Labs
The weekly graded labs — twelve short, domain-specific deliverables, marked as a set rather than individually agonised over
Due: 28 May 2027 · Weight: 20%
Twelve labs, one mark. Grading each of the twelve on its own would either force twelve separate high-stakes deadlines onto a course whose entire argument is that the last mile is where things break, or it would force us to pretend a rough week doesn’t happen. Neither is honest, so the scheme instead does what a full semester should: it tolerates exactly two bad weeks — sick, swamped, or just a lab you didn’t click with — without asking you to explain yourself, and it counts the other ten in full.
What you submit, each week#
Every lab page under the labs tab states its own artefact — a diary entry, a calculation with justification, a zine, a pitch recording, a debate script — and its own group size. Submit what that week’s page asks for, by that week’s due date, through the same course drop-box used for the two larger assessments.
Why “best 10, drop 2” instead of a weighted average#
A weighted average of twelve scores quietly punishes the two weeks you were least prepared for exactly as hard as it rewards the ten you were most prepared for — which is precisely the kind of unforgiving, no-recovery curve week 5 spends its own lecture arguing is the wrong model for low-stakes, frequent work. Dropping the worst two outright is the one grading scheme that matches this course’s own argument: it puts the “last mile” pressure where it belongs, on actually turning up and doing the work most weeks, and removes it from the one place it teaches nothing, which is punishing a single bad week you can’t retroactively fix.
The spec
The brief poses the problem and leaves room for a response. The spec is the fixed part: what a marker checks. Some lines can be checked mechanically; the rest call for judgement.
- a submission exists for at least 10 of the 12 labs, each meeting that lab's own spec
- each submission was made by that lab's own due date (5pm the Friday of its teaching week) — a late submission is treated the same as a missed one, and simply becomes one of the two you can drop
- group submissions name every group member, and every named member contributed to the artefact submitted
How it is marked
Each lab is marked pass/needs-work against that week's own spec, then the twelve marks are combined by keeping your best 10 and dropping your worst 2 — including any you miss outright. There is no separate extension process for a missed or weak lab; the drop is the extension. A lab that scores "needs work" still counts as one of the ten if it is your tenth best, so there is no benefit to redoing a submitted lab.