Why permission worksOn Failure

Perfectionism comes in two parts that pull in opposite directions. Perfectionistic strivings are the high standards themselves. Perfectionistic concerns are the worry about falling short of them: fear of mistakes, of being judged, of the gap between what you meant and what you managed.

A 2017 meta-analysis pooled studies of about 10,000 people. Worry about falling short went with more procrastination. High standards on their own went with less. The associations are small to medium and correlational, so they describe a pattern rather than prove a cause, but the pattern is the one this site is built on. It is not wanting to do well that stops people starting. It is being afraid of doing badly.

In a series of experiments, people were asked questions they could not possibly answer, then shown the answer. Others simply read the question and answer together. The people who had tried and failed remembered more. The failed attempt was not wasted effort; it made the correct answer stick.

A 2017 review of this research in the Annual Review of Psychology reached the same conclusion across many studies, with one condition that matters: the error has to be followed by the correction. Getting things wrong helps when you then find out why.

Error management training tells learners outright that mistakes are expected and useful, and lets them explore. A meta-analysis of 24 studies and more than 2,000 people found it did better than training designed to keep people from making errors, and did best of all when people had to use the skill on a new problem later.

Classrooms show something similar. A 2021 meta-analysis of 53 studies compared letting students attempt a problem before being taught how to solve it, which usually means failing at it first, against teaching first and practising after. Attempting first came out ahead, by a moderate margin. It is not universal: for the youngest learners and for general skills, the trend ran the other way.

The fear is that forgiving yourself for a failure lets you off the hook. Across four experiments, it did the opposite. After failing a difficult test, people who had been prompted to treat the failure with self-compassion spent longer studying for the next one than people in the comparison groups, one of which had been given a boost to their self-esteem instead.

That is the mechanism the pledge is reaching for. Permission to fail is not permission to stop. It removes the part of failing that makes people avoid trying again.

Students who did a brief writing exercise about their worries immediately before an important exam scored better than those who did not, and the students who were habitually anxious about tests gained the most. The researchers' reasoning was that worrying about an exam is part of what drags performance down, and writing the worries out beforehand eases it.

This is not the same exercise as signing a pledge, and no one has tested the pledge. But it is a close cousin: take the thing that is stopping you, write it down by hand, and deal with it before the moment that counts.

In a study of 51 work teams, Amy Edmondson found that the teams whose members felt safe to take interpersonal risks, the kind involved in admitting a mistake or asking for help, did more of the things that make a team learn. That learning, in turn, was what connected the feeling of safety to better performance.

Google later studied more than 180 of its own teams to find out what made the effective ones effective. Of the five dynamics it found, it put the same one first.

Creativity researchers call it the equal-odds rule, after work by the psychologist Dean Keith Simonton on the careers of composers and scientists: the more ideas a person produces, the more good ones they tend to produce.

A 2015 study tested it directly: people who came up with more ideas on a creative task also came up with better ones, as rated by independent judges. A 2019 study added a caveat worth keeping, that the effect is stronger in some people than in others. The practical reading is still the one the pledge makes. You cannot pick the good attempt in advance, so the strategy is to make more of them, knowing most will not be the good one.