复合对称情况下（0 + factor | group）和（1 | group）+（1 | group：factor）随机效应规格的等价关系

道格拉斯·贝茨（Douglas Bates）指出，以下模型是等效的：“如果向量值随机效应的方差-协方差矩阵具有一种特殊形式，称为复合对称”（本演示文稿中的幻灯片91）：

m1 <- lmer(y ~ factor + (0 + factor|group), data)
m2 <- lmer(y ~ factor + (1|group) + (1|group:factor), data)

具体而言，贝茨使用以下示例：

library(lme4)
data("Machines", package = "MEMSS")

m1a <- lmer(score ~ Machine + (0 + Machine|Worker), Machines)
m2a <- lmer(score ~ Machine + (1|Worker) + (1|Worker:Machine), Machines)

具有相应的输出：

print(m1a, corr = FALSE)

Linear mixed model fit by REML ['lmerMod']
Formula: score ~ Machine + (0 + Machine | Worker)
   Data: Machines
REML criterion at convergence: 208.3112
Random effects:
 Groups   Name     Std.Dev. Corr     
 Worker   MachineA 4.0793            
          MachineB 8.6253   0.80     
          MachineC 4.3895   0.62 0.77
 Residual          0.9616            
Number of obs: 54, groups:  Worker, 6
Fixed Effects:
(Intercept)     MachineB     MachineC  
     52.356        7.967       13.917  

print(m2a, corr = FALSE)

Linear mixed model fit by REML ['lmerMod']
Formula: score ~ Machine + (1 | Worker) + (1 | Worker:Machine)
   Data: Machines
REML criterion at convergence: 215.6876
Random effects:
 Groups         Name        Std.Dev.
 Worker:Machine (Intercept) 3.7295  
 Worker         (Intercept) 4.7811  
 Residual                   0.9616  
Number of obs: 54, groups:  Worker:Machine, 18; Worker, 6
Fixed Effects:
(Intercept)     MachineB     MachineC  
     52.356        7.967       13.917

谁能以直观的方式解释模型之间的差异以及如何m1简化m2（给定复合对称性）？

在此示例中，三个机器（A，B，C）和六个工人的每种组合都有三个观测值。我将使用表示n维恒等矩阵，使用1 n表示n维矢量为1。假设y是观察的向量，我将假设它是由工人排序的，然后是机器然后进行复制。设μ为对应的期望值（例如，固定效应），设γ为与期望值的组特定偏差的向量（例如，随机效应）。以γ为条件，可以写出y的模型：$I_n$$n$$1_n$$n$$y$$\mu$$\gamma$$\gamma$$y$

$$y \sim \mathcal{N}(\mu + \gamma, \sigma^2_y I_{54})$$

$\sigma^2_y$

$\gamma$

$$y \sim \mathcal{N}(\mu, \sigma^2_y I_{54} + \Sigma)$$

$\Sigma$$\gamma$$\Sigma$

m1

$$\gamma = Z \theta$$

$Z = I_{18} \otimes 1_3$$\theta^T = [\theta_{1,A}, \theta_{1,B}, \theta_{1,C} \dots \theta_{6,A}, \theta_{6,B}, \theta_{6,C}]$

$$\theta \sim \mathcal{N}(0, I_6 \otimes \Lambda)$$

$\Lambda$$\Lambda$$\sigma_\theta$$\tau$

$$\Lambda = \left[\begin{matrix} \sigma^2_\theta + \tau^2 & \tau^2 & \tau^2 \\ \tau^2 & \sigma^2_\theta + \tau^2 & \tau^2 \\ \tau^2 & \tau^2 & \sigma^2_\theta + \tau^2 \end{matrix}\right]$$

$\Lambda$

$\Sigma = Z (I_6 \otimes \Lambda) Z^T$$\sigma^2_\theta + \tau^2 + \sigma^2_y$$i, j$$u, v$

$$\mathrm{cov}(y_{i,u}, y_{j,v}) = \begin{cases} 0 & \text{if } i\neq j \\ \tau^2 & \text{if } i=j, u\neq v \\ \sigma^2_\theta + \tau^2 & \text{if } i=j, u=v \end{cases}$$

对于您而言m2，随机效应分解为：

$$\gamma = Z \omega + X \eta$$

$X = I_6 \otimes 1_9$$\omega^T = [\omega_{1,A}, \omega_{1,B}, \omega_{1,C}, \dots, \omega_{6,A}, \omega_{6,B}, \omega_{6,C}]$$\eta^T = [\eta_{1}, \dots, \eta_{6}]$

$$\eta \sim \mathcal{N}(0, \sigma^2_\eta I_6)$$ $$\omega \sim \mathcal{N}(0, \sigma^2_\omega I_{18})$$ $\sigma_\eta^2, \sigma_\omega^2$

m2$\Sigma = \sigma^2_\omega Z Z^T + \sigma^2_\eta X X^T$$\sigma^2_\omega + \sigma^2_\eta + \sigma^2_y$$i, j$$u, v$

$$\mathrm{cov}(y_{i,u}, y_{j,v}) = \begin{cases} 0 & \text{if } i\neq j \\ \sigma_\eta^2 & \text{if } i=j,u\neq v \\ \sigma^2_\omega + \sigma^2_\eta & \text{if } i=j,u=v \end{cases}$$

$\sigma^2_\theta \equiv \sigma^2_\omega$$\tau^2 \equiv \sigma^2_\eta$ m1

简洁不是我的强项：这仅仅是一个漫长而复杂的说法，每个模型都具有两个用于随机效应的方差参数，并且只是同一“边际”模型的两种不同编写方式。

在代码中...

sigma_theta <- 1.8
tau         <- 0.5
sigma_eta   <- tau
sigma_omega <- sigma_theta
Z <- kronecker(diag(18), rep(1,3))
rownames(Z) <- paste(paste0("worker", rep(1:6, each=9)), 
                     rep(paste0("machine", rep(1:3, each=3)),6))
X <- kronecker(diag(6), rep(1,9))
rownames(X) <- rownames(Z)
Lambda <- diag(3)*sigma_theta^2 + tau^2

# marginal covariance for m1:
Z%*%kronecker(diag(6), Lambda)%*%t(Z)
# for m2:
X%*%t(X)*sigma_eta^2 + Z%*%t(Z)*sigma_omega^2