Inferences & Supporting Evidence
Learn how to draw inferences a nonfiction text actually supports, test them for overreach, and cite the exact lines that license each claim you make.
What you'll do in this lesson
A voice-first session with the Crimsora tutor on Inferences & Supporting Evidence, then targeted practice and FRQs — with the tutor adapting to where you get stuck.
What this lesson covers
A nonfiction writer almost never spells out everything. An article about a shrinking town tells you the high school merged with the one two counties over; it does not add "and this means families with children are leaving." You supply that. That move — combining what the page states with what you already know about the world — is an inference, and it is the single most-used reading skill in this course.
The catch is that inferences can drift. It is easy to slide from what the text supports to what you personally suspect, and once you cross that line your reading stops being a reading and becomes an opinion wearing a costume. This lesson teaches you to make the leap and then prove you landed somewhere the text can hold: locate the specific lines that license your inference, name the background knowledge you added, and test the claim against readings that fit the same evidence.
The catch is that inferences can drift. It is easy to slide from what the text supports to what you personally suspect, and once you cross that line your reading stops being a reading and becomes an opinion wearing a costume. This lesson teaches you to make the leap and then prove you landed somewhere the text can hold: locate the specific lines that license your inference, name the background knowledge you added, and test the claim against readings that fit the same evidence.
Stated, Supported, and Overreached: Three Kinds of Claims
An inference is a conclusion you reach that the text does not state directly but does make reasonable, using its evidence plus knowledge a normal reader brings. Two parts matter equally. Drop the text and you are guessing; drop the background knowledge and you are only paraphrasing.
Consider a sentence from a report: "In the three months after the library shortened its hours, résumé printing at the employment desk fell by more than half."
The overreach is tempting because it feels emotionally true, but notice what it requires: you had to invent an attitude no line describes. A supported inference fills a gap the text leaves open; an overreach installs new furniture in a room the author never built.
Students most often go wrong in one of two opposite directions. Some refuse to infer at all and answer questions by copying sentences, which never explains anything. Others infer freely and then cannot find a line to point to. The fix for both is the same habit: state the claim, then immediately ask which words on the page put it there.
Consider a sentence from a report: "In the three months after the library shortened its hours, résumé printing at the employment desk fell by more than half."
| Type of claim | Example | Why it lands there |
|---|---|---|
| Stated | Résumé printing dropped after the hours changed. | The sentence says it in plain words. |
| Supported inference | Some job seekers depended on the library's printers and lost access. | The drop plus common knowledge that printing requires a printer makes this the most economical explanation. |
| Overreach | The library board did not care about unemployed residents. | Nothing in the sentence describes anyone's motives or feelings. |
Students most often go wrong in one of two opposite directions. Some refuse to infer at all and answer questions by copying sentences, which never explains anything. Others infer freely and then cannot find a line to point to. The fix for both is the same habit: state the claim, then immediately ask which words on the page put it there.
Finding the Lines That License the Inference
"Cite the specific lines" means specific. Pointing at a whole paragraph is a way of saying you have not found the evidence yet.
Work backward from the claim. Ask what would have to be true on the page for your inference to hold, then hunt for that. If you infer that a scientist quoted in an article is skeptical of a new treatment, you need words that carry doubt — a hedge like "preliminary," a contrast word like "however," a comparison to a failed trial. Find the actual syllables doing the work.
The most productive places to look in nonfiction are numbers and their timing, quoted speech and who was chosen to speak, contrast words such as "but," "instead," and "despite," and details that seem oddly specific for no stated reason. When a writer includes a detail without explaining it, that detail is usually the invitation to infer.
Once you find the words, quote briefly and precisely. A four-word quotation embedded in your own sentence proves more than three full lines dropped in as a block, because the short quote shows you know which part matters. Compare these:
Weak: The author says, "In the three months after the library shortened its hours, résumé printing at the employment desk fell by more than half."
Strong: The drop happened "in the three months after" the change, a timeline tight enough to link the two.
