us the opportunity to consider, which we might otherwise overlook. Embedding behavioral science within the design of systems from early on will ultimately be more powerful than just kind of applying it ad hoc, which is what most practitioners seem to be doing with it right now.

Nathan: That sounds like a great place to wrap this up. Thank you very much for sitting down to talk to me. 

The Hidden Power of Intellectual Humility

The 2008 financial crisis was undoubtedly a convoluted event in history. Despite its complexity, in its aftermath, many economists aimed to provide bulletproof explanations for what went wrong. Russ Roberts, a notable economist, contributed his own explanation through an essay and subsequent book, both titled Gambling With Other People’s Money. Yet he later revised his essay, admitting that while he still believed in his own theory, he is now more open to other explanations, and is continually learning more about issues that he was supposed to be an expert on.1

In an interview, when asked why he changed his mind, Roberts said: “I became deeply aware of my ignorance. I had no idea how the housing market worked … I’ve become more epistemologically humble, which is a fancy word for saying ‘I don’t know.'”2

This one statement highlights a subject of interest in science today: intellectual humility, or “considering that one might be wrong.” Turns out, there is a surprising amount of benefit to this taking approach. 

The science behind “we don’t know what we don’t know”

We are cognitively wired to think we are right more often than we think we are wrong. The Dunning–Kruger effect, a well-known cognitive bias, causes us to overestimate our knowledge or ability in any given area. The researchers Dunning and Kruger discovered this effect by testing study participants on tasks including humor, logic, and grammar, and subsequently asking participants to rate their own ability. Those who scored lower on the tasks overestimated their ability significantly: on average, they predicted they would rank in the 62nd percentile, when in fact they were in the 12th.3

Essentially, the less we know about something, the less we realize we know nothing about it. In contrast, the more we know about something, the more aware we are of all the things we don’t know about it. It’s a unique effect that is both frustrating and enlightening: it shows that time and effort may not just improve our ability, but our realization of our lack of ability. 

Adapted from: Squad. (2018, December 13). Dunning-Kruger Effect: Definition, Test, Examples & Quiz. Science Terms. https://scienceterms.net/psychology/dunning-kruger-effect/

Why does it happen? First, it’s more difficult to realize we are bad at something when we don’t understand it well. Think of a newbie learning proper form at the gym, or someone with poor grammar trying to write a book. Getting better takes practice, and with that comes the realization of weak points. In addition, we are also cognitively wired to take mental shortcuts, which means our brains avoid spending time thinking about our abilities in any given topic. Yet, with increased reflection on how we learn, practice, and perform, we can improve. 

But it’s not only novices who are prone to overconfidence. Daniel Kahneman and Amos Tversky found that even trained statisticians overestimate their ability to use proper sample sizes in their work. In a famous study on the so-called “law of small numbers,” they gave statisticians realistic research proposals and asked them to choose sample sizes, estimate the chance of success or failure, and give advice to a fictional grad student. The majority of the respondents made errors in their work.4 Evidently, even experts fall victim to overestimating their abilities. 

One might think that this occurrence doesn’t happen in science. The concept of questioning existing beliefs is at the center of the scientific method, which guides researchers as they look to prove or disprove theories. Yet, according to Julia Rohrer, self-correction isn’t as embedded in the culture of science as we may think. 

Rohrer, working with other scientists from the US and Europe, founded the Loss-of-Confidence Project last year. The group invited psychology researchers to submit statements about times they had lost confidence in their previous work, and why. They found that although many scientists privately disclosed a lack of confidence in their findings, those views never become public, which is detrimental to the advancement of science.6 The Loss-of-Confidence Project shows that, as scientists, we need to improve our ability to admit our downfalls. 

More often than we’d like, situations arise in which the “right” answer does not exist. Think back to the viral sensation of “The Dress,” which threw the world into an argument over whether or not the colors of the dress were black and blue, or white and gold.

Source: The ACTUAL colour of The Dress revealed. (2015, February 27). The Independent. http://www.independent.co.uk/life-style/fashion/news/the-dress-actual-colour-brand-and-price-details-revealed-10074686.html 

This viral photo highlights how an ultimate truth may not exist in some cases.6 Yet we still avoid admitting we might be wrong—especially when results and opinions matter. This desire to be correct contributes to our overestimation of our abilities, and lead us to make potentially dangerous mistakes. 

However, research shows that intellectual humility may help us overcome this. 

