How to spot it

This one lives on questions that ask how well something is supported: does the figure support the claim, do these results back the hypothesis, is the conclusion justified by the study.

The choice reads the data correctly. It sees the same rise you saw. Then it renders a verdict that the data cannot carry, in either direction.

Overstating is the common half. A trend across four points becomes proof, one experiment becomes established, a correlation becomes a cause. Understating is the other half and it is easier to miss: a figure where the line rises at every single reading gets called inconclusive, or a claim the passage states outright gets called unsupported because the wording differs.

The tell is that you agree with the choice's description of the data and disagree with its last few words. If the sentence would be right with fully changed to partly, or with unsupported changed to supported, you are looking at this pattern rather than a misreading.

What it looks like

Figure: melt volume measured at five temperatures, rising at every step.

Question: Do the data support the claim that warmer conditions increase melt?

Wrong answer: They partially support it, because melt rises at some temperatures but not others.

Correct answer: They support it, because melt rises at every temperature measured.

The wrong choice is not misreading the figure. It is grading it, and grading it wrongly.

Why it works on people

Hedging feels like rigour. Partially supported sounds more careful than supported, and careful sounds more correct, so under pressure the hedge is the comfortable choice even when the figure is unambiguous.

The reverse happens where a claim sounds strong. Warmer conditions increase melt reads like a big statement about the world, so a student looks for a reason the modest data cannot license it, and finds one, when the question only asked about the data in front of them.

Both errors come from the same place: answering about the claim's ambition rather than about the fit between the claim and the evidence shown.

How to beat it

Answer the fit, not the size of the claim. Ask one question: does every relevant data point go the way the claim says. If yes, the answer is supported, however sweeping the claim sounds. If some do and some do not, that is what partial means, and it needs a specific exception you can point at.

Before choosing a hedge, name the exception out loud. If you cannot say which reading breaks the pattern, you have not found partial support, you have found a habit.

Check the direction of the last few words separately from the reading. Say the data description and the verdict as two sentences, then test them one at a time. Most wrong choices here pass the first test and fail the second, which is precisely why reading them as a whole feels fine.

How often it shows up

Counted in September 2026: 1,006 of the 19,627 questions in our practice bank contain this trap, across 1,979 separate answer choices. Some questions use it more than once.

Related patterns

• Claim goes too far: that one overstates what a text says rather than what evidence supports, which is the same failure moved from data to prose.
• Read the wrong row: the error this one is often confused with, where the number came from the wrong place. Here the number is right and the verdict on it is wrong.
• Sounds right, never said: that one asserts something the source never established at all.