Practice thinking in real situations.

Students compare speakers in a climate dispute using a position spectrum and a perspective matrix, distinguishing their views, relevant expertise and evidence.

Students check the source, identities, story details and historical context of a photograph, identifying which claims the evidence supports.

Students compare CRAAP and CRRAAB to assess both source reliability and whether a source can answer their research question.

Students use FLICC to examine two climate videos with different positions, checking both for selective evidence and explaining their judgments.

Students reassess the value and limitations of a 1915 British recruitment poster as they learn about its creator, publisher and archival context.

Students distinguish facts, opinions and value judgments, then choose appropriate checks: scope and samples, reasons, or evaluative criteria.

Students examine a proposed casino on Boracay through tourism income, environmental capacity, governance and community interests, comparing priorities and trade-offs.

Students identify specific claims in a statement about short-term rentals in Venice, check multiple sources and use the evidence in writing.

Students review how they reached a judgment, checking premises, missing reasoning steps and perspectives they have not considered.

Before deciding that two papers conflict, students compare their subjects, data, scope and measurement methods.

Students assess sources about Columbus by examining authorship, audience, purpose and context, then compare the questions each source can help answer.

A post claims NASA found the Earth greening, and that China and India are why. Does it go in the essay? Read laterally back to the primary source, see what the retelling lost on the way, and write it up as an essay that takes a position and can defend it.

Students compare three responses to overtourism in Venice, using documentary evidence to recommend a policy and explain its trade-offs.

Students plan an academic essay, develop paragraphs, check citations and describe their use of AI as they review their draft.

Using Haltinner and Sarathchandra's paper on types of climate scepticism, the reading is not “what did they find” but “how sure are they of each finding”. Suggests, indicates, demonstrates: the certainty is in the wording.

Students examine axes, comparison groups, denominators and time intervals to understand how data presentation can influence interpretation.

Students choose methods and charts based on a research question, interpret the results and check whether each analysis addresses that question.

Finding sources is not something you do once, when choosing a topic. It runs through drafting, writing and revision, and each stage wants a different kind: early on to understand the ground, later to build evidence, at the end to check.

Using Tesla’s impact report, students formulate checkable questions and compare evidence across environmental, social and governance issues.

Students practise checking reports, images and videos using original sources, provenance, metadata and contextual evidence about time and place.

“The IOC recognises esports as sport” and “Germany refuses to call esports sport” can both be true, because they are not saying the same thing. One body is classifying for administration, another is recognising an organisation, a third merely added an event to its own games.

Method is not the formatted paragraph in the paper; it is where a conclusion's credibility comes from. The course pushes it from “what tools I used” to why the evidence was gathered this way, how it was analysed, and how far the conclusion can reach.

The course does not rule on who is right about the climate. It handles something else: the judgment that surfaces after you watch a film — how much of it came from the evidence, and how much from the lighting, the framing and the score.

Students check whether a concession addresses relevant counterevidence and whether that evidence calls for qualifications or changes to their argument.

It starts with a sentence that goes down easily, lets the student be persuaded for a moment, then turns back on it: was I convinced by the evidence, or pushed there by numbers, fear, authority, a switched term, or a false choice?

An experiment first: the same crime data, half the class told crime is a beast, half told it is a virus. The solutions split visibly — capture and punish on one side, diagnose and treat on the other. Watching it happen to yourself beats being told ten times that language steers judgment.

Someone sends you a video: CO₂ is not a pollutant, fossil fuels took life expectancy from 30 to 73, so cutting emissions holds civilisation back. It has data, charts and a certain plausibility. So follow the money, and find out who paid for the argument.

“I want to research AI's effect on employment” is not a research question; it is a topic. This course walks the hardest stretch of the road — narrowing, finding an angle, and landing on a question that material can actually answer.

Students divide a long-term inquiry into stages and weekly tasks, track decisions, dependencies and risks, and adjust their plan as work progresses.

Students tend to think a better prompt is a longer one — more role, more tone, more format. Here they ask about something they actually have to do: predict the answer, read it, then open the sources and check. The finding that matters is not that AI errs, but why fluent, confident prose quietly takes your judgment away.

IFS translated into a cast of inner characters a teenager can use. No test, no diagnosis, no lecture about screen time. It does something more basic: it rewrites “I am anxious” into “a part of me is anxious” — and that one rewrite moves you from the stage back into the audience.

Built on Kegan's subject-object shift and immunity to change, but deliberately without stage rankings, self-tests or verdicts on a person. It helps a student see one thing only: whether the change she says she wants and the thing she is actually protecting are the same.

A short video about Anthropic puts training data, copyright, labour and energy in one frame, and it is easy to land straight on “AI companies sacrifice a lot of people”. The course does not ask you to take a side first. It starts from a simpler question: who is actually paying?