D Cognitive Lessons from Behavioral Economics
In the preceding chapter I made an analogy between the way we evaluate real physical objects, including money, and the way we evaluate ourselves, our self-esteem, measured in fictitious units called creds.
If we push the analogy a bit more, we get a new perspective on the hypothesis and scientific thinking.14.D.1 Take a Chance
How much we're willing to expose our egos or how much we need to protect them depends on the balance in our cred accounts. How many creds can we afford to risk in putting forward a hypothesis? If we play it too safe, we short change ourselves, our colleagues, and the taxpayers, donors, and funding agencies who support us. We should be willing to take intellectual chances, as Stuart Firestein14 recommends. Maybe we can apply that same boldness to our intellectual projects. Granted, encouraging and rewarding risk-taking will mean that the scientific community and its funders must be willing to undertake structural and cultural changes in how science is done. We can take modest, and inexpensive steps in the direction of intellectual boldness by being less timid and more forthright in stating our hypotheses explicitly.
14.D.2 The Greater Good
In a way, science faces a situation like the one described by the behavioral economist and Nobel Prize winner Richard Thaler15 who once asked a group of high- level money managers at a large financial firm if they would each take on a mildly risky project that had a good chance of turning a big profit for the firm. The odds of success were good, but success was not guaranteed. A large majority, 85%, of the managers reported (anonymously) that, despite the favorable odds, they would not take on such a project because, if it failed, they would look bad; it would reflect poorly on their individual records.
When the firm's CEO was asked what he would want his managers to do, he said he hoped that they would all take on the risk.
The CEO's reasoning was straightforward: if each project had a good chance of succeeding, the combined odds for success in the group of them were so good that the firm, as a whole, would be almost guaranteed to come out ahead, even if a few individual projects didn't pan out. For the firm to maximize its gain, all the managers would have act courageously, so the challenge was to merge the perceived self-interests of the individual managers with the overall goals of the firm, to ensure that managers would be rewarded for taking chances.Likewise, the scientific community stands to profit when individual scientists are rewarded for sharing the fruits of their scientific training, their thinking, especially their hypotheses, more openly than they now do. In fact, science may be in a better position than the business firm studied by Thaler. Whereas losing a monetary bet would cost the firm in real terms, putting forward a legitimate hypothesis that is falsified is a benefit, not a cost, to science. We advance by trial and error, as Popper stresses. Error is an inevitable and valuable part of the process. Remember the engineers at the TESTCo tire company (Chapter 2): you can only make better tires by trying lots of designs, most of which are destined to fail.
One problem is that we're raised from an early age to want to be right. We've always been rewarded for it. How can we turn around and see our falsified hypotheses for the positive contributions that they are? One possibility is that we can learn from the economic rule to ignore sunk costs.16
14. D.3 Ignore Sunk Costs
At one time or another probably most of us have been confronted with a situation like this: we've spent a significant amount of money to go to a future event, say a basketball game, and then, when game day arrives, we're ill—headache, plugged sinuses, the works. We begin to dread going to sit for hours with hundreds of screaming, semi-inebriated fans while feeling terrible to boot.
We go anyway because, if we don't “the money will be wasted.”This decision makes no sense to a rational thinker: the money that's been spent is gone, and you won't get it back no matter what you do—it's a sunk cost. “Honoring” sunk costs by allowing them to influence your future behavior is irrational, and if you honor them you are committing the sunk cost fallacy. Folk wisdom recognizes the fallacy when it warns you not to “throw good money after bad” and to “cut your losses.” The fallacy is probably linked to loss aversion (Chapter 11) because the anticipation of loss of money or creds scares us into sticking with a set plan even when sticking with it no longer makes sense. And the sunk cost fallacy goes beyond money matters: countries keep fighting obviously lost wars—“So our soldiers will not have died in vain”—although continuing to fight will cost thousands more lives. You should keep fighting only if you think you have a chance of winning the war, not because of the heroic sacrifices your soldiers have already made. Likewise, you shouldn't go to the game just because you bought the tickets; you should only go if you expect to enjoy going more than you'll enj oy staying home.
The sunk cost fallacy can influence scientific thinking. Our egos get involved. You invest creds when you state a hypothesis, and, in one way, this is a good thing because you'll deserve the rewards that come from advancing science with a new idea. But what happens if your hypothesis is wrong? What do you do? Doggedly defend your hypothesis because you've invested creds in it? Put up objections or carry out ad hoc revisions? In short, do you insist on its correctness simply
because you’re afraid of “wasting” your creds? If so, you may have fallen for the sunk cost fallacy.
