ABSTRACT
Not long ago, a visiting speaker, a distinguished cognitive scientist, came to Princeton. He visited our laboratory, and one of us, by way of entertaining him gave him a tricky inferential problem, one that almost everyone gets wrong. He got it wrong, too. Later, he explained:
Despite his insight-a metacognitive one, our visitor failed to come up with an appropriate strategy for dealing with the inference-as indeed had we when we first attempted the inference in checking the output of a computer program.
