What is SAT error analysis?

Classifying every wrong answer by content domain, by skill, and by cause — rather than only marking it wrong. The cause is the part that decides what to do next, because a question lost to a missing method and one lost to the clock need opposite responses.

Isn't 'careless mistake' a real category?

Rarely, and it is the most over-used label in SAT preparation. Most errors filed as careless turn out to be a method that only works on easy cases, a habit of not re-reading what the question asked, or the first sign of time pressure. Calling them careless ends the investigation exactly where it should start.

How do I know if a mistake was a timing problem?

Redo the question untimed before reading any explanation. If you get it right, it was pacing. If you still cannot get it, it was content or method. That single test separates two problems that look identical on a score report.

How often should I redo the analysis?

After every full-length test, and at the end of each block of focused work. The point of comparison is the previous profile: a skill that has closed leaves the plan, and one that has not gets a different approach rather than more repetitions of the same one.

Do I need a tutor to do this?

No, and doing it alone is far better than not doing it. The step that is genuinely hard alone is assigning the cause honestly, because the labels a student reaches for are the comfortable ones. If you work alone, force yourself to write a sentence explaining each label.

The method

Four causes explain almost every wrong answer

Marking a test tells you how many you missed. Classifying it tells you what to do. These are not the same activity, and only the second one changes a score.

SAT error analysis means labelling every wrong answer three ways: which content domain it came from, which skill it tested, and why you got it wrong. The third label is the one that decides the work. Nearly every SAT error falls into one of four causes — a content gap, a method error, a misread question, or running out of time — and each has a different fix.

The reason this matters is that the four look identical on a score report. A student who sees "Advanced Math: 60%" will go and revise Advanced Math. If the real cause was that the last six questions were rushed because a Problem-Solving question earlier ate four minutes, that revision will change nothing and the student will conclude, wrongly, that they are bad at Advanced Math.

The four causes

How to tell them apart, and what each one needs

The distinguishing test is simple: redo every wrong question untimed, before reading any explanation.

Content gap

How you know: You still cannot do it untimed, even after seeing the topic named.

The fix: Reteach the underlying concept, then drill it in a filtered question set until it is reliable at the difficulty band you are failing at. This is the slowest of the four to fix and the most durable once fixed.

Method error

How you know: You know the topic and can do easy versions, but your approach breaks on harder ones.

The fix: Replace the method rather than practising it more. Students often have an approach that works on textbook cases and collapses under SAT framing, and repetition entrenches it.

Misread question

How you know: On review you can immediately see what was being asked and could have answered it.

The fix: A checking routine, not more content. Underline the actual question, and for Reading and Writing, name what the correct answer must do before looking at the options.

Out of time

How you know: You get it right untimed, quickly, with no help.

The fix: A pacing problem, and usually caused earlier in the module than where it shows up. Find the questions that ate the minutes, not the ones that were rushed at the end.

Worked example

The same score, two different plans

Two students sit the same diagnostic and both score 1200. Their error profiles have almost nothing in common.

Two 1200 diagnostics compared
	Student A	Student B
Total	1200	1200
Math / Reading and Writing	670 / 530	540 / 660
Errors cluster in	Craft and Structure, Information and Ideas	Advanced Math, Problem-Solving and Data Analysis
Dominant cause	Misread question, then out of time in module 2	Content gap in nonlinear functions
First four weeks	A checking routine for R&W, plus timed sets to build module-2 stamina	Reteach quadratics and exponentials, then drill filtered sets untimed
What would be wasted	Any hour spent on Math content	Any hour spent on reading speed

Put both students in the same 1200-to-1400 course and roughly half of each one's hours go to material they did not need. That is the argument for diagnosing first, and it is a scheduling argument rather than a marketing one.

Do it yourself

A review routine that works without a tutor

If you take nothing else from this page, take this sequence. It costs about ninety minutes after a full-length test.

1
Before reading any explanation, redo every question you got wrong, untimed. Mark which ones you now get right.
2
For each wrong answer, write one sentence saying why. Not 'careless' — an actual sentence, such as 'I solved for x when the question asked for 2x'.
3
Assign one of the four causes to each. Ones you got right on the redo are almost always timing, not content.
4
Tally the causes by domain and by skill. Look for the two or three cells with the most errors.
5
Write next fortnight's plan against those cells only. Everything else waits.
6
Keep the tally. After the next test, compare the two profiles rather than the two scores.

Then feed it forward: turning an error profile into a weekly study plan.

Error analysis FAQ

Get the classification done properly once

The hardest part to do alone is labelling causes honestly. A diagnostic and an analysis session gives you a profile you can then maintain yourself.

Start with a diagnostic Talk to TestPrep