The strong version isolates the evidence — the timing — and that isolation is the beginning of your explanation. Also name where the line lives: paragraph 4, line 12, or the third paragraph of the section on funding. Your reader should be able to put a finger on it in seconds.
Work backward from the claim. Ask what would have to be true on the page for your inference to hold, then hunt for that. If you infer that a scientist quoted in an article is skeptical of a new treatment, you need words that carry doubt — a hedge like "preliminary," a contrast word like "however," a comparison to a failed trial. Find the actual syllables doing the work.
The most productive places to look in nonfiction are numbers and their timing, quoted speech and who was chosen to speak, contrast words such as "but," "instead," and "despite," and details that seem oddly specific for no stated reason. When a writer includes a detail without explaining it, that detail is usually the invitation to infer.
Once you find the words, quote briefly and precisely. A four-word quotation embedded in your own sentence proves more than three full lines dropped in as a block, because the short quote shows you know which part matters. Compare these:
Weak: The author says, "In the three months after the library shortened its hours, résumé printing at the employment desk fell by more than half."
Strong: The drop happened "in the three months after" the change, a timeline tight enough to link the two.
The strong version isolates the evidence — the timing — and that isolation is the beginning of your explanation. Also name where the line lives: paragraph 4, line 12, or the third paragraph of the section on funding. Your reader should be able to put a finger on it in seconds.
Stress-Testing an Inference for Overreach
Before you commit, run three checks.
Check the direction. Does the evidence point at your claim, or merely sit near it? A statistic about falling enrollment is near a claim about teacher quality but does not point at it. If you cannot draw a straight line from quotation to conclusion, you are leaning on assumption.
Check for rival readings. Ask what else would explain the same evidence. The résumé printing could have fallen because a new employment center opened nearby. If a rival reading fits just as well, your inference is not wrong — it is just not yet warranted at full strength. Either find a line that rules the rival out, or soften your wording: "suggests," "likely," "one explanation the text supports." Hedging is not weakness; it is accuracy, and it is what careful nonfiction writers do themselves.
Check what you imported. Every inference adds outside knowledge. Say what yours is. Adding "printing requires access to a printer" is safe, because every reader knows it. Adding "library boards are indifferent to poor residents" is an opinion, not background knowledge, and it does not belong in an inference about this text.
A useful distinction: relevant background knowledge is information almost any reader would accept and that the author could reasonably assume. Personal opinion is a position other reasonable readers dispute. When you catch yourself importing the second kind, you have found the exact moment your inference left the text.
One more trap is causation. Nonfiction is full of sequences — this happened, then that happened. Sequence alone licenses "these may be connected," not "this caused that." You need the author to supply a mechanism, an expert's statement, or a controlled comparison before you claim cause.
Check the direction. Does the evidence point at your claim, or merely sit near it? A statistic about falling enrollment is near a claim about teacher quality but does not point at it. If you cannot draw a straight line from quotation to conclusion, you are leaning on assumption.
Check for rival readings. Ask what else would explain the same evidence. The résumé printing could have fallen because a new employment center opened nearby. If a rival reading fits just as well, your inference is not wrong — it is just not yet warranted at full strength. Either find a line that rules the rival out, or soften your wording: "suggests," "likely," "one explanation the text supports." Hedging is not weakness; it is accuracy, and it is what careful nonfiction writers do themselves.
Check what you imported. Every inference adds outside knowledge. Say what yours is. Adding "printing requires access to a printer" is safe, because every reader knows it. Adding "library boards are indifferent to poor residents" is an opinion, not background knowledge, and it does not belong in an inference about this text.
A useful distinction: relevant background knowledge is information almost any reader would accept and that the author could reasonably assume. Personal opinion is a position other reasonable readers dispute. When you catch yourself importing the second kind, you have found the exact moment your inference left the text.
One more trap is causation. Nonfiction is full of sequences — this happened, then that happened. Sequence alone licenses "these may be connected," not "this caused that." You need the author to supply a mechanism, an expert's statement, or a controlled comparison before you claim cause.