If your hypothesis is truly wrong, let it go. You made a positive contribution by getting others to think about a part of nature in a new light. You may have prompted your competitors to build on your idea.
Be proud of that and move on.Notice that the question here, as it was with the game tickets and the wars, is not whether you should abandon your hypothesis at the drop of a hat; if you have good reason to think that it is correct, then, by all means, defend it. But focus on finding new ways of testing it or challenging the rival hypothesis, not on protecting your ego.
14.D.4 The Sunk Cost Fallacy and Opportunity Cost
The sunk cost fallacy may affect individual scientists well before their results are published. I am pretty sure that I am not the only researcher who stayed with an experimental project for too long, working on it past the time when I began to realize that it wasn’t going to lead to the nifty discovery that I had first dreamed that it was. In part, I kept at it to keep from acknowledging to myself that I’d already wasted a lot of time, effort, and money on it. A classic case of falling for the sunk cost fallacy.
Unfortunately, committing the fallacy was not my only blunder. By sticking with the disappointing project, I also lost opportunities to work on more productive ones. Economists, because they are economists, reckon missed opportunities as real monetary costs, called opportunity costs. If you spend $5 on the latest designer cup of coffee, you’re not only out $5, but you’ve given up the chance to buy something that might have had more lasting value. Evaluating the opportunity costs when it comes to buying and selling is straightforward; you know how much money is involved. Opportunity costs are not always monetary; if you stay home alone to read a book, you give up the chance to go to a movie with your friends.
People are more willing to abandon an investment if there is not a lot at stake. If you’ve spent only a little money for those basketball game tickets or got them for free, then you are more likely to stay at home nursing your cold. If you’ve spent a fortune on them, doing the rational thing is much harder. However, it is not only what you actually paid that is important: how you perceive the cost matters, too, and the influence of a cost wears off with time.
People who’ve paid up-front membership fees to join a health club around New Year’s Day are conscientious in going to the club at first, but then taper off, only to experience a new burst of energy when the semi-annual bill comes due or bathing suit season arrives.In other words, if real or perceived costs to you are low, you should be better at ignoring sunk costs. When it comes to expressing your scientific thoughts, making a habit of explicitly stating your hypothesis would make it seem less like a risky exception and would decrease the costs associated with it. Fewer personal creds would be on the line, and it would easier to walk away from a falsified hypothesis.
The history of science assures us that the overwhelming majority of hypotheses will, sooner or later, be falsified at some level. Ideally, there would be no downside to putting forward your hypothesis explicitly or, contrariwise, no reason to slip it, sotto voce, into your papers in an implied state. We invest creds to take part in the intellectual exchange that is science, hoping to gain and accepting the risks of loss. This vision will undoubtedly strike many readers as naive. This is where the ability to think like a trader comes in handy.
14.D.5 “Think Like a Trader”
How can we overcome our anxiety about losing creds whenever we put forward a hypothesis? Maybe we can learn a lesson from people whose livelihood depends on risking money for profit: the financial traders on Wall Street who buy and sell shares in companies. If their shares go up in price, they make money; if their shares go down, they lose. In order to stay sane, traders take a broad view of their holdings: they don't invest too much ego or hope in any one investment. Instead, they make lots of investments and expect to win some and lose some (winning a few more, of course). If you think like a trader, you are less likely to suffer from loss aversion, to be more willing to let go of a lost cause, or be less likely to fall for the sunk cost fallacy.
What would thinking like a trader mean for a scientist debating about whether to express a hypothesis? You can put forward multiple hypotheses for any one phenomenon, and you can step back and take the long view of your career.
Generating multiple hypotheses is probably the closest scientists come to actually behaving like a trader. After you've thought of one explanation for your phenomenon, you put it aside and think of another. Besides demanding that you think seriously about the phenomenon, generating multiple hypotheses lets you stay a decent emotional distance away from each of them. When thinking like a trader, you know that some of your ideas will be right and some will be wrong. Having multiple hypotheses is like having a diversified financial portfolio with investments in several different stocks—or even better, in different kinds of entities entirely—if you want to be buffered against sweeping changes. Typically, when the prices of stocks go up, the prices of bonds (municipal bonds, government bonds) go down. If you're a creative scientist you can try to come up with hypotheses that are quite different kinds of explanations. In any case, with multiple hypotheses, you don't have all of your eggs in one basket.
Another way to think like a trader is to take the long view and realize that throughout your career you'll be in the position of advancing hypotheses many times. If they are carefully thought out and rigorously tested, there is every reason to think that some will be correct. Your intellectual investments will undoubtedly pay off, but only if you make them in the first place. The way a trader would.
14.