Hard Inferences questions make the conclusion less visible. The passage may combine a baseline, an exception, and a later result; the answer follows only after you connect all three and resist dropping a qualification midway.
The Inferences skill
- Valid inference
- A conclusion that follows logically from the text's explicit facts and strongest implications without requiring unsupported information.
Turn the passage into premises
Identify the hinge premise—the sentence that changes how another fact should be interpreted. Words such as only, unless, despite, whereas, and even when often mark that hinge.
Evidence, inference, and assumption
Baseline
What is initially shared. Example: Fresh applications block water.
Hinge
The condition that separates outcomes. Example: Repeated bending affects thicknesses differently.
Integrated inference
The conclusion explaining both. Example: Durability may depend on thickness.
Control scope and certainty
A repeatable method
Premise-to-conclusion workflow
- Index
Number the relevant premises in passage order.
- Find the hinge
Locate the premise that qualifies, contrasts with, or redirects another.
- Chain
Write P1 + P2 + P3 → conclusion in plain language.
- Audit
Check every noun, condition, and degree word in the proposed answer.
Worked SAT-style example
Follow only what the text supports
A design team found that users completed a task quickly with an icon-only toolbar after receiving a tutorial. New users without the tutorial completed the task more slowly than users of a labeled toolbar. After one week of regular use, however, the untutored icon-only group matched the labeled-toolbar group. The findings suggest that ______
Question: Which choice most logically completes the text?
- icon-only controls may impose an initial learning cost that decreases with instruction or experience
- labeled controls become impossible to use after one week
- tutorials permanently eliminate every difference among interface designs
- new users immediately understand icon-only controls better than labeled controls
- Extract
Tutorial group: icon-only use is fast after instruction.
- Connect
Untutored new users: icon-only use is initially slower than labeled use.
- Verify
After regular use: the gap closes. Together, the clues support an initial, reducible learning cost.
Why each choice works or fails
- Correct: it explains all three time-and-training conditions with appropriate qualification.
- Unsupported: nothing says labeled controls become unusable.
- Overstated: one task does not prove permanent elimination of every design difference.
- Contradicted: untutored new users were initially slower with icon-only controls.
Check your reasoning
Mini check
A coating blocked water when freshly applied. After repeated bending, thin applications cracked, while thicker applications remained intact. What follows?
- The coating's durability under bending may depend on application thickness.
- Every coating fails when bent once.
- Thin coatings blocked more water before bending.
- Bending improved all applications equally.
Show answer and explanation
Answer: The coating's durability under bending may depend on application thickness.
The starting performance is shared, but outcomes diverge after bending by thickness. The inference must integrate both stages.
Key takeaways
- Number the premises and find the hinge.
- Carry every condition and qualifier into the conclusion.
- Reject answers that explain only one stage of the evidence.
- Hard questions reward disciplined chaining, not imaginative guessing.
Put these notes into practice
Apply the ideas with SAT-style questions, then reinforce key details with flashcards.