Use of the Box-Cox Transformation with Binary Response Models
将Box-Cox幂变换应用于比值比，推广了逻辑斯蒂模型并参数化了一类拟合不足，同时考虑了剂量-响应问题中设计变量的变换。
The power transformation suggested by Box & Cox (1964) is applied to the odds ratio to generalize the logistic model and to parameterize a certain type of lack of fit. Transformation of the design variable within the context of the dose-response problem is also considered.