Idris Fenn
Runs the measurement-heavy Nap Labs, and is the reason you will be asked for a median rather than a mean.
- Role
- Tutor
- Affiliation
- Slop University School of Invented Disciplines
- idris.fenn@slop.university
- Contact
- Ask in the Nap Lab — that is what the time is for. Email if something cannot wait until the next one, or if it is not a question you want to ask in front of the room.
My background is in psychophysiology, mostly the unglamorous part: whether the instrument measured what somebody claimed it measured. I spent three years on the validation side of consumer sleep devices, which is why week 10 exists and why I am harder on your tracker than you will initially think is fair.
I run the labs where something is measured on you — weeks 2, 3, 8 and 10 — and I share weeks 1, 4, 6, 7, 11 and 12. If you find yourself irritated that a nine-point scale you filled in by hand is being taken more seriously than a device that cost two hundred dollars, that irritation is the intended curriculum, and week 10 is where it gets resolved one way or the other.
The two things I will keep saying#
Take the median. Ten reaction-time trials, and you report the middle one, not the average. One distracted trial should not be allowed to speak for the set, and a semester of means is a semester of data that cannot support anything.
Record the uncertainty instead of resolving it. “I am not sure whether I slept” is a real observation and most of you will have it in week 2. Writing down a confident guess destroys the most interesting thing about your own data, and it is precisely the failure mode that week 10 shows the devices making at scale.
What to come to me about#
Anything to do with the measurements, the Nap Log format, or a number in your own data that looks wrong. That last one is worth an email on its own — about half the time something has genuinely gone wrong with the procedure, and the other half it is a real result the student was about to delete.
If a lab’s requirements are difficult for you — deferring a meal, taking caffeine, napping in a room with other people — tell me beforehand. Every lab has an alternative condition, no explanation is needed, and the control conditions are scientifically useful rather than a consolation.