Writing the Inference So a Reader Can Follow It
An inference on paper has three moves, and skipping the third is the most common problem in student writing.
First, claim: state the inference in one sentence, in your own words, hedged if the evidence is partial. Second, evidence: quote the shortest phrase that does the work and locate it. Third, reasoning: explain how that phrase gets you to that claim, including the background knowledge you added. The reasoning step is where the thinking becomes visible.
A complete version reads like this: "The passage suggests the library serves as basic job-search infrastructure, not only as a place for books. Résumé printing fell 'by more than half' in the three months after hours were cut (paragraph 2), and a staff member explains that 'people don't have printers at home' (paragraph 3). Because printing a résumé requires equipment most applicants lack, the timing indicates that closing the doors earlier removed a service people were relying on for work."
Notice the ratio: one line of claim, one of quotation, two of reasoning. Quote-heavy paragraphs with no explanation are sometimes called quote dumps, and they leave your reader to do your thinking.
Two habits keep the write-up honest. Match your verb to your certainty — "states" for what is on the page, "suggests" or "implies" for what you inferred. And never quote a phrase you have not explained; if the words are worth including, they are worth a sentence of your own analysis. When you build inference paragraphs this way, the same structure carries directly into the longer analytical writing you will do later in the unit.
First, claim: state the inference in one sentence, in your own words, hedged if the evidence is partial. Second, evidence: quote the shortest phrase that does the work and locate it. Third, reasoning: explain how that phrase gets you to that claim, including the background knowledge you added. The reasoning step is where the thinking becomes visible.
A complete version reads like this: "The passage suggests the library serves as basic job-search infrastructure, not only as a place for books. Résumé printing fell 'by more than half' in the three months after hours were cut (paragraph 2), and a staff member explains that 'people don't have printers at home' (paragraph 3). Because printing a résumé requires equipment most applicants lack, the timing indicates that closing the doors earlier removed a service people were relying on for work."
Notice the ratio: one line of claim, one of quotation, two of reasoning. Quote-heavy paragraphs with no explanation are sometimes called quote dumps, and they leave your reader to do your thinking.
Two habits keep the write-up honest. Match your verb to your certainty — "states" for what is on the page, "suggests" or "implies" for what you inferred. And never quote a phrase you have not explained; if the words are worth including, they are worth a sentence of your own analysis. When you build inference paragraphs this way, the same structure carries directly into the longer analytical writing you will do later in the unit.
Key terms
- Inference.
- A conclusion not directly stated in a text that a reader reaches by combining the text's stated evidence with relevant background knowledge.
- Explicit statement.
- Information the text says outright; quoting it is paraphrase or citation, not inference.
- Textual evidence.
- The specific words, phrases, numbers, or lines in the text that support a claim, identified precisely enough that a reader can locate them.
- Warrant (reasoning).
- The explanation connecting a quotation to a claim — the sentence that says why this evidence produces that conclusion.
- Relevant background knowledge.
- Widely shared information an author can reasonably assume a reader has, which fills gaps the text leaves open; distinct from personal opinion.
- Overreach.
- A claim that goes beyond what the text can support, usually by inventing motives, feelings, or causes the passage never establishes.
- Hedge.
- A qualifying word such as 'suggests,' 'likely,' or 'may' that matches the strength of a claim to the strength of the evidence.
- Rival reading.
- An alternative explanation that fits the same evidence; a strong inference either rules it out with a cited line or is stated more cautiously.
Worked example
Read the passage, then state one inference the text supports, cite the exact words that license it, and explain your reasoning.
(1) When Ridgeway's public library shortened its weekday hours, the branch manager expected complaints from parents. (2) Instead, the loudest objections came from the employment desk on the second floor. (3) In the three months after the change, résumé printing at that desk fell by more than half. (4) "People don't have printers at home," one staff member said, "and the bus stops running out here at seven." (5) The library restored evening hours the following spring and has since added a second printing station.
