Words-in-Context System

Predict the word yourself before you read a single choice — that one habit converts the most common R&W question on the test from a coin-flip into a near-certainty. Words in Context sits inside Craft & Structure (≈28% of R&W, the largest domain) and is the single most frequently appearing R&W question type (College Board domain weighting; type-frequency per PrepScholar/UWorld analysis — exact count is an estimate, not official). Win this type and you bank a large, reliable block of points every test.

Sources: method per Erica Meltzer (The Critical Reader), UWorld, ArborBridge. All example items below are MERIDIAN-original.

Point convention: each item = 1 of ~54 R&W questions; a Module-1 miss is a ceiling-risk miss (System Map).

---

#WHAT IT TESTS

A short passage has a blank (or an underlined word). You choose the word that best fits the logic and register of the sentence. It is not a vocabulary-size test — it is a reading-the-context test. The hard versions hinge on the secondary meaning of a familiar word (qualify, table, economy, novel, sound), so "I know this word" is a trap, not a shortcut.

---

#THE METHOD — PREDICT, THEN MATCH

1. READ the full sentence — and the sentence before it if the logic depends on it.
2. COVER the four choices. (Literally ignore them at first.)
3. PREDICT your own simple word for the blank, from the context's logic.
4. MATCH the choice closest to your prediction.
5. PLUG IT BACK IN and confirm the whole sentence still makes sense.

Why this works (mechanism): the four choices are engineered to bait you — two will "sound smart," one will be the dictionary-primary meaning of a word used here in a secondary sense. If you read the choices first, they overwrite your sense of the sentence. Your own prediction is the shield: you compare each choice to your word, not to your gut. [This is the difference between a ~700 and a 780 on this type — UWorld.]

The two-finalist tool: if you're stuck between two choices, find the single difference between them and test that difference against the sentence's logic. The right answer is the one whose connotation/direction matches.

---

#TRAP TAXONOMY (the engineered wrong answers)

• Primary-meaning bait: the common meaning of a word used here in a less-common sense. ✗ choosing "criticize" for qualify when the sentence means "limit/soften."
• Synonym-in-isolation: a word that means your prediction in a dictionary but is wrong for this sentence's register or logic.
• Two near-synonyms: if two choices mean nearly the same thing, both are wrong — there is one best answer, so a duplicated meaning can't be it. [eliminates 2 at once]
• Too-extreme: a word stronger/more loaded than the neutral context supports.

---

#SAME ITEM, TWO LEVELS (MERIDIAN-original)

Passage: "The committee's findings were far from final; the report explicitly ___ its conclusions, noting that further data could overturn them." Choices: (A) endorsed (B) qualified (C) criticized (D) publicized

~700 reasoning (trap): "The report said something about its conclusions — criticized sounds critical, and they're being cautious." → picks (C). Trap: read the choices first, matched a vibe ("critical/cautious"), ignored that criticized ≠ "limited." [−1; a Module-1 ceiling risk]

780+ reasoning (precise): Cover choices. Context = "far from final… could overturn them" → predict "limited / hedged." Match: qualified (to qualify = to limit/soften a claim). Plug back: "the report limited its conclusions" ✓. Eliminate: endorsed/publicized = wrong direction; criticized = attacks, doesn't limit. → (B). The move: predicted "limited" before the choices, then recognised the secondary meaning of qualified. [secures the question; Module-1 safe]

---

#DIAGNOSE YOUR ANSWER

• Did you read the choices before predicting? If yes, that's the leak — they hijacked you. Fix: cover-and-predict, every time.
• Did you pick a word's primary meaning? Check whether the sentence needs a secondary sense. Fix: define the word as used here, not as you first learned it.
• Stuck between two? You skipped the single-difference test. Fix: isolate the one differing connotation, test it against the logic.
• Picked the strongest word? Check the register — neutral context rejects loaded words.

---

#DRILLS (timed · self-marked · MERIDIAN-original)

Drill 1 — Predict-first (target: 6/6, ≤60 s). Cover choices; write your own word for each blank, THEN match.

1. "Once ___ by critics, the theory is now standard." (predict: rejected/dismissed)
2. "The data were ___: every replication agreed." (consistent)
3. "She gave a ___ reply, revealing nothing." (evasive/guarded)
4. "His prose is ___, stripped of any excess." (spare/concise)
5. "The plan was ___ at first but gained support." (controversial/contested)
6. "They ___ the rule, allowing rare exceptions." (relaxed/softened) PASS: your predicted word matches the intended sense on 6/6. If <6, you're reading too fast — slow the context read and redo tomorrow.

Repetition protocol: fail → repeat next day. Two clean passes running → predict-first is automatic.

---

#TRANSFER

Abstract rule: the answer is the word your own prediction from context selects — topic-independent. Second context: a science passage ("the enzyme ___ the reaction, slowing it") and a history passage ("the treaty ___ tensions, easing them") are the same question — predict ("slowed" / "reduced"), then match. What stays constant: cover-predict-match-confirm and the trap list. What changes: only the topic and the specific word.

---

#REFERENCE CARD — WORDS IN CONTEXT

1 Read sentence (+ prior if needed)   2 COVER choices   3 Predict your own word
4 Match closest   5 Plug back to confirm
TRAPS: primary meaning of a 2nd-sense word · synonym-in-isolation ·
       two near-synonyms = both wrong · too-extreme
Stuck on two? find the ONE difference, test it against the logic.
Each item = 1 R&W question; Module-1 miss = ceiling risk.

---

MERIDIAN™ · © 2026 · For Neo's personal study use. Not affiliated with or endorsed by the College Board. SAT® is a trademark of the College Board.

MERIDIAN · FORGE · WORDS-IN-CONTEXT SYSTEM | v1.0