Command of Evidence — 700-800

Command of Evidence: The 700-800 Patterns

Part 1 of 3 — Matching Evidence to the Exact Shape of the Claim

Hard evidence items are not reading tests; they are logic-matching tests. The conclusion or claim has a precise logical shape, and exactly one choice has the same shape. Everything else is on-topic, accurate — and shaped wrong.

Archetype 1: Two-Part Conclusions Need Two-Part Evidence (data tables)

"The change sped the step it targeted without improving outcomes." "Support is highest nearby, but most supporters live far away." Count the claims: if the conclusion compares, contrasts, or says "only," the evidence must contain every comparison in one choice. Distractors cite one true row — a single value can never show a decrease, a contrast, or an "only."

Archetype 2: The Headline Number That Argues for the Other Side

When the student's conclusion is skeptical ("the data cannot show the program worked"), the most tempting distractor is the table's flashiest comparison — which supports the opposite conclusion. Before reading choices, ask: which side is this student on? Then reject anything a believer would cite.

Archetype 3: The Quotation With the Right Topic and the Wrong Mechanism

Literary claims specify a mechanism: influence exercised through others, adaptation without confrontation, grief conveyed through record-keeping. The trap quote shows the right phenomenon by the wrong route — she speaks publicly herself, he objects to his face. One word ("to his face," "at the podium") defeats it. Underline the claim's restrictive phrase and demand the quote satisfy it.

Archetype 4: Discriminating Evidence Between Rival Hypotheses

When two explanations compete, supporting evidence must be something the favored hypothesis predicts and the rival doesn't. Findings consistent with both (the phenomenon exists; the correlate grew) discriminate nothing. The strongest form is the dissociation: the effect appears where only the favored cause is present, and fails to appear where only the rival cause is present.

Command of Evidence: Traps & Speed

Part 2 of 3 — Why the Wrong Choice Feels Right

Every hard evidence distractor exploits the same weakness: strong students verify that a choice is true and on-topic, then stop. The item is testing a third check — is it the right shape?

The trap logic

The Half-Supporter. The conclusion has two clauses; the distractor nails one, usually the first. It feels responsive because it echoes the sentence you just read. Defense: before reading choices, write the conclusion as "A and/but B." Any choice serving only A or only B is dead, no matter how accurate.

The Wrong-Team Citation. On skeptical conclusions ("the data cannot establish...", "gives little reason to credit..."), the table's headline comparison — the biggest gap, the flashiest gain — supports the believer, not the skeptic. Studies of these items show this is the single most-picked wrong answer among high scorers, because the headline number is what your eye extracted first. Defense: identify the student's side, then ask of each choice, "whose lawyer would cite this?"

The One-Word Kill (quotations). The claim's restrictive phrase — "without confrontation," "never publicly," "through record-keeping" — is a tripwire. The trap quote trips it with a single detail: "to his face," "at the podium," "she rose and said." Defense: underline the restriction; scan each quote only for a violation before evaluating anything else. Violations eliminate faster than merits select.

The Consistent-With-Both Finding (rival hypotheses). A finding both accounts predict has zero discriminating power, however impressive it sounds. Defense: for each choice ask, "would the rival account also predict this?" If yes, it's out — even if it's the most detailed choice on the page.

Speed dividend

These checks run in about fifteen seconds and typically eliminate three choices without a second read of the passage. On table items, you often never need to re-read the prose at all — the conclusion's shape plus the numbers decide everything.

Command of Evidence: Timed Drill

Part 3 of 3 — Four Items at Full Difficulty

Pacing target: 60 seconds per question (table items can finish in 45). Before the choices, always:

1. Write the claim's shape: "A and B," "A but not B," "X rather than Y," or "predicts P, rival doesn't."
2. Identify the student's side on skeptical conclusions.
3. Underline any restrictive phrase in literary claims — it's the tripwire the trap quote will trip.
4. Eliminate by shape first; verify arithmetic last, and only for survivors.