(1) When Ridgeway's public library shortened its weekday hours, the branch manager expected complaints from parents. (2) Instead, the loudest objections came from the employment desk on the second floor. (3) In the three months after the change, résumé printing at that desk fell by more than half. (4) "People don't have printers at home," one staff member said, "and the bus stops running out here at seven." (5) The library restored evening hours the following spring and has since added a second printing station.
Step 1 — Find the detail the author includes but does not explain. Line 4 mentions the bus schedule. Nothing in the passage says why a bus matters to a library, which is exactly the signal that the author expects you to connect it yourself.
Step 2 — Draft a claim. Try: the shortened hours pushed the library's services out of reach for people who could not get there earlier in the day, particularly those depending on public transit.
Step 3 — Locate the licensing lines. Three pieces do the work: the objections came from "the employment desk" (line 2), printing fell "by more than half" "in the three months after the change" (line 3), and the staff member's pairing of "people don't have printers at home" with "the bus stops running out here at seven" (line 4).
Step 4 — Name the background knowledge. Job applications usually require printed or uploaded documents; someone without a home printer must use a public one; someone without a car depends on the bus. All three are facts most readers accept, not opinions.
Step 5 — Test for overreach and rival readings. Could printing have fallen for another reason, such as fewer people looking for work? Possible, but line 5 helps: the library restored the hours and added printing capacity, a response that only makes sense if staff attributed the drop to access. That is still an inference about institutional reasoning, so hedge slightly.
Step 6 — Write it with reasoning attached. "The passage suggests the library functions as job-search infrastructure for residents without printers or cars. Complaints came from 'the employment desk,' printing fell 'by more than half' 'in the three months after the change,' and a staff member links the loss to two facts: 'people don't have printers at home' and 'the bus stops running out here at seven.' Because applying for work requires printed materials and getting to the branch requires transportation, cutting weekday evening hours effectively removed the service for the people who needed it most — which is why restoring those hours and adding a printing station reads as a correction."
What to avoid: claiming the manager did not care about unemployed residents. Line 1 says only what she expected, and line 5 shows the branch reversed course.
Step 2 — Draft a claim. Try: the shortened hours pushed the library's services out of reach for people who could not get there earlier in the day, particularly those depending on public transit.
Step 3 — Locate the licensing lines. Three pieces do the work: the objections came from "the employment desk" (line 2), printing fell "by more than half" "in the three months after the change" (line 3), and the staff member's pairing of "people don't have printers at home" with "the bus stops running out here at seven" (line 4).
Step 4 — Name the background knowledge. Job applications usually require printed or uploaded documents; someone without a home printer must use a public one; someone without a car depends on the bus. All three are facts most readers accept, not opinions.
Step 5 — Test for overreach and rival readings. Could printing have fallen for another reason, such as fewer people looking for work? Possible, but line 5 helps: the library restored the hours and added printing capacity, a response that only makes sense if staff attributed the drop to access. That is still an inference about institutional reasoning, so hedge slightly.
Step 6 — Write it with reasoning attached. "The passage suggests the library functions as job-search infrastructure for residents without printers or cars. Complaints came from 'the employment desk,' printing fell 'by more than half' 'in the three months after the change,' and a staff member links the loss to two facts: 'people don't have printers at home' and 'the bus stops running out here at seven.' Because applying for work requires printed materials and getting to the branch requires transportation, cutting weekday evening hours effectively removed the service for the people who needed it most — which is why restoring those hours and adding a printing station reads as a correction."
What to avoid: claiming the manager did not care about unemployed residents. Line 1 says only what she expected, and line 5 shows the branch reversed course.
Practice questions
A science article states: "The tagged sea turtles returned to the same beach within nine days of last year's arrival date, despite currents that had shifted more than sixty miles east." Which inference does this sentence most directly support?
- The turtles navigate using something other than ocean currents.
- Ocean currents have no effect on marine animal migration.
- Researchers had tagged the turtles at that beach the previous year.