Intellectual humility: how admitting our downfalls makes our work better 

In essence, intellectual humility is recognizing the possibility that we might be wrong. It is a mode of thinking in which we see our limitations and are more open to the experiences and feedback of others.8 Those displaying intellectual humility are generally more curious, driven to pursue knowledge, and receptive to feedback.9

Intellectual humility improves memory and mastery of techniques 

Researchers from the University of California, Davis, and the University of Pittsburgh examined how intellectual humility helps individuals seek out challenges and persist through setbacks. They found, across several studies, that intellectual humility boosts our ability to learn, and that those who were encouraged to be more intellectually humble put more effort in learning about a topic which they failed initially to master.9 In addition, improving intellectual humility improved participants’ memory for the information they read, and increased their awareness of their own knowledge.10 

Intellectual humility encourages open-mindedness 

Besides improving task mastery and memory, intellectual humility also helps us be more open to other viewpoints. In a time of polarization, debate, and disagreement, we are continually provided with information that goes against our opinions or what we hope is true, but we don’t necessarily pay attention to it. Confirmation bias is our tendency to remember and interpret information in a way that is favorable to our pre-existing beliefs. It can get in the way of innovation and success, and can even harm relationships.10 

Individuals displaying more intellectual humility are less susceptible to this bias. A study found that those with more intellectual humility spent more time reading sentences expressing differing opinions than their own, compared to those with lower intellectual humility. In addition, intellectual humility boosts open-mindedness and reduces social vigilance, which helps with collaboration.9 By recognizing and embracing the possibility that we might be incorrect about our beliefs, we become more willing to learn about alternative perspectives and make better decisions because of it. 

Intellectual humility boosts innovation

Intellectual humility can also help us innovate, by making us more open-minded. This is because the ability to make connections with individuals who are different than ourselves, and a willingness to learn about other cultures, helps spur creative thinking. A large study by several experts from the United States and Europe found that individuals who actively exposed themselves to new cultures were more productive and more likely to become an entrepreneur.11 Take, for example, Mark Zuckerberg and Steve Jobs. Jobs was an avid supporter of “going East,” and regularly visited ashrams in India to gain clarity when he needed to make difficult decisions. He advised Zuckerberg, who at the time was struggling to position Facebook, to do the same. This had profound benefits for the future direction of Facebook: while he was travelling, Zuckerberg witnessed the close connections people had with one another, and the experience affirmed his sense of Facebook’s mission.12 By helping us remain open-minded, intellectual humility doesn’t just help us become more tolerant of opposing views, but also helps us remain open to experiences that allow innovation and productivity to flourish. 

Intellectual humility and certainty

Despite the benefits of intellectual humility, those who have it still have their share of convictions. One study found that those with more intellectual humility were found to have more strongly held beliefs.12 The reason was that these individuals have more information available to support their strong beliefs—due to their tendency to question their own beliefs and ask questions of the other side.13 

This level of confidence is often useful, as there are moments that require certainty in our answers. Careers in science, consulting, and other fields require employees to have high confidence in their abilities. Intellectual humility forces us to recognize our downfalls and question them to provide more certainty. But, it is important to be cognizant of our convictions. Michael Lynch, a philosophy professor, states that “It’s bad to think of problems like this like a Rubik’s cube: a puzzle that has a neat and satisfying solution that you can put on your desk.”7 Although certainty is desired in many fields, in many cases it isn’t achievable and may do more damage in the pursuit. 

The downside of intellectual humility

While intellectual humility is a fundamental tool for improving memory, learning, and open-mindedness, it comes with its downsides. By definition, intellectual humility does not necessarily suggest a lack of confidence in one’s ability. However, humility exists in both “appreciative and self-abasing” forms.14 Some humble individuals have experienced success, and are more confident in their abilities while still being open to feedback. But for others, humility arises after experiencing consistent failure, causing individuals to underestimate their ability and concede to others in order to avoid negative feedback.14 Therefore, it is important to recognize your behavioral “blind spots” without becoming self-effacing or losing faith in yourself. 

Using intellectual humility to your advantage

It’s difficult to be wrong. Sometimes it’s difficult to even admit the possibility of it. Yet doing so can help improve how we learn and how we work, and make us more agreeable to others around us. In an age where we are constantly bombarded with information, asked to deliver results, and left to navigate a polarized and political world, we can use intellectual humility to ensure we make better decisions. 

Here are some takeaways on cultivating intellectual humility, in life and at work: 

  • Ask more questions of those different than you. Josef Pieper, a German philosopher, suggests that “the natural habitat of truth is conversation.”2 Engaging in dialogue and conversation with diverse opinions is the key to unlocking intellectual humility. And do so with a touch of empathy.13 Taking time to understand others and why they may disagree will help bolster your understanding of issues and your work. 
  • Celebrate failure: Before we can learn from our mistakes, we have to make them. But few of us want to be the only person in the room who admits being wrong—even if there are benefits for learning and doing better work. Leaders should focus on building a culture that embraces failure, to inspire better discussion and better results. The culture of science also needs to focus on building a community of scientists that aim to prove themselves wrong. 
  • Remember, it’s not all about you: It’s hard to admit when we are wrong, but it may not do the damage we think it does. Those who admit they are wrong are typically not viewed as incompetent, and instead are seen as more friendly and open-minded.9,16 

Do you want to live in a delusion of what you think is true, or do you want to live in the reality of what is true?

Neil deGrasse Tyson

Certainly, this year has been one to prove us wrong. Travel plans have changed, education has moved online, companies have pivoted, leaders have changed their policies. I certainly was wrong about most of my predictions for what 2020 would hold. However, research shows that embracing our errors and admitting our potential to be incorrect can in fact strengthen us. So, just as Russ Roberts admitted his doubts about the 2008 financial crisis, we can embrace uncertainty about today’s era and work together to create a better outcome.