我不知道这是否曾经被问过,但是我什么也没找到。我的问题是,是否有人可以提供一个很好的参考,以学习如何获得混合效应模型中固定和随机因素分别解释的方差比例。
4
Good question, but I don't have (a reference for) a good answer. There's more than one level of variation in mixed models, so there's more than one component of variance to explain, plus it's debateable whether random effects can really be said to 'explain' variance. I think the whole concept of 'proportion of variance explained' is less useful in mixed models.
—
onestop
Gelmans的“贝叶斯ANOVA”方法也可能有用。
—
N Brouwer'5