Limits of predictability in forecasting in the behavioral sciences

A series of methodological problems in forecasting which arise out of the humanness of the predictor and/or predictee are reviewed. These include (1) perceptual disordering, in which the imperfect nature of data collected by human sensation is investigated, (2) model disordering, in which the imperfect nature of models and theories arising out of human information processing limitations is investigated, and lastly, (3) obtrusive reactive disordering, in which the human tendencies of the predictee, to guess the forecast, and to alter his or her behavior so as to reinforce or interfere with the forecast, are explored. The implications of these methodological dilemmas for forecasting are discussed.