结论当有自变量不符合比例优势假定条件时，偏比例优势模型是更为合理的选择。

Conclusion The partial proportional odds model is more appropriate when proportional odds assumption was violated.

youdao

方法利用在陕西某地开展的一项前瞻性研究数据，分别采用比例优势模型和偏比例优势模型进行分析，对二者的结果进行比较。

Methods the proportional odds model and partial proportional odds model were used for the data from a cohort study in Shaanxi Province, and the results were compared.

youdao