摘要: A bstract . Regression models have been used in the social sciences at least since 1899, when Yule published a paper on causes of pauperism. are now to make causal arguments wide variety applications, and it is perhaps time evaluate results. No definitive answers can be given, but this chapter takes rather negative view. Snow's work cholera presented as success story for scientific reasoning based nonexperimental data. Failure stories also discussed, comparisons may provide some insight. In particular, suggests that statistical technique seldom an adequate substitute good design, relevant data, testing predictions against reality settings Introduction his changes “out-relief” cause pauperism: He argued providing income support outside poorhouse increased number people relief. At present, regression science crude four-point scale useful: usually works, although (like anything else) imperfect sometimes go wrong. works hands skillful practitioners, isn't suitable routine use. might work, hasn't yet. can't work.