作者: Michael W. Gillespie , John Fox
DOI: 10.1177/004912418000800302
关键词:
摘要: This article attempts to explain negatively correlated disturbances in "parallel" simultaneous-equation models which each equation contains essentially the same variables as others but is written for a different position of social system under study. The two general methodological explanations presented are: random error measures endogenous and upward bias calculation (positively signed) reciprocal effects. We show how these effects can be produced by following types specification error: failure control exogenous variables, omission "cross-effects " parallel common causes, contamination b y variables. illustrate consequences last four through analysis contrived data