Standardized tests — SAT, GRE, GMAT, LSAT, the alphabet soup — get treated as mysterious beasts requiring expensive secrets. They're not. They're pattern-recognition tests under time pressure, and the preparation that works is the same learning science that works everywhere else, pointed at a specific format.

What makes them feel different is that the stakes compress the timeline and the format is rigid — which is actually an advantage. A standardized test is the most predictable exam you'll ever take.

What the test is actually measuring

Strip away the branding and every standardized test measures the same two things:

• Do you have the underlying skill? Math, reading, reasoning — the content layer. This is where prep books and content review help, and it's necessary but not sufficient.
• Can you deploy it fast under the test's specific constraints? Time pressure, question format, the particular way the test disguises easy content in hard-looking wrappers. This layer is where most prep fails, because it gets trained separately or not at all.

The insight most test-takers miss: the second layer is trainable, and it responds to the same mechanisms as the first. Timed practice, error analysis, spaced re-testing of weak points — the learning science doesn't change because the test is standardized.

The plan that works

Start with a baseline. Take a real timed practice test before you study anything. Not to feel bad — to find out which layer needs work. If the content gaps are huge, that's the plan. If you know the content but can't finish in time, that's a different plan. Studying without a baseline is prescribing without diagnosing.

Practice in timed mixed sets, not untimed blocks. The test doesn't ask you "20 algebra questions" — it asks you "the next question, whatever it is, in 90 seconds." Practice that way: mixed topics, real timing, actual test format. Untimed blocks build knowledge; timed mixed sets build test performance, and the test is what's scored.

Keep an error log that captures why. This is the piece most prep skips. Every miss gets logged with the cause — content gap, misread the question, ran out of time, strategy error, careless slip. "I got it wrong" tells you nothing; "I misread 'except' and answered the opposite" tells you exactly what to fix.

Write a decision rule for each pattern. A logged error is data; a rule is the fix. "If the question says 'except,' I underline it before answering." "If two answers both look right, I check which one the question actually asks." Rules convert your error history into test-day behavior — they're the bridge between knowing and scoring.

Re-test the error items on a schedule. The misses are your highest-yield study material — they're the specific places your preparation is leaking. Re-test them at expanding intervals (1/3/7/14 days) instead of doing fresh problems on topics you already know. The error log is the syllabus; the spacing is the schedule.

What to practice, in what proportion

A weekly cadence that works for most tests:

• 3 timed mixed sets — the core training. Mixed topics, real timing, actual format. This is where the second layer (deployment under pressure) gets built.
• 2 targeted drills — focused work on whatever the error log says is weakest. This is where the first layer (content) gets repaired.
• 1 review block — error log maintenance: new misses logged, rules written, old errors re-tested.
• Every 1–2 weeks: one full-length test — the measurement tool, not the practice. It updates the baseline and tells you if the plan is working; it doesn't replace the daily work.

The format-specific layer

Each test has its own quirks, and they're worth learning — after the fundamentals:

• SAT/GRE math: the questions are usually easier than they look — the test disguises basic algebra in word-problem wrappers. Practice translating the wrapper to the math.
• Reading sections: the answer is always in the passage. The test's trick is making you rely on what you "know" instead of what's written. Practice going back to the text, not to your impression of it.
• Adaptive tests (GRE, GMAT): early questions matter more — they set the difficulty track. Front-load accuracy; the end-of-section scramble matters less than the start.
• Time allocation: every test has a point where a question costs more time than it's worth. Practice recognizing it and moving on — the strategic skip is a score-preserving skill, not a surrender.

Where LearnX fits

LearnX's Exam Sprint is built for exactly this: timed, exam-aligned questions generated from your actual materials, an error log that becomes the re-test schedule automatically, and variants that keep you from memorizing question wording instead of learning the underlying pattern. The Sprint's built-in spacing handles the re-test schedule, and the performance tracking surfaces which weaknesses are actually costing points — which is what the error log is for, just automated.