Skip to contents

Example 9.4 from Generalized Linear Mixed Models: Modern Concepts, Methods and Applications by Walter W. Stroup (p-288)

Source: R/Exam9.4.R
Exam9.4.Rd

Exam9.4 Relationship between BLUP and Fixed Effect Estimators

References

  1. Stroup, W. W. (2012). Generalized Linear Mixed Models: Modern Concepts, Methods and Applications. CRC Press.

See also

DataSet9.4

Author

  1. Muhammad Yaseen (myaseen208@gmail.com)

  2. Adeela Munawar (adeela.uaf@gmail.com)

Examples

data(DataSet9.4)
DataSet9.4$a <- factor(x = DataSet9.4$a)
DataSet9.4$b <- factor(x = DataSet9.4$b)

library(lmerTest)
Exam9.4lmer <- lmer(y ~ a + (1|b) + (1|a/b), data = DataSet9.4)
#> Warning: unable to evaluate scaled gradient
#> Warning:  Hessian is numerically singular: parameters are not uniquely determined
#> Warning: Model may not have converged with 1 eigenvalue close to zero: 4.4e-10
summary(Exam9.4lmer)
#> Linear mixed model fit by REML. t-tests use Satterthwaite's method [
#> lmerModLmerTest]
#> Formula: y ~ a + (1 | b) + (1 | a/b)
#>    Data: DataSet9.4
#> 
#> REML criterion at convergence: 155.2
#> 
#> Scaled residuals: 
#>      Min       1Q   Median       3Q      Max 
#> -1.51149 -0.48875 -0.06897  0.59000  1.38365 
#> 
#> Random effects:
#>  Groups   Name        Variance Std.Dev.
#>  b:a      (Intercept) 6.256    2.501   
#>  b        (Intercept) 4.024    2.006   
#>  a        (Intercept) 2.380    1.543   
#>  Residual             3.647    1.910   
#> Number of obs: 32, groups:  b:a, 16; b, 8; a, 2
#> 
#> Fixed effects:
#>             Estimate Std. Error     df t value Pr(>|t|)    
#> (Intercept)   14.819      1.973 28.295   7.510 3.29e-08 ***
#> a2            -2.544      2.604 22.996  -0.977    0.339    
#> ---
#> Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
#> 
#> Correlation of Fixed Effects:
#>    (Intr)
#> a2 -0.660
#> optimizer (nloptwrap) convergence code: 0 (OK)
#> unable to evaluate scaled gradient
#>  Hessian is numerically singular: parameters are not uniquely determined
#> 
library(emmeans)
emmeans(Exam9.4lmer, spec = ~a)
#>  a emmean   SE   df lower.CL upper.CL
#>  1   14.8 1.97 83.5    10.89     18.7
#>  2   12.3 1.97 83.5     8.35     16.2
#> 
#> Degrees-of-freedom method: kenward-roger 
#> Confidence level used: 0.95