What My Dissertation Was Supposed To Be About

my dissertation was supposed to be about food.

The pattern

I registered a question about nutrition and got an answer about perception. Seventy-four adults, a model that ignored every supplement in it, and the age comparison I have not stopped thinking about since 2022 — including the part where I argued against my own finding.

The mechanism behind it

My master's dissertation was supposed to be about food.

Eating habits, nutrient deficiency, supplements — what people put into their bodies to manage stress and anxiety. That was the registered question, and for months it was the one I believed in. Seventy-four adults, nineteen to sixty-three: what they ate, what they took, the state of their health, and how anxious they had been over the previous two weeks.

The supplements predicted nothing. Neither did whether the person was, medically speaking, healthy.

One variable in the model predicted their stress, and it was how they read their own life. Twelve questions about what someone believed about their life and their health. That score moved, and the anxiety moved with it.

The finding I still turn over is the age comparison. The older participants had more health problems and less anxiety than the physically healthy young ones. Worse bodies, quieter minds.

Seventy-four people measured at a single moment settles nothing about which direction the relationship runs, and I wrote a full section arguing against my own result. That section is still the most honest thing I have put my name on.

But I have built everything since on the possibility that it runs the way the data pointed — that a life is not made well by what gets added to it, and that the thing worth defending is a person's reading of their own days.

Which is why I now build systems that take the work off people rather than programmes that ask more of them.

What to do with it

eating habits, deficiency, what people take to feel less wrecked. that was the question i registered, and for months it was the question i believed in. seventy-four adults, nineteen to sixty-three, all of them telling me what they ate, what they took, what their body was doing, and how anxious they had been for the last two weeks.

the supplements predicted nothing. neither did whether the person was actually, medically healthy.

one thing in the model predicted their stress, and it was how they read their own life. twelve questions about what they believed about their life and their health. that score moved and the anxiety moved with it.

the part i still turn over: the older group had more health problems and less anxiety than the physically healthy young ones. worse bodies, quieter minds.

seventy-four people at one moment in time settles nothing about which direction it runs. i wrote a whole section arguing against my own finding, and it is still the most honest thing i have put my name on.

but i have built everything since on the possibility that it runs the way the data pointed. that a life is not made well by what you add to it, and that the thing worth protecting is a person's reading of their own days.

that is what all of this is. protect the reading. give the rest to the machines.

comment PROTOCOL and i'll send you what the dissertation actually found — free.

The bigger picture

my dissertation was supposed to be about food. eating habits, deficiency, supplements — what people take to handle stress.

seventy-four adults, nineteen to sixty-three, telling me what they ate, what they took, what their body was doing, and how anxious they had been for two weeks.

the supplements predicted nothing. neither did whether they were actually healthy.

the thing that predicted their stress was how they read their own life. and the older group, carrying more health problems, was less anxious than the physically healthy young ones.

seventy-four people at one moment settles nothing about direction — i wrote a section saying exactly that. i have still built everything since on the possibility that it runs that way.