Making Inferences

What You Need to Know

An inference is a logical conclusion based on evidence + prior knowledge. It's NOT directly stated in the text—you figure it out.

The Inference Formula:

Text Evidence + Background Knowledge = Inference

Text evidence must do most of the work. Background knowledge may connect clues, but it cannot replace missing evidence. If two answers are possible, prefer the one explained by more independent details and requiring fewer assumptions.

Four Evidence-Based Reasoning Products

A prediction is not a guess about anything that could happen. It projects the most supported next event. On every task, point to the words that support the answer and the detail that rules out a tempting rival.

Think of it as being a detective: You gather clues (evidence) and use what you already know to solve the mystery (reach a conclusion).

🧠 Memory Trick

"What does the text SUGGEST but not SAY?"

If you can point to a sentence that directly states it, it's NOT an inference—it's a detail.

Inference = Reading BETWEEN the lines Detail = Reading the lines themselves

How to Make Inferences

Step-by-Step:

1. Read carefully — notice details, word choice, what's included AND excluded
2. Ask: "What does this suggest?"
3. Use your background knowledge — what do you know about this topic/situation?
4. Check your inference — is it supported by evidence in the text?

5. Avoid over-reaching — stick close to what the text supports

6. Name the evidence pair — strong inferences often depend on two clues, such as an action plus a setting or a trend plus a comparison.

7. Check the boundary — “the owner may be preparing to close” is safer than “the business will permanently close tomorrow” unless the text supplies that certainty.

Types of Inferences

An Evidence Ladder

The west gallery will close at 4 p.m. today. Crates marked “Fragile” stand beside its doorway, and the new exhibition opens Monday.

• Directly stated: The gallery closes early.
• Supported inference: Staff are probably changing the exhibition.
• Supported prediction: Visitors will likely be redirected after 4 p.m.
• Overreach: The museum is selling its entire collection.

The last claim is imaginable, but nothing bridges the crates to a sale. Good distractors often borrow one real clue and then stretch it beyond its scope.

✏️ Worked Examples

Example 1: Basic Inference

What we're looking for: What does the text SUGGEST but not directly SAY?

Text: "Maria pushed her food around her plate but didn't take a single bite. She kept glancing at the clock on the wall."

Step 1 — Read the sentences and notice the specific details. Two behaviors: (1) playing with food but not eating, (2) repeatedly checking the clock. Both are described deliberately — the author chose these details for a reason.

Step 2 — Ask: What do these behaviors normally mean in real life? When someone pushes food around without eating, they're usually not hungry OR distracted/anxious. When someone keeps checking the clock, they're usually waiting for something or worried about time.

Step 3 — Combine the clues into a logical inference. Both behaviors together — not eating AND clock-watching — suggest Maria's mind is elsewhere. She's either anxious, distracted, or waiting for something.

Step 4 — Check: Is this supported by the text or am I making it up? Yes — the specific behaviors the author chose (food-pushing, clock-glancing) are textbook signs of distraction or anxiety. Supported. ✓

What we CAN infer:

• Maria isn't hungry OR is too anxious/distracted to eat
• She's waiting for something or worried about time
• Something is bothering her

Step 5 — Rule out over-reaching inferences. What we CANNOT infer (not enough evidence):

• She's sick — possible, but nothing in the text points there
• The food tastes bad — she hasn't even tried it; we can't blame the food
• She's meeting someone specific — nothing says who or what she's waiting for

Example 2: Infer a concern without ruling out other needs

Text: “Before the debate, Jo asked the organizer three times whether her teammate had arrived. Each time she checked the empty chair beside her and reread their shared notes.”

Repeated questions plus attention to the teammate's chair suggest that the teammate's absence is on Jo's mind. It is reasonable to infer that she wants the teammate present. The text does not establish whether Jo is frightened, angry or physically unwell; do not rule out needs the passage never discusses.

Example 3: Research Context Inference

What we're looking for: What does this data suggest — and what limits what we can conclude?

Text: "The experimental group showed a 45% improvement, while the control group showed only 3% improvement. However, the study included only 12 participants."

Step 1 — Read the numbers. 45% vs. 3% — that's a 42-point gap. On the surface, that looks impressive.

Step 2 — Apply what you know about research. But hold on — "only 12 participants." Twelve participants provide limited evidence for generalizing. The passage does not describe assignment, measurement or uncertainty, so the size of the reported difference alone does not establish a reliable treatment effect.

Step 3 — Balance both pieces of information. The treatment looks effective (big difference between groups) BUT the sample size undermines confidence in that result (too few people to be sure it's real).

Inferences:

• The treatment appears promising (45% vs. 3% is the reported gap, not proof of statistical significance)
• BUT the small sample size (12 participants) limits how much we can trust this
• More research with a larger group is needed before drawing strong conclusions

TEAS Tip: When a passage shows both positive results AND a limitation, the best inference usually acknowledges BOTH — not just the positive result.

Example 4: Separate growth from its possible causes

Text: “Applications to the evening design course rose from 80 to 120 while available seats stayed at 60. A waitlist formed in both years. Tuition and the admission process did not change.”

The course received more applications for the same number of places, so competition for a seat likely increased. The passage does not explain why applications rose or establish graduates' salaries, course quality or demand for designers. Programs, applications, enrollment and jobs are different quantities: never silently substitute one for another.

💡 Pro Tips

• Look for emotional clues: Body language, word choice, and actions reveal feelings
• Notice what's missing: Sometimes what ISN'T said is important
• Consider context: Time period, setting, and situation inform your inference
• Stay grounded: Your inference should be the most LIKELY explanation, not just a possible one
• On the TEAS: Wrong answers often go too far beyond the text

⚠️ Common Mistakes to Avoid

• Confusing inference with stated information: If the text says it directly, it's not an inference
• Over-inferring: Drawing conclusions not supported by the text
• Ignoring evidence: Making inferences based on your opinions, not the text
• Choosing the most dramatic option: TEAS inferences are usually reasonable, not extreme

Quick Reference

Making Inferences

Inference vs. Stated Information

Red Flags for Wrong Answers

• Too extreme ("definitely," "always," "proves")
• Not supported by text evidence
• Based on outside knowledge only
• Contradicts information in the passage

Inference skills unlocked! 💪 Next up: Drawing Conclusions