Which Model Presents or Implies Given Information (EMI 302 and 403)

Paired-viewpoint passages give you two or more scientists explaining the same observation with different causes. A question will hand you one fact and ask whose model it belongs to. The trick is that the fact is never random: it's a load-bearing piece of one scientist's argument and irrelevant, or only decorative, to the other's. Your job is to trace each model's chain of reasoning - cause, mechanism, effect - and ask which chain would fall apart if that fact were false.

Work it by elimination, not by memory. Take the fact, then ask each model in turn: does this model's explanation require this to be true? If Scientist 1 says sea level rose because land ice melted into the ocean, then a fact about glacier mass dropping is the engine of that claim - no melting ice, no rise, by that model. Scientist 2's rise comes from ocean water expanding as it warms, so glacier mass is beside the point for them. The fact sorts cleanly into one camp.

This skill sits right next to judging shared facts, which is trickier. When both scientists agree on a number - both accept the ocean warmed by some amount - that number isn't automatically neutral. It becomes a test the moment you pair it with a model's required outcome: does that much warming produce enough thermal expansion to reach the claimed rise? If not, the shared fact quietly argues against that model, even though both scientists started by agreeing on it. Reading models this way - sorting facts by what they support rather than by who said them - is the core move tested across this whole category, including judging whether a prototype or hypothesis succeeded.