Assessment

Four pieces of work, what each one is actually testing, and why there is no exam.

There is no exam, and the four tasks below are not four samples of the same skill. Each one is aimed at a different way this course can fail to land.

The calibrated quizzes (15%) test whether you know what you know. You declare a confidence with every answer and a confident error costs six marks where a hedged one costs nothing, which means the scoring measures the gap between how right you were and how right you thought you were. That gap is the subject of week 12, and this is it arriving in week 3 in your own results.

News and Views (25%) tests whether you can get to a primary source and report honestly on what it says. You take a claim that circulates, find where it came from, and characterise the difference — including the case where there is no difference, which marks well and is rarer than you would expect.

The position statement and revision (20%) tests whether anything moved. You write down what you think intelligence is in week 2, before the reading, and revise it in week 11 against a document timestamped before the course could reach it. The memo is what gets marked, and both over-revision and under-revision fail.

The cognitive profile (40%) tests whether you can use the model rather than recite it. One real organism, nine fixed sections, closing on what that organism lets you see about human cognition that you cannot see from inside it.

Two things are worth knowing before you plan your semester. The week 2 position statement is a hurdle: it carries no marks, it cannot be extended past the end of week 3, and without it you cannot submit the revision that does carry marks. And the profile is decided by section 7 — the failures with no possible fix — so bring two organism options to Rosalind before the end of week 9 rather than committing to one on your own.

Late work, extensions, integrity and tool use are on the policies page, and every brief below links back to it rather than restating it.