Factual Knowledge of Oregon College Students

摘要

This paper explores usefulness of algorithmic decomposition in estimatinguncertain quantities. In a series of experiments, experimental questions were used that people are unlikely to know the precise answer to, but could estimate from other facts or estimates. The three studies conducted indicate that the subjects who volunteer for our experiments frequently lack knowledge of basic facts. A mild manipulation intended to persuade the subjects to reevaluate and improve their estimates had little effect. The data further suggest that efforts to develop estimation aids based on algorithmic decomposition techniques may fail, not because people are unable to make such decompositions but because they have and hold extensive stores of misinformation. When this misinformation is brought to bear on estimation problems, performance may not improve but become worse than previously expected. Keywords: Algorithmic decomposition; Acquisition; Subjective estimates; Experimental data. (cp)