求助大家,做了一个横断面研究,就是简单的P物质在A病与正常对照组差异,两组年龄性别是匹配的。做的时候也没注意,T检验有意义后,用协方差分析控制了B,C,D,性别年龄这些协变量。然后审稿人提的问题是不应该把年龄,性别当做协变量,并建议我用混合线性模型来控制变量,并且把年龄性别当做随机变量。我就不知道该怎么做了,结果的表达也不清楚,求助大家。审稿人的意见如下:It is also illegitimate to adjust for matching variables. Thus, the authors should not use age or sex as adjustment covariates because the two groups are already matched by those factors. If the authors want to compare groups after adjusting for covariates (apart from age and sex) then they should use a mixed-effects regression with matching variables (age and sex) as random variables.
我自己试了一下,不知道该怎么把年龄性别当做随机变量。