The decision

Here's where the argument has got to. The doubt and the step back are separate, and the second is a habit. The habit pays in quick relief and costs you evidence, practice, hours and other people's picture of you. The feeling of doubt tracks exposure and caring more than ability. Other people's calm is mostly surface. Extra preparation and extra checking soothe the feeling without settling anything. Not trying protects a possibility and nothing else. Self-doubt isn't humility. Your record has been kept unfairly. And confidence comes after doing, so waiting for it doesn't work.

If most of that seems right to you, then there's a decision to make, and I'd put it like this.

A doubt that names something I can check gets checked, once. A doubt that only names me doesn't get to decide what I do.

Notice what the rule leaves alone. It doesn't ask you to stop feeling doubt, which you couldn't do anyway. It doesn't ask you to feel confident, or to think well of yourself on demand. The doubt can come along. It can sit in the meeting with you and say its usual sentence. What changes is that it loses its vote on what you do next.

The rest of this chapter is method. None of it is complicated.

Learn your own step backs. They're quick, and the first job is to catch them. You'll have your usual ones. It may be silence in meetings, or the fifth reread, or the "quick check" with a colleague, or "you choose", or the advert you look at and don't answer. Write down your two or three most common. For the first week, you're only trying to notice them as they happen, or soon afterwards. Catching one an hour later still counts.

Ask "what exactly?" When you catch the doubt, put the question from chapter 3 to it. If it names something you can check, check it. If it says something about you, note that it's the general doubt, and carry on with what you were about to do.

Check once, with a limit. A specific doubt gets one proper check. Look up the March figure, find out whether it includes refunds, and move on. If the same doubt comes back after it's been answered, it has changed into the other kind, and more checking won't help. For bigger pieces of work, set the preparation in advance as in chapter 5, by the job. Two run-throughs, three figures checked, questions prepared, stop.

Go early. This matters more than it looks. In a meeting, the doubt grows with every minute you wait, and the longer you've been silent the bigger a deal it feels to break the silence. So say something in the first ten minutes, even if it's only a question. After that you're a person who's speaking in this meeting, and the second contribution is far easier. The same goes for sending things. Decide beforehand which draft you'll send, the second, say, and send it when you reach it. Put the application in on the day you decide to apply, or the next morning. Don't leave it for the week before the deadline.

Drop the apology in front. I mean the little run-up. "This is probably a stupid question, but." "I'm sure you've already thought of this." "I'm no expert." It feels like politeness, and like insurance in case you're wrong. What it does is tell people how to take what follows, and they'll mostly take you at your word. Chapter 4 mentioned that groups give weight to people who sound sure. You don't need to sound surer than you are. Just stop sounding less sure than you are. "Could it be the form on mobile?" is a complete sentence. If you truly are uncertain, say what you're uncertain about. "I've only tested it on my own phone" is useful information, and "this is probably nothing" isn't.

Write down the prediction, then the result. This is the most important of the seven, because it's how the evidence finally gets in. Before you do the thing, write one line saying what you expect. Make it specific enough to be wrong. "They'll say it's already been checked and I'll look foolish." Afterwards, write one line saying what happened. "Dev checked, it was broken, fixed by lunch." Keep these lines in the same notebook or file as the record from chapter 8. Therapists use the same technique under the name behavioural experiments. One review of fourteen clinical studies found some evidence that it works better than simply facing the feared situation, though the authors said the studies were too limited for firm conclusions. For our purposes it does one specific job. In chapter 2, the prediction survived because it was never tested and never even stated. This puts it on paper next to what happened. Some predictions will come true, and the next chapter deals with that. My expectation is that most won't, and that the ones that do will turn out milder than forecast.

Bank the good results. Add them to the record from chapter 8, in plain words. When someone says something good about your work, say "thank you" and write it down later. After a few months, read the record through before anything that frightens you. You'll be reading facts about what you've already done.

Start small, and I mean that as method, not as kindness. Chapter 10 explained why early successes matter. Choose situations where a bad result would cost little. It might be a question in a small meeting, or an email sent on the second draft, or a preference stated at home. The smallest real version you worked out at the end of chapter 10 belongs on the list too. Do those for a few weeks before you take on the big presentation. But do start this week, because the first prediction you write down and test is worth more than another chapter of reading.

It helps to decide in advance exactly what you'll do at the moment you know is coming. Plans of the form "when this happens, I'll do that" have been tested in a lot of studies, and across them people with such a plan were noticeably more likely to follow through than people with only a good intention. The reason seems to be that the decision is already made, so you don't have to make it while the doubt is talking. Write two or three, fitted to your own step backs. For example:

• When the team meeting starts, I'll ask or say one thing in the first ten minutes.
• When I've finished the second draft of an email, I'll send it.
• When someone asks where I'd like to eat, I'll name a place.

What should you expect? Not calm, at first. The first dozen times, your body will behave exactly as it always has. If the argument of this book is right, the feeling follows later, as the record fills and the predictions keep failing. I can't promise you how quickly, and for some situations it may never go completely. Many people who speak in public for a living are nervous every time. They've stopped treating it as a reason not to speak.