One Step Beyond the Evidence, Not Two

A text gives you facts, and the question asks what those facts support. The correct choice makes exactly one move past the evidence: it generalizes or connects what's already there without adding anything new. Wrong choices usually make two moves. The first move is the same honest step. The second move smuggles in an assumption the text never earned, usually because it sounds like something that's probably true in the real world. That second move is the trap.

To catch it, draw a line at the edge of what the text actually says, then check each answer against that line. If an answer only restates or combines facts already given, it sits on the right side of the line. If it needs you to believe something extra, like that a pattern in one group holds for a different group, or that a cause explains a correlation, it has stepped over. Counting assumptions is a concrete way to do this: zero new assumptions, keep it; one or more, cut it. This matters because a conclusion can be plausible, even true, in the world, and still be invalid as an inference from this text, since the text itself never supplied the extra piece.

This skill sits right next to central-idea and evidence questions in the same unit. Central-idea questions ask what the whole text is driving at; evidence questions ask what detail supports a claim; inference questions ask what new claim the details support. All three reward staying tied to the page instead of reasoning from outside knowledge.