Week 10 Lab: The Subject Has Changed
This week
We run week 5's experiment again, on the same person, and get a different answer.
Before you run this
- When to stop
- The same cap as week 5 applies, and more strictly: this is a repeat, not an escalation. Stop at four days, or sooner if the removal phase stops feeling like an experiment.
- What this is not
- Running a protocol twice does not license running it harder. Nothing about the second pass should be longer, stricter or more ambitious than the first, and anyone who took the alternative in week 5 takes it again now.
- If this one isn't for you
- Re-run whichever inert version you used in week 5. The comparison you need is between your own two attempts, so any variable works as long as it is the same one.
What you bring to the lab
The lesson below is how you get there. The lab is where you show it.
- you re-run week 5's protocol without modification
- you report both results side by side, not averaged
- you offer at least two explanations for the difference, one of which is not a mistake
Do week 5 again. Same variable, same three phases, same measure, same length.
Then put the two results next to each other.
The answer will have moved#
For most of you the second result will differ from the first, sometimes substantially, occasionally in the opposite direction. There is a strong instinct about what this means: one of the two runs was done wrong.
Sometimes that is true, and looking for the mistake is the right first move. But after you have looked and found nothing, there is a second explanation, and it is the one this course has been building towards.
You are non-stationary. The thing being measured changed between March and May. Your sleep changed, your workload changed, the daylight changed, and you have spent nine weeks learning to watch yourself, which by week 4’s logic has changed you too.
Why this breaks the usual promise#
Almost all advice about daily life assumes a stable subject. “Work in twenty-five minute blocks” is a sentence about a person who stays the same. So is every result you have produced this semester.
If the subject drifts, then a finding has a shelf life. Not because your method was poor, but because the thing your method measured has moved on. This is why the course cannot end with a list of what works for you: any specific answer expires, and the only durable thing is the ability to run the measurement again.
The long version#
Santorio Santorio, in Venice, built himself a weighing chair and spent something close to thirty years on it: weighing what he ate, what he excreted, and himself, trying to account for the difference. He published in 1614.
Thirty years is the part to notice. He was not running an experiment; he was maintaining an instrument, on a subject he knew would keep changing. That is much closer to what this course is actually teaching than any single result in it.
In the lab#
Both numbers on the board, unaveraged. Then we ask everyone the same question: which of your two answers would you act on, and what would you have to believe about yourself to choose?
Evidence on file
- 800 people blinding not possible
Continuous glucose monitoring across 46,898 meals found that responses to the same standardised meal varied widely from person to person, which the authors took as a limit on universal dietary advice.
Zeevi D et al., Personalized Nutrition by Prediction of Glycemic Responses, Cell 2015;163(5):1079-1094- History, not measured
Santorio Santorio spent roughly thirty years weighing himself, his food and his waste on a balance of his own construction, looking for what he could not account for. He published the results as a set of aphorisms in 1614.
Santorio Santorio, Ars de statica medicina, Venice 1614
Teaching team
- Tobias Veck, tutor