3.2 Inferencing
Key Takeaways
- Inferencing on AL means drawing a conclusion that is strongly licensed by stated content plus necessary implication — not inventing unsupported world knowledge
- AQL items often ask what can be inferred, what the writer implies, or which statement must be true if the passage is accurate
- Distinguish bridging inferences (connecting two stated ideas), elaborative inferences (filling a plausible gap), and predictive inferences (what follows next) — each has a different evidence bar
- Common traps over-generalise local claims, reverse cause and effect, or treat possibilities as certainties
Inferencing: reading what the text licenses
If communicative function asks what a sentence does, inferencing asks what a reader may conclude from what is stated and what is necessarily or strongly implied. Academic Literacy on the AQL deliberately rewards students who can operate one step beyond verbatim matching — the same skill university lectures and research articles demand when writers leave connections partly unspoken.
An inference is not a free association. On AL, a defensible inference must be text-licensed: if the passage is true, the conclusion should hold (or be the best supported option), without importing controversial outside claims.
Types of inference you will meet
| Inference type | Reader move | Evidence bar on AL | Example trigger |
|---|---|---|---|
| Literal restatement | Rephrase what is already explicit | Lowest — often a foil, not a true “infer” item | Synonym swap of a stated fact |
| Bridging | Connect two stated pieces so the text coheres | High — both pieces must be in the passage | Cause in sentence 1 + effect in sentence 3 |
| Elaborative | Fill a gap the writer leaves obvious | Medium — must be the option closest to the text | Unnamed motive strongly suggested by actions |
| Causal | Infer why something occurred or what produced an outcome | High — direction of cause must match the text | “After the policy change, attendance rose” |
| Predictive | Infer a likely next outcome | Medium-high — avoid certainty language the text does not justify | Trends described without a final number |
| Attitude / stance | Infer the writer’s evaluation | Medium — rely on evaluative vocabulary and concessions | Cautious verbs like appear, may, provisionally |
AL stems commonly use wording such as:
- “It can be inferred that…”
- “The passage implies that…”
- “Which of the following is a reasonable conclusion from the text?”
- “The writer’s comments suggest that…”
When you see suggests or implies, the correct answer is usually one careful step from the wording — not a dramatic new thesis.
Mini passage C — tutorial attendance (invented)
In the first semester, the Department of Geography moved weekly tutorials from Friday afternoons to Tuesday mornings. Average attendance rose from 54% to 79%. Student comments collected on an anonymous form frequently mentioned clash with part-time weekend work and long taxi journeys after Friday lectures. The department had not altered tutorial content, assessment weightings, or the tutors assigned to groups. A minority of responses still complained that rooms were overcrowded after the change.
Stated facts: time slot changed; attendance increased; content/assessment/tutors unchanged; many comments cite weekend work and Friday travel; some overcrowding complaints remain.
Licensed inferences:
- Scheduling — not tutorial substance — is the most plausible driver of the attendance rise (bridging + causal: same content, different time, higher attendance, reasons pointing to Friday friction).
- At least some students previously skipped Friday tutorials because of work or transport constraints (elaborative but tightly tied to comment themes).
- The change did not solve every satisfaction issue (overcrowding remains).
Not licensed:
- “Tuesday tutorials are pedagogically superior.” (content unchanged)
- “All students prefer mornings.” (over-generalises)
- “Taxi journeys became shorter.” (not stated; Friday timing was the issue)
- “Overcrowding caused the attendance increase.” (complaints persist after the rise; causation reversed or confused)
Worked reasoning for a typical stem
Stem: “It can be inferred that the attendance increase was most likely related to…”
Option analysis pattern:
- Eliminate options that invent new variables (new tutors, easier tests).
- Eliminate options that contradict “content unchanged.”
- Prefer the option that links the time-slot change to the comment themes about Friday work and travel.
- Treat “overcrowding” as a remaining complaint, not the cause of higher attendance.
That process mirrors how AQL AL items are built: several options are locally true topics, but only one is the inference the passage warrants.
Based on Mini passage C, which conclusion is best supported?
An AL option claims: “Student transport times became shorter after the change.” Why is this a weak inference from Mini passage C?
Applying insights beyond the literal — without leaving the text
University reading often requires you to apply a passage insight to a slightly new case. AL items may ask which additional scenario is most consistent with the writer’s findings or principles. The rule is still textual: the new scenario must preserve the same causal or logical structure.
Mini passage D — library fines (invented)
When the university replaced daily overdue fines with a flat “extended use” fee charged only after 14 days, short-term returns increased and confrontational disputes at the desk declined. Librarians reported that students who previously kept books “one more day” to finish a chapter now returned them within the fine-free window. However, a small group of long-term borrowers who routinely kept items for weeks showed little change in behaviour.
Literal facts: fine structure changed; short-term returns improved; disputes fell; chronic long-term borrowers largely unaffected.
Application-style inference: a policy that removes small daily penalties but keeps a later threshold cost can nudge borderline late returners more than habitual long-term keepers. An AL option that says “all overdue behaviour disappeared” fails; an option that says “students near the old daily-fine margin were more responsive than chronic keepers” fits.
Inference traps AQL loves
| Trap pattern | What it looks like | Fix |
|---|---|---|
| Absolute language | all, never, proves, only possible cause | Prefer most likely, suggests, consistent with when the text is cautious |
| Scope blow-up | Local campus finding → national law | Keep the geographic/population scope of the passage |
| Causal reversal | Effect treated as cause | Check time order and the writer’s linking words |
| Opposite polarity | Turns a concession into the main claim | Re-read although / however frames |
| Outside expertise | Imports textbook knowledge the passage never uses | If the passage does not need it, discard it |
| True but irrelevant | Correct world fact unrelated to the question | Ask “does the passage support this conclusion?” |
Stance inferences
Sometimes the “fact” is the writer’s attitude. Words such as provisionally, under-studied, promising but, merely, and failed to signal evaluation. Inferring stance means naming that evaluation accurately — cautious optimism is not the same as rejection; a minority complaint is not the author’s main conclusion.
Exam-day routine for inference items
- Underline the constraints in the stem (must be true, most likely, implies).
- List 2–3 stated anchors from the relevant paragraph.
- Predict the inference in your own short clause before looking at options.
- Eliminate absolutes, reversals, and outside-knowledge options.
- Choose the remaining option that stays inside the passage’s scope and polarity.
Inferencing sits at the centre of AL meaning-making: it is how academic readers turn sentences into knowledge structures. Pair it with communicative function (what the sentence does) and you will also reason more cleanly about vocabulary in context, because many “word meaning” items are really miniature inference tasks about how a term is being used.
From Mini passage D, which application is most consistent with the reported findings?
A passage says a finding is “promising but provisional until broader monitoring is published.” What stance inference is most accurate?