외고 상위권 논증·추론 도전 세트

A coastal town rated evacuation plans by the number of residents reaching a shelter within twenty minutes. Plan A scored higher than Plan B in a summer rehearsal. Officials initially recommended A for the entire year. The rehearsal, however, had taken place when the northern bridge was open and most elderly residents were away at a community event. In winter, the bridge closes regularly, while those residents are more likely to be at home. None of this makes the summer result false. It makes the result an answer to a narrower question than the officials had asked. A useful measure becomes misleading when the conditions supporting it are quietly removed from the conclusion.

A coastal town rated evacuation plans by the number of residents reaching a shelter within twenty minutes. Plan A scored higher than Plan B in a summer rehearsal. Officials initially recommended A for the entire year. The rehearsal, however, had taken place when the northern bridge was open and most elderly residents were away at a community event. In winter, the bridge closes regularly, while those residents are more likely to be at home. None of this makes the summer result false. It makes the result an answer to a narrower question than the officials had asked. A useful measure becomes misleading when the conditions supporting it are quietly removed from the conclusion.

A coastal town rated evacuation plans by the number of residents reaching a shelter within twenty minutes. Plan A scored higher than Plan B in a summer rehearsal. Officials initially recommended A for the entire year. The rehearsal, however, had taken place when the northern bridge was open and most elderly residents were away at a community event. In winter, the bridge closes regularly, while those residents are more likely to be at home. None of this makes the summer result false. It makes the result an answer to a narrower question than the officials had asked. A useful measure becomes misleading when the conditions supporting it are quietly removed from the conclusion.

Two forecasting systems predict demand at a food bank. System P has a smaller average error across all branches. At the smallest branch, however, it repeatedly underestimates demand, whereas System Q does not. The director rejects an immediate switch to Q everywhere: the small branch represents only one unusual pattern, and Q performs poorly elsewhere. She instead asks whether the factors that distinguish that branch can be identified before a forecast is made. If so, different systems could be used under different conditions. The point is not to abandon comparison, but to examine where an aggregate ranking stops being informative.

Two forecasting systems predict demand at a food bank. System P has a smaller average error across all branches. At the smallest branch, however, it repeatedly underestimates demand, whereas System Q does not. The director rejects an immediate switch to Q everywhere: the small branch represents only one unusual pattern, and Q performs poorly elsewhere. She instead asks whether the factors that distinguish that branch can be identified before a forecast is made. If so, different systems could be used under different conditions. The point is not to abandon comparison, but to examine where an aggregate ranking stops being informative.