Inferences

An inference is a conclusion the text supports without necessarily stating it directly. The task is usually to choose the best-supported conclusion, not one that is certain under every imaginable circumstance. Evidence can support a reasonable explanation or prediction even when it does not prove it.

Use three moves: identify the relevant facts, explain their connection, and check the answer against both. Ask, What makes this conclusion more justified than the alternatives? An unsupported answer might be true in the world; the problem is that this passage does not establish it. Do not turn that distinction into a claim that the answer is false.

For a logical-completion question, summarize the unfinished reasoning before studying the options. A passage may introduce an expectation, report conflicting evidence, and need a revised conclusion. Or it may describe a mechanism and need its consequence. Words such as however, therefore, and although help identify the relationship, but they cannot replace reading the evidence.

Predict an idea, not exact wording. Then insert each serious candidate into the blank. Does it complete the whole argument, or merely repeat one earlier observation? A conclusion can qualify an explanation without rejecting it entirely.

Some inference answers are supported by two details that perform different jobs. One may establish a pattern; another may eliminate an explanation. Mark both before choosing. If you can justify an answer using only a sentence that also fits several competitors, you may have missed the decisive clue.

Avoid importing an extra step from your own knowledge. A useful bridge should follow from the passage: The material traveled, but the personnel did not; therefore, the material could spread independently of those personnel. It should not invent a motive, a funding source, or a historical event.

Before accepting an inference, check its boundaries: who, when, how many, and under what conditions? Evidence about sampled participants does not automatically describe everyone. A result observed once can establish that something is possible; it rarely establishes that it is typical. Conversely, do not reject a strong statement when the evidence actually warrants it.

Watch the direction of conditions. If a rule permits selection only for proposals with a budget, a budget is necessary for eligibility under that rule. It is not necessarily sufficient: other requirements may apply. An observed pattern alone does not establish such a rule. For negation, rewrite carefully: not all succeeded means at least one failed, not that none succeeded.

A prediction should extend an observed relationship to a situation where the relevant conditions still apply. Identify the observed difference and the conditions in which it appeared before predicting what happens next. Do not assume that a change improves everything: a material may gain strength but lose flexibility, and a strategy may work only in a particular environment.

For comparisons, name both sides and the measurement. More sensitive to moisture does not mean stronger overall. A change from each group's starting point also does not establish which group finishes higher unless the starting points are known.

An association can support a causal explanation, but its strength depends on the design and competing explanations. Ask what changed, what was measured, and what else could produce the pattern. A comparison group, random assignment, or a mechanism can strengthen causal reasoning. Their absence does not make all evidence useless.

Keep the outcome precise. Fewer insects counted at a lamp might mean fewer insects were attracted there; it does not necessarily mean fewer insects survived in the region. Choose a qualified causal conclusion when it fits the evidence, rather than either declaring proof or refusing every causal inference.

Literary passages often reveal an attitude through a contrast: what someone says versus does, an object kept versus dismissed, or a detail noticed versus ignored. Identify the behavior first, then ask which interpretation explains its context. Mixed feelings are common, but do not select a mixed-feelings answer merely because it sounds sophisticated.

Separate character, narrator, and speaker. A narrator's description is not automatically a character's belief, and a poem's speaker is not automatically its author. Infer no more biography than the passage supplies. Distinguish a reasonable emotional reading from a precise motive that would need additional evidence.

For a difficult question, build a short chain: evidence → permitted connection → conclusion. Then look for the vulnerable link. Does the answer turn surviving records into all past events, a minimum into an exact count, or evidence of use into evidence of origin? Close distractors often preserve most of the chain and alter one link.

Test the strongest competing answer explicitly. Name the extra fact it would need. Do not automatically prefer the weakest wording: choose the most informative conclusion supported by the full passage. Evidence can establish a firm lower bound while leaving the complete history uncertain.

Keep these checks in mind

• Identify the conclusion the question needs: an implication, explanation, prediction, or logical completion.
• Anchor the answer in the relevant clues and explain how those clues connect.
• Preserve the people, conditions, time, quantity, comparison, and measured outcome.
• Reject the closest alternative by naming its unsupported step; do not confuse unsupported with false.

Begin with four brief reasoning drills, then complete sixteen multiple-choice questions. Later items generally demand closer distinctions. Identify the evidence for each answer and the specific flaw in its nearest competitor.