Franklin and the Art of Small Experiments on Yourself

How do you find out whether a new routine actually works for you? You try it for a fixed, short time, write down a few plain facts as you go, and then read your own notes as honestly as you would read a stranger's. That, in one sentence, is the habit Benjamin Franklin describes in his Autobiography, and it is a good one to borrow. It needs a pencil, a notebook and about two weeks. It does not need willpower of heroic size.

I should say at the start that I am still learning to do this well. I keep a list of my wrong guesses, and a fair few of them come from experiments I ran carelessly. This article is partly what I have worked out from those.

What did Franklin actually do?

In the Autobiography, Franklin describes a project of his younger years. He wanted to improve his own conduct, so he chose thirteen qualities he hoped to practise, among them order, moderation and industry. He did not try to work on all thirteen at once. He gave each one a week of his full attention, then moved to the next. Thirteen weeks make a quarter of a year, so he could go round the whole list four times in a year.

The part I find most useful is the record-keeping. He ruled a small book into a grid, with the days of the week across the top and the qualities down the side. Each time he noticed he had fallen short of the quality he was working on, he put a mark in the right square. His aim was to keep that week's row clean. The other rows he left alone, expecting a few marks to appear.

He did not reach perfection, and he says so himself. He still considered himself a better and happier person for the attempt. I like that he does not pretend otherwise.

One caution: Franklin's grid was about character, and I am not qualified to say anything about that. What I am borrowing is the method, which is to pick one thing, give it a short and fixed period, and keep a written tally. The method works just as well for smaller questions, such as "Do I think better when I get up earlier?" or "Does this way of taking notes help me remember things?"

Why a short written record beats trusting your memory

Suppose I ask you, after two weeks of waking at 6 am, how it went. You will give me an answer, and it will feel like a memory. But memory is not a recording. It is rebuilt each time you recall something, and the rebuilding is influenced by how you feel right now, and by the most recent and most vivid days. If today was a good day, the whole fortnight tends to look rosier.

Here's why this matters. You usually want an experiment to work, because you chose it and put effort in. Wanting a result colours what you remember. A written note made on the day cannot be nudged by that wish, because it was set down before you knew how the trial would turn out.

This is the same reason scientists keep lab notebooks. The notebook is not there because scientists are forgetful or dishonest. It is there because everyone, however careful, remembers in a friendly direction.

So the record is a kind of fairness to your future self. It lets that person look at the facts without being talked round by your hopes.

How to run a two-week trial, step by step

Here is the shape I use. Each step has a reason, so you can change it to suit you.

1. Ask one small question. For example: "If I get up thirty minutes earlier, do I feel better in the afternoon?" One question per trial. If you change your wake-up time, your breakfast and your commute all at once, you will not know which change did what.

2. Decide beforehand what you will record. Choose one or two things that take under a minute each evening. A score from 1 to 5 for afternoon energy is enough, plus a note of the time you actually got up. Deciding in advance matters because otherwise you will be tempted to record whatever supports the story you like.

3. Collect a few days of "before" first. If you only note things during the new routine, you have nothing to compare them with. Five ordinary days of scores will do. This is called a baseline, which just means "how things look when I change nothing".

4. Run the change for the rest of the fortnight. Be gentle and realistic. If the plan is thirty minutes earlier, do not make it two hours.

5. Write down what would change your mind. This is the question a retired bus driver, a former neighbour of mine, taught me to ask. Before you start, write one sentence: "If my afternoon scores are no better than before, I will drop this." Having that line in advance stops you from quietly moving the goalposts later.

6. Read the notes at the end, not in the middle. Looking every day invites you to react to one odd day. Wait for the whole stretch.

A worked example, with the numbers

Imagine I want to know whether getting up thirty minutes earlier lifts my afternoon energy. I score each afternoon from 1 to 5.

Over the five baseline days I score: 3, 2, 3, 3, 2. That adds up to 13, and 13 ÷ 5 gives an average of 2.6.

Then I get up earlier, and over the next five days I score: 4, 3, 3, 4, 2. That adds up to 16, and 16 ÷ 5 gives an average of 3.2.

The average went up by 0.6, which looks encouraging. But look at the spread. Before the change, my scores ranged from 2 to 3. During the trial they ranged from 2 to 4, and one of the trial days was a 2, exactly as low as my worst baseline day. So the new routine's results overlap with the old ones. Day to day I am bouncing around by a point or so anyway.

What can I fairly say? Something like: "The numbers lean in the right direction, but ten days is thin, and the difference is not much bigger than my ordinary ups and downs." That is a perfectly respectable result. It does not prove the early start works, and it does not show it fails. It tells me the question is worth another week, perhaps with a note about how well I slept the night before.

This is the heart of the idea. A record cannot give you certainty, but it can stop you from claiming more than the evidence supports. Many of the things that vary a person's day, such as weather, sleep, a difficult phone call and a good lunch, are mixed into those scores. A small experiment on yourself is a way of looking through that noise, not of removing it.

Common misconception: "If it didn't work, the experiment failed"

It is easy to feel that an experiment has only succeeded if the new routine turns out to be a keeper. But an experiment succeeds when it gives you an honest answer, and "this doesn't suit me" is an honest answer.

My own trials often fizzle. I tried a 5.30 am start and managed nine days, and my notes show that my afternoon scores sank rather than rose, which I had not predicted at all. I tried taking notes by hand in a tidy two-column layout, and by the second week I had stopped, because I was spending more time ruling lines than listening. Neither of these felt like a win at the time.

But each one taught me something specific. The early start taught me that, for me, sleep matters more than the hour on the clock. The note-taking trial taught me that I need a method that is quick enough to survive a busy meeting. Those are small, true facts about how I work, and I did not have them before. Franklin's own words on the matter are that he fell short of his hopes yet gained from the trying. That is a comforting thing to hold on to when a trial wobbles.

Franklin also tells a short story about a man buying an axe from a smith. The man wanted the whole face of the axe polished as bright as the edge, and the smith agreed on condition that the man turned the grindstone. The turning was tiring, and the smith pressed the axe hard against the stone. After a while the man gave up and said he would take the axe as it was, because he thought he liked a speckled axe best. Franklin says he took much the same view for himself, in that he felt tempted to settle for a speckled result rather than chase perfection.

I read that story as a gentle warning against two opposite mistakes. One is demanding perfect polish and giving up when it is not achieved. The other is calling the speckles a choice you always wanted. A written record helps with both, because it shows you what you actually got.

A quick check for understanding

Try these before you start your own trial, and see whether the answers come easily.

• What is my one question? If it has the word "and" in it, split it in two.
• What will I record each evening? If it takes more than a minute, trim it.
• What counts as "not working"? Write the sentence now.
• What else might explain a good or bad day? Jot down anything unusual, such as a poor night's sleep or a visitor, so you can discount it later.

If you can answer all four, you are ready. If you cannot, that is not failure. It simply means the question needs a bit more shaping, and that is worth knowing before you spend a fortnight on it.

Something to wonder about

I have noticed that the most useful parts of my notebooks are often not the scores but the odd little remarks I scribbled beside them: "slept badly", "rained all day", "lunch late". Looking back, those remarks explain more than the numbers do.

So here is a question I am still turning over. What is one thing in your own days that you have never written down, because it seemed too ordinary to count? And what might you see in it after two weeks of simply looking?

Thank you for reading. If you run a trial, I hope it fizzles interestingly.