For an SAT inference question, choose the conclusion that the passage supports, not a claim that merely sounds plausible. Your answer may connect ideas rather than repeat a sentence, but every important part of it must stay within the evidence.
The most useful habit is to ask, “What would I have to add to the passage to make this answer true?” If the answer needs a new assumption, broader population, or stronger causal claim, it may go beyond what the text establishes.
College Board includes comprehension and reasoning within the Information and Ideas domain of Reading and Writing. The examples below are original teaching exercises, not official SAT questions.
Inference is a controlled conclusion
An inference is not a free interpretation. It connects the information supplied to a conclusion that follows from it.
Suppose a passage says a museum's newly discovered letter was written before a previously known draft. You may infer that the letter could change the documented timeline. You cannot automatically infer that the letter proves every existing biography is false.
The distinction is scope. The first conclusion stays close to what the discovery affects. The second expands from one piece of evidence to an entire body of scholarship.
Do not choose an option because it is the most interesting implication. Choose it because the text supports it best.
Use a four-step process
- Identify the task. Are you completing a conclusion, predicting a supported response, or drawing an inference?
- List the relevant facts. Keep them short and preserve qualifications.
- State the supported connection. Use your own plain language.
- Test every part of the answer. One unsupported phrase can make the whole option wrong.
A short prediction helps prevent answer choices from rewriting your memory of the passage.
For example: “The result applies to the tested conditions; they have not checked other conditions.” That prediction can reject both an overconfident generalization and a claim that the result has no value at all.
Original example 1: do not invent causation
A school opened a quiet study room in January. During the same month, it extended library hours and began offering after-school tutoring. The number of students completing weekly assignments rose. The school has not separately measured the effects of the three changes.
Which conclusion is best supported?
A. The quiet room alone caused the increase in completed assignments.
B. Extending library hours had no effect on students.
C. The available information does not isolate the effect of the quiet room.
D. The school should close the tutoring program.
Answer: C.
The passage describes several changes and explicitly says their effects were not measured separately. C preserves that limit.
A attributes the result to one change without separating it from the others. B excludes a possible contribution without evidence. D turns an observation into a recommendation the passage never supports.
Notice that C does not claim the quiet room was useless. “The effect has not been isolated” and “there was no effect” are different statements.
Original example 2: preserve the population
Researchers tested a coating on twelve ceramic tiles exposed to a controlled cycle of heat and moisture. Compared with untreated tiles in the same conditions, the coated tiles developed fewer visible cracks. The researchers have not tested the coating on metal or wood.
Which conclusion is best supported?
A. The coating prevents all damage on every building material.
B. The coating reduced visible cracking in the ceramic tiles under the tested conditions.
C. Metal and wood cannot benefit from protective coatings.
D. The researchers proved that moisture never damages coated surfaces.
Answer: B.
B matches the material, outcome, and conditions described. It does not expand the result to untested materials.
A changes “fewer visible cracks” into “prevents all damage” and “ceramic tiles” into “every building material.” C invents a negative result for materials not tested. D introduces an absolute claim and narrows the cause in a way the experiment does not establish.
During practice, underline the exact words that expand an answer beyond the passage. This makes “too broad” a specific explanation rather than a vague label.
Original example 3: connect two facts
Every surviving copy of a small newspaper's April issue contains a correction to a date printed in March. A newly discovered March copy has the original date crossed out and the April correction written in the margin. The handwriting has not been dated.
Which inference is most defensible?
A. The handwritten correction may have been added after the April correction became available.
B. The editor personally corrected every March copy before publication.
C. The March copy was printed after the April issue.
D. The handwritten note proves that no uncorrected March copies ever existed.
Answer: A.
A offers a possibility consistent with the sequence without claiming certainty about the handwriting. B invents the writer, timing, and number of copies. C confuses a later annotation with the original printing date. D makes an unsupported claim about every copy.
The qualification “may” matters here because the handwriting has not been dated. But do not turn this into a rule that the most tentative option is always correct. The evidence determines how strong the answer can be.
Watch for these common answer changes
| Passage says | Wrong answer changes it to |
|---|---|
| Some participants | All participants |
| Under tested conditions | In every setting |
| Associated with | Caused by |
| Not yet measured | Proven to have no effect |
| One explanation | The only possible explanation |
| A possibility | A certainty |
| A stated observation | A recommendation or value judgment |
These shifts can be subtle. An option may repeat several correct details while changing only one relationship.
Read the whole answer, especially its final clause. A correct opening does not rescue an unsupported ending.
Do not import outside knowledge
A topic may be familiar from science or history class. Use that knowledge to understand vocabulary, but answer from the information in the passage.
If the text describes a hypothetical study, do not replace its conditions with what you remember about a real study. If a passage concerns an unfamiliar author, do not assume the author's views from the title of a book.
A useful test is to ask whether another careful reader could justify the answer using only the provided material. If your explanation depends on facts the passage never gives, reconsider it.
When two choices remain
Name the precise difference. Does one claim causation while the other claims a relationship? Does one cover all cases while the other covers the tested group? Does one attribute a view to the wrong person?
Then return to the relevant sentence, not the entire passage by default. A targeted reread is often more useful than repeatedly starting at the beginning.
If both choices still seem possible, compare how much each asks you to assume. The better answer should fit the task and evidence more directly.
Do not use answer length or sophisticated vocabulary as a tiebreaker. Neither establishes support.
Practice inference with explanations
After a short set, write one sentence proving each correct answer and one sentence identifying the flaw in your most tempting wrong answer.
For the coating example, a useful explanation is: “B preserves the tested material and outcome; A expands both.” That is more transferable than “B sounded more reasonable.”
Try changing one condition in an original example. If the researchers later tested metal, which conclusions would become possible? This exercise helps you see how evidence controls the answer.
Track accuracy, but do not convert a short inference set into a scaled score. The SAT score calculator organizes digital accuracy; it cannot reconstruct official digital scoring from raw counts. Its paper Practice Test 6 mode is separate and form-specific.
Your goal is a repeatable evidence check: preserve who, what, when, and under which conditions before you select the conclusion.
Keep Reading
- Digital SAT Reading and Writing Strategy: A Plan for Both Modules
- SAT Cross-Text Connections: Compare Two Short Passages
- SAT Words in Context: Vocabulary Strategy Without Endless Word Lists
Explore SHSPrep’s SAT program. Study with lessons, targeted practice, and two original full-length mock tests with raw performance results.
Official sources checked: September 6, 2026. This draft awaits editorial review.