source：bbc | by Tom Stafford
诚实或为上策，然则谎言亦有其优势，哪怕是当我们自欺欺人的时候。很多研究证实了那些善于自欺欺人的人在球类运动和商业中更容易获得成功。他们或许比那些对自己诚实的人更快乐。但是相信自己撒的谎会有坏处吗？
耶鲁大学的Zoe Chance独创了一个研究来看看人们在测试上撒谎会带来什么后果。
Chance和他的同事设计了一个实验让一些学生去回答一些跟智力和常识有关的问题。一半的学生会给一份“不小心”把答案印在了最下方的试卷。这就意味着这些学生必须抵抗住去检查答案或者修改答案让自己获得更高的分数的诱惑。
正如你所预期的，有些学生确实忍不住作弊了。所以总体来看拿到答案的那个组的学生的成绩会比另一个组的好，哪怕这两个组的学生都是从同一个学校里面随机抽选分组的，所以他们的能力水平实际上是相似的。（我们不确定是谁撒了谎，可能一些有答案的学生在没有拿到答案的情况下也能获得高分，这就意味着这个组的平均表现既可能是有个别人很聪明的原因，也可能是有答案在手的原因。）
Chance的研究的一个重要问题就是：在“撒谎组”的人知道他们的好成绩是靠答案得来的吗？还是把自己的好成绩归因为个人的聪明才智呢？
研究者用来解答这个问题的方法就是让学生去预测他们接下来的测试成绩。学生被允许快速地浏览一下接下来要做的测试，这样学生就会发现接下来的测试跟之前的是一样的，而且试卷的下方没有任何答案。研究者预计那些作弊的学生得知第二轮测试没有答案之后，会把自己的成绩预期降低。
但是并没有。学生的自欺欺人行为占了上风。那些之前看到过答案的学生总体上都认为接下来的测试他们会做得更好——他们觉得自己在IQ测试上至少可以提高10分。当然了，第二轮测完以后，他们的成绩远低于他们的预期。
研究者设计了另外一个实验证实了这个现象确实是因为撒谎者对自己能力的自信膨胀。在这个实验里面，研究者给学生一笔现金奖励他们预测自己第二轮成绩的准确性。毫无疑问的是，那些给予了作弊机会的学生还是过高地估计自己的能力，导致相比于另一个组的学生少了20%的奖金。
Chance的实验隐含着这样一个事实——我们都欺骗自己，觉得自己比真实的自己聪明。这对我们来说既有好的方面——有利于提高自信、满足感、更容易获得别人的信任，也有不好的方面。当情境发生了变化，我们真的需要去准确预测自己的水平的时候，相信自己的水平比实际的高是要付出代价的。
这种自欺欺人的行为产生了代价的现象有一些很有趣的寓意。从道德层面看，多数人都会说自欺欺人是不对的。但是不管自欺欺人的行为可不可取，我们期望所有人都能一定程度上欺骗自己（因为这是有好处的），但是又希望欺骗自己有个限度（因为这是要付出代价的）。
自欺欺人不是程度越大就越好的，虽然大多数情况下自欺欺人的益处超出了它的害处。我们或许都会在一定程度上欺骗自己。挺讽刺的是，因为是欺骗自己，所以我们并不知道自己有多经常会这样干。
译者有话说：我对这个实验的设计存在一个疑惑，当我作弊以后，让我预测在没有答案的情况下再回答一遍同样的题目，我的预测比实际分数高，未必是因为我对自己的智商有信心，有可能是我对自己的记忆力有信心——我知道自己没那么聪明，但是我觉得我记得住那些我写过的答案。错误估计自己的记忆力可以算是一个自欺欺人的行为吗？
原文：
Honesty may be the best policy, but lying has its merits – even when we are deceiving ourselves. Numerous studies have shown that those who are practised in the art of self-deception might be more successful in the spheres of sport and business. They might even be happier than people who are always true to themselves. But is there ever a downside to believing our own lies?
An ingenious study by Zoe Chance of Yale University tested the idea, by watching what happens when people cheat on tests.
Chance and colleagues ran experiments which involved asking students to answer IQ and general knowledge questions. Half the participants were given a copy of the test paper which had – apparently in error – been printed with the answers listed at the bottom. This meant they had to resist the temptation to check or improve their answers against the real answers as they went along.
As you"d expect, some of these participants couldn’t help but cheat. Collectively, the group that had access to the answers performed better on the tests than participants who didn"t – even though both groups of participants were selected at random from students at the same university, so were, on average, of similar ability. (We can"t know for sure who was cheating – probably some of the people who had answers would have got high scores even without the answers – but it means that the average performance in the group was partly down to individual smarts, and partly down to having the answers at hand.)
The crucial question for Chance"s research was this: did people in the “cheater” group know that they"d been relying on the answers? Or did they attribute their success in the tests solely to their own intelligence?
The way the researchers tested this was to ask the students to predict how well they"d do on a follow-up test. They were allowed to quickly glance over the second test sheet so that they could see that it involved the same kind of questions – and, importantly, that no answers had been mistakenly been printed at the bottom this time. The researchers reasoned that if the students who had cheated realised that cheating wasn’t an option the second time around, they should predict they wouldn"t do as well on this second test.
Not so. Self-deception won the day. The people who"d had access to the answers predicted, on average, that they"d get higher scores on the follow-up – equivalent to giving them something like a 10-point IQ boost. When tested, of course, they scored far lower.
The researchers ran another experiment to check that the effect was really due to the cheaters’ inflated belief in their own abilities. In this experiment, students were offered a cash reward for accurately predicting their scores on the second test. Sure enough, those who had been given the opportunity to cheat overestimated their ability and lost out – earning 20% less than the other students.
The implication is that people in Chance"s experiment – people very much like you and me – had tricked themselves into believing they were smarter than they were. There may be benefits from doing this – confidence, satisfaction, or more easily gaining the trust of others – but there are also certainly disadvantages. Whenever circumstances change and you need to accurately predict how well you"ll do, it can cost to believe you"re better than you are.
That self-deception has its costs has some interesting implications. Morally, most of us would say that self-deception is wrong. But aside from whether self-deception is undesirable, we should expect it to be present in all of us to some degree (because of the benefits), but to be limited as well (because of the costs).
Self-deception isn"t something that is always better in larger doses – there must be an amount of it for which the benefits outweigh the costs, most of the time. We"re probably all self-deceiving to some degree. The irony being, because it is self-deception, we can"t know how often.
