Causal Stereotypes and Conditional Probability David Bergner, Ph. D. p. 1 Causal Stereotypes ▪Cab involved in accident, less-than-perfect witness ▪85% Green, 15% Blue ▪Witness says Blue, correct 80% of the time. ▪What is the probability the cab was blue? p. 2 Causal Stereotypes (cont.) ▪ Cab involved in accident, less-than-perfect witness ▪ 85% of accidents involve Green Cabs, and only 15% involve blue cabs. ▪ Witness says Blue, correct 80% of the time. ▪ What is the probability the cab was blue? Insights: ▪ Statistical base rates are generally underweighted or neglected altogether when other information, or a representative story, is available. ▪ Causal base rates are treated as applying to the specific case at hand, and are easily combined with other information in the narrative. ▪ System 1 can deal with stories in which the elements are causally linked, but is weak at statistical reasoning. p. 3 Summary ▪Note that causal stereotyping actually improves the intuitive statistics in the Green/Blue cab story. ▪System 1 can deal with stories in which the elements are causally linked, but is weak at statistical reasoning. ▪Stereotypes are good, because they are usually right, but compromise social justice due to the unacceptable cost of “false positives” ▪Dr. Martin Luther King, “Injustice anywhere is a threat to justice everywhere.” p. 4
Purchase answer to see full attachment