Judgmental versus statistical prediction: information assymmetry and combination rules

Theuse of statistical models against human judgment was compared in terms of predictive accuracy. Current trends recognize that human judgment and statistical models can be complementary using the relative assets of both. Combination of both models utilized their respective strengths in prediction under different conditions. Statistical prediction was more useful when valid cues were few and human judgment was better when numerous valid and redundant cues were available. Performance depended on the amount of contextual data and equal distribution to humans and statistical models.