Grammar Mastery Ladder

MERIDIAN™ · © 2026 · original teaching + practice (not real SAT questions). Not affiliated with or endorsed by the College Board. SAT® is a trademark of the College Board.

---

Standard English Conventions is the most winnable domain on the whole exam. It is a closed rule set — roughly a dozen rules, all of them decidable, none of them a matter of taste. Once you can name the rule, the question has one answer and you can prove it. That is not true of any other domain on the test.

So this domain deserves a proper finish, and this is it. Three documents, run in order.

---

#THE LADDER

Do not skip to stage 6 because it looks like the real thing. It is the real thing, and taking it cold tells you only that you were not ready.

---

#WHY IT IS BUILT THIS WAY

Three design decisions here are deliberate, and each is doing work. If you understand them you will not be tempted to shortcut them.

#1 · The answer sits immediately under the question in the drills, and at the back in the bank

This is not an inconsistency. They are different tools.

In the drills, you are learning to recognise a rule, and feedback delayed by twenty minutes of scrolling is feedback you have stopped caring about. Retrieving an answer and being told immediately whether it was right is the single highest-return thing in all of learning research — larger than any content change, and nearly free. So the drills give it to you instantly, with the reason, and with a note on what each wrong choice was trying to catch you with.

In the bank and the unit test, you are simulating. A back-of-book key is correct there, because on test day nobody tells you after each question.

#2 · Drills are blocked by sub-skill. The unit test is not.

Mixing problem types beats drilling them in blocks — but only after the individual procedures are learned, and only among things you might confuse. Mixing before you can perform each skill is not desirable difficulty, it is just confusion, and it produces the demoralising pattern where nothing seems to be improving.

So: block until competent, then interleave. The drills block. The bank half-blocks (grouped, but adjacent). The unit test interleaves completely.

#3 · Nothing in the unit test is labelled

Every item in the drills is tagged with the sub-skill it tests. Every item in the unit test is not, and this is the most important difference between them.

On the real exam, naming the rule is half the question. A subject–verb agreement item does not announce itself; it looks like a sentence with four punctuation-and-word options, and you have to notice that the subject is "the collection of letters" and not "letters". A drill that tells you which rule is being tested has quietly removed the harder half of the task. If your drill accuracy is 90% and your unit test accuracy is 60%, that gap is not carelessness — it is precisely this, and the fix is the diagnostic table at the end of the unit test.

---

#THE ONE PROCEDURE, FOR EVERY ITEM

By the time you reach the unit test this should be automatic. It is four steps, in this order, always.

Step 1 — Look at what varies across the four choices. The SAT tells you what it is testing by what it changes. Four choices differing only in punctuation is a Boundaries item. Four choices differing in was/were/is/are is subject–verb agreement. Four choices differing in its/it's/their/there is possessives. Read the choices before you re-read the sentence; it takes two seconds and it names the rule for you.

Step 2 — Run the Boundary Test. Cover the blank. Is what comes before it a complete sentence? Is what comes after it a complete sentence? Two completes means you need a full stop, a semicolon, a comma + FANBOYS, a colon, or a dash — and never a bare comma. This one test decides roughly half the domain.

Step 3 — Find the true subject, or the true antecedent. Strike out every prepositional phrase and every non-essential clause. What is left is what the verb has to agree with. "The box of old photographs and letters was found" — the subject is box, not letters.

Step 4 — Prove three wrong, not one right. Name the specific defect in each of the other three: comma splice, run-on, subject–verb comma, unmatched supplement pair, dangling modifier, broken parallel. If you cannot name a defect in a choice you have eliminated, you have not eliminated it, you have merely disliked it.

---

#WHAT COUNTS AS FINISHED

Not "I read all four documents". Read is not a measure of anything.

Finished is: on a mixed, unlabelled, timed set you have not seen, you score 19 of 22 or better, and for every item you missed you can name the rule and the trap. That is the standard the unit test is calibrated to, and it corresponds to giving away close to nothing in this domain on the real thing.

If you get there, stop drilling Conventions. It is a closed rule set and there is no further score in it — the time belongs to Information & Ideas or to Advanced Math, where the ceiling is higher.

---

#REFERENCE CARD — THE LADDER