- Warming water temperatures are altering the turtles' migration.
Answer: The turtles navigate using something other than ocean currents.
The key word is "despite," which sets the arrival against the shifted currents. If the turtles arrived on schedule while the currents moved, current-following alone cannot explain the route — so some other navigational cue is at work. The claim that currents have no effect on migration overreaches from one species and one observation to all marine animals. The tagging detail may be true but the sentence never says where tagging happened. Warming water is not mentioned anywhere; that answer imports outside information the passage does not raise.
A student writes: "The author clearly resents the new zoning rules, since paragraph 5 says the ordinance 'passed after four public hearings.'" Explain what is wrong with this inference and revise it into a claim the evidence can actually support.
Answer: The quoted phrase describes a procedure, not an attitude, so it cannot license a claim about the author's resentment; a supportable revision would be that the ordinance faced substantial public scrutiny before passing, since four hearings is an unusually long process for a routine measure.
Two problems. First, the evidence points in a different direction than the claim: 'passed after four public hearings' reports how the rule was adopted and contains no evaluative language, no sarcasm, and no loaded word. Second, the student imports an attitude rather than background knowledge — resentment is a state of mind no reader can verify from a neutral procedural detail. The revision keeps the same quotation but scales the claim down to what the words reach: the number four, compared with the single hearing most routine measures get, supports an inference about the level of scrutiny. If the student wants to claim the author disapproves, they need different evidence entirely — word choice, the balance of sources quoted, or which consequences the author chooses to describe.
A report notes that a school moved its start time from 7:25 to 8:30 and that tardies dropped by a third the following semester. A student concludes: "The later start time caused fewer students to be late." Is this inference fully warranted? Explain.
Answer: Not fully — the passage shows a sequence, not a mechanism, so the claim should be hedged to something like 'the later start likely contributed to fewer tardies' unless the report rules out other explanations.
Two events in order do not by themselves establish that the first produced the second. Plenty of rival readings fit the same evidence: the school may have changed how tardies are recorded, added a new attendance policy, or the semester may have had milder weather affecting bus routes. A reader can legitimately infer a connection, because the timing is tight and there is a plausible mechanism — more sleep and a schedule better matched to bus service. But 'caused' asserts more than the evidence delivers. Writing 'likely contributed to' keeps the inference honest, and pointing out what additional information would settle the question shows genuine control of the material.
FAQ
- What is the difference between an inference and a prediction?
- A prediction says what will happen next; an inference says what is true now but unstated. In nonfiction you are almost always inferring, not predicting — figuring out an author's unstated attitude, an unmentioned cause, or the significance of a detail. Predictions about future events usually cannot be supported by lines already on the page, which is why they tend to slip into overreach.
- How much background knowledge am I allowed to bring in?
- Bring in whatever a general reader would accept without argument — that printers need paper, that a drought reduces crop yields, that a bus schedule limits when people can travel. Do not bring in contested opinions, facts from a different article, or assumptions about a group of people. A quick test: if a classmate could reasonably say 'that's just your view,' it is opinion and does not belong in the inference.
- How do I cite lines if my copy of the text has no line numbers?
- Use paragraph numbers, section headings, or a short signal phrase that locates the quotation — 'in the third paragraph,' 'in the section on funding,' 'in the sentence describing the 2019 vote.' The goal is that a reader can find your evidence quickly. Always quote the words themselves as well; naming a location without quoting leaves your reader guessing which part you meant.
- Can two students draw different inferences from the same passage and both be right?
- Yes, as long as each inference is licensed by cited lines and neither contradicts the text. Rich nonfiction supports several defensible readings. What separates them is quality of support: the stronger inference points to more specific evidence, explains its reasoning, and acknowledges the rival reading instead of ignoring it. Two readings can coexist; a reading with no line behind it cannot survive.
Learn this with a teacher, not a page
The Crimsora tutor teaches Inferences & Supporting Evidence live — explaining on a whiteboard, asking you questions, and adapting to where you get stuck.