the knowledge platform

flashcard design rules: make cards that actually work

turn notes into retrieval prompts: minimum information, discriminating cues, and exam-style decision rules.

Most flashcards fail for one reason: they test recognition, not retrieval. If a card feels easy because you can ‘see the answer’ in the question, it is training familiarity — not exam performance. Good cards force you to generate the answer, then give fast, corrective feedback.

Why this works

Evidence reviews consistently rate practice testing (retrieval) and distributed practice (spacing) as high-utility strategies. Flashcards work when they behave like tiny tests — not mini-notes.
1

Rule 1 — One decision per card

A card should test one thing: a definition, a threshold, a first-line choice, a red flag, a contraindication. If it tests three things, you’ll get ‘partial recall’ that feels like progress but doesn’t move scores.
2

Rule 2 — Write questions like an examiner

Use prompts that look like the action you need on exam day: 'Best next step', 'Most likely diagnosis', 'First-line', 'Most appropriate investigation'.
3

Rule 3 — Add a discriminating cue

The cue should separate similar options. Example: include the detail that flips the answer (age group, timing, key risk factor, hallmark symptom).
4

Rule 4 — Include the trap

Add a short “don’t confuse with…” line or a single exception. Exams love near-misses; your cards should train discrimination, not only recall.
5

Rule 5 — Prefer cloze only when it forces recall

Cloze deletions are fine if they require generation. If your brain can guess the blank from grammar alone, it’s not retrieval — rewrite it.
6

Rule 6 — Attach a source (lightweight)

Add a tiny reference line (guideline name / page / short rationale). This prevents myth-propagation and makes later updates easy.

Common failure mode

Copy-pasting paragraphs into the back of cards. You create an illusion of work, but you’re training scrolling and rereading — not recall.
SourceDunlosky et al. — Evidence review of learning techniques (practice testing + spacing)
Open Link
SourceRoediger & Karpicke (2006) — Testing effect (PubMed)
Open Link
SourceCepeda et al. (2006) — Distributed practice (spacing) meta-analysis (PubMed)
Open Link