Resources

Most people read difficult material exactly once, slowly, and hope. A better approach is three deliberate passes with different goals.

The first pass is structural and should take minutes: headings, figures, the first line of each section, the conclusion. You are building a map, not learning anything.

The second pass is the real read, and it is where you should allow yourself to be confused. Mark the confusions rather than resolving them immediately — most resolve themselves two pages later.

The third pass is reconstruction. Close the material and write the argument in your own words. Whatever you cannot write is what you did not understand, and it is a much shorter list than it felt like during pass two.

Topic — Study practice

Describe before you judge. Say what is actually on the page, in neutral language, until the maker agrees with your description.

Separate three things every time: what was intended, what was executed, and what the effect is on a person meeting it cold.

Rank your observations by cost to fix and cost of leaving. Lead with the expensive-to-leave items.

End with one specific thing to do next. Critique without a next action is entertainment.

Topic — Critique

Making and finishing use different faculties. Making rewards openness; finishing rewards closure, and the two are genuinely uncomfortable together.

The practical consequence is that finishing must be scheduled rather than awaited. A piece of work that is ninety per cent complete is not nearly finished; it is at the point where the remaining work is least pleasant.

Define done before you begin. Not perfectly — a paragraph is enough — but concretely, so that you can be honest with yourself later about whether the extra week was work or avoidance.

The best test remains the crude one: would you rather have this version in the world next week, or a better version in three months that may never arrive?

Topic — Creative practice

Open with consent, purpose and permission to stop. Two minutes, every time, no exceptions.

Warm up with recent, concrete behaviour: “Walk me through the last time you…”. Never open with preferences.

Spend the middle third on the specific decision you are investigating. Follow surprise; abandon your own script when something more interesting appears.

Close by asking what you should have asked. It is the single highest-yield question in the guide.

Topic — Research

Measure first. Somewhere between sixty and seventy-five characters per line for continuous text; everything else is negotiable.

Leading follows measure. Longer lines need more space between them, and the relationship is closer to linear than most people expect.

Set your scale before you set anything else, and keep it short. Three or four sizes will carry an entire publication.

Space is punctuation. Before you reach for a rule, a box or a colour, try changing the space.

Topic — Typography

Start with published post-mortems from organisations that write honest ones. Read three from the same company across several years; the pattern is more instructive than any single account.

Move to the human-factors literature next. Most technical failures are decision failures under uncertainty, and that field has been studying those for far longer than software has existed.

Add one queueing theory text. You do not need the mathematics, but you do need the shape of the curve.

Finish with anything on organisational drift. The most expensive failures are the ones a system was slowly reconfigured to permit.

Topic — Engineering

The commonest structural problem is that the essay begins one section too early. Delete the first section and read again; it is usually better.

The commonest line problem is the hedge. “Somewhat”, “arguably”, “it could be said that” — remove them and see whether you still believe the sentence.

Long sentences are not the problem; unvaried sentences are. Read your draft aloud and mark where you run out of breath.

If a paragraph resists editing, it is usually because it is doing two jobs. Split it and the problem tends to dissolve.

Topic — Writing

Never ask what someone would pay. Ask what they currently spend on the problem, in money and in hours.

Anchor on the alternative, not on your costs. Your costs are your business; the alternative is theirs.

Keep one sheet with three columns: value metric, unit price, and the customer’s own arithmetic. If the third column does not work, nothing else matters.

Discounts are measurements. Track them, because the shape of your discounting is the most honest description of your pricing power you will ever get.

Topic — Business

Unplug the mouse. Tab through the whole thing. Note every place focus disappears, jumps or gets trapped — this alone finds more than any automated tool.

Then read the heading outline. If it does not describe the page, the page has a structure problem, not a styling problem.

Check the forms next: label association, error text, and whether an error moves focus anywhere useful.

Only then run a contrast check. It is the easiest to fix and the least likely to be the real barrier.

Topic — Accessibility

Assignments are written responses assessed against the brief rather than against a model answer. Length is a constraint, not a target.

Quizzes exist to check that a concept survived the lesson. They are low-stakes, retryable and explained after submission.

A course is complete when every lesson has been read, every quiz passed and every assignment submitted.

Completion produces a demo certificate. It is a record of work done inside this template, not an accredited credential.

Topic — Study practice