Tension is rising in the Harford household as exams approach and we try to persuade Miss Harford Sr to relax, and Miss Harford Jr to be slightly less relaxed. I’m sure many readers have vivid memories of the exam room, recent or otherwise. But here’s a question about exam technique that suggests a much wider lesson. In a multiple-choice test, you sometimes write down an answer and then have second thoughts. Is it wise to stay with your first instincts, or better to switch?
Most people would advise that the initial answer is usually better than the doubt-plagued second guess. Three-quarters of students think so, according to various surveys over the years. College instructors think so too, by a majority of 55 to 16 per cent. The 2000 edition of Barron’s How to Prepare for the GRE Test is very clear that students should be wary of switching. “Experience indicates that many students who change answers change to the wrong answer.”
This confidence would be reassuring, were it not utterly erroneous. Researchers have been studying this question since the 1920s. They have overwhelmingly concluded both that individual answer changes are more likely to be from wrong to right, and that students who change their answers tend to improve their scores. This gap between perception and reality is stark enough to have earned a name: the “first-instinct fallacy”. No doubt our first instincts are often right, but when we start to have second thoughts, the second thoughts are usually occurring for a reason. It is better to switch. So why don’t we?
Justin Kruger, a psychologist at New York University, has been studying this question. (Prof Kruger is more famous as co-discoverer of the Dunning-Kruger effect: people who are incompetent are too incompetent to realise how incompetent they are.) With his colleagues Derrick Wirtz and Dale Miller he replicated the longstanding findings that college students believe you should trust your first answer in a multiple choice question, and yet that switching to a second answer tends to improve your grades. Then the trio started to explore why.
In one study they showed subjects video based on the TV show Who Wants To Be A Millionaire? and asked them to imagine that they were watching a teammate play, accumulating cash for the team whenever he or she gave the correct answer. Some subjects were shown teammates who always switched on 50/50 questions, while others were shown teammates who always stuck with their first instinct. In the study, both strategies produced identical results, yet subjects watching a switching teammate were more frustrated and critical and had a good memory for the errors.
Another study by Prof Kruger and his colleagues showed that we also have a warped recollection of our own errors in multiple choice tests. We have a rosy memory of sticking to our first instincts, forgetting the failures and exaggerating the successes. We vividly recall switching to the wrong answer and overestimate how often we did so. In short, we remember sticking as having been the best tactic, when in fact switching was better. No wonder the Barron’s guide remarks that “experience” tells us switching is a bad idea. Our own experiences do indeed tell us that, but only because we misremember the lessons of previous switches.
If you — or a loved one — are about to enter exam season, perhaps this evidence-based strategy will be of use. But it’s hard not to see a broader lesson. How often in life do we make a choice and then stick to it despite mounting doubts?
In politics, such questions are aggravated by questions of partisanship and pride. Nobody wants to admit that they were wrong in the face of jeers from those on the opposite side of the political fence. The U-turn is one of the greatest sins in politics, if only because it is so easy to criticise. Either you were wrong before or you are wrong now.
But even in everyday life, we find ourselves clinging to bad choices. Steven Levitt, the co-author of Freakonomics, once conducted a study in which people hesitating over big choices — to leave a spouse, to adopt a child, to quit a job, to start a business — agreed to be guided by a coin toss. Those who had been nudged to act ended up being happier several months on than those who had been nudged to stick with the status quo.
We are prone to cling tightly to the devil we know. The likely explanation is that we are seeking to minimise regret. Forget the old cliché about regretting the things you didn’t do more than the things you did: we regret misfiring action more than misfiring inaction. We starkly remember the times we changed things for the worse, and we more easily forget the times when we failed to change things for the better. It’s not that impulsive action is always the best option, any more than your first answer on a test is always wrong. Instead, the lesson is that if you are hesitating over whether to leave things as they are, you probably needed to make a change some time ago.
Written for and first published in the Financial Times on 10 May 2019.