Skip to Main Content

HLM Textbook Examples Introduction to Multilevel Modeling by Kreft and de Leeuw Chapter 3: Varying and Random Coefficient Models


Table 3.3 on page 48.

We will use imm10.sav and imm10_lev2.sav for this example. Variable schid is the ID variable that links the level-1 data file with the level-2 data
file. We assume that an SSM file has been created based on these two files.

c

Sigma_squared =     43.07100
 Tau
 INTRCPT1,B0     69.30426     -31.76171 
 HOMEWORK,B1    -31.76171      22.45254 
Tau (as correlations)
 INTRCPT1,B0  1.000 -0.805
 HOMEWORK,B1 -0.805  1.000
 ----------------------------------------------------
  Random level-1 coefficient   Reliability estimate
 ----------------------------------------------------
  INTRCPT1, B0                        0.920
  HOMEWORK, B1                        0.928
 ----------------------------------------------------
The value of the likelihood function at iteration 6 = -8.819772E+002
 The outcome variable is     MATH
 Final estimation of fixed effects:
 ----------------------------------------------------------------------------
                                       Standard             Approx.
    Fixed Effect         Coefficient   Error      T-ratio   d.f.     P-value
 ----------------------------------------------------------------------------
 For       INTRCPT1, B0
    INTRCPT2, G00          44.770591   2.743644    16.318         9    0.000
 For HOMEWORK slope, B1
    INTRCPT2, G10           2.040469   1.554194     1.313         9    0.222
 ----------------------------------------------------------------------------
 The outcome variable is     MATH
 Final estimation of fixed effects
 (with robust standard errors)
 ----------------------------------------------------------------------------
                                       Standard             Approx.
    Fixed Effect         Coefficient   Error      T-ratio   d.f.     P-value
 ----------------------------------------------------------------------------
 For       INTRCPT1, B0
    INTRCPT2, G00          44.770591   2.602659    17.202         9    0.000
 For HOMEWORK slope, B1
    INTRCPT2, G10           2.040469   1.474242     1.384         9    0.200
 ----------------------------------------------------------------------------
The robust standard errors are appropriate for datasets having a moderate to
large number of level 2 units.  These data do not meet this criterion.
 Final estimation of variance components:
 -----------------------------------------------------------------------------
 Random Effect           Standard      Variance     df    Chi-square  P-value
                         Deviation     Component
 -----------------------------------------------------------------------------
 INTRCPT1,       U0        8.32492      69.30426     9     153.18126    0.000
 HOMEWORK slope, U1        4.73841      22.45254     9     133.36740    0.000
  level-1,       R         6.56285      43.07100
 -----------------------------------------------------------------------------
 Statistics for current covariance components model
 --------------------------------------------------
 Deviance                       = 1763.954337
 Number of estimated parameters = 4

Table 3.4 on page 50.

Variable PUBLIC is included as level-2 predictor.

c

 Sigma_squared =     42.96022
 Tau
 INTRCPT1,B0     45.84914     -32.24989 
 HOMEWORK,B1    -32.24989      23.96176 
Tau (as correlations)
 INTRCPT1,B0  1.000 -0.973
 HOMEWORK,B1 -0.973  1.000
 ----------------------------------------------------
  Random level-1 coefficient   Reliability estimate
 ----------------------------------------------------
  INTRCPT1, B0                        0.885
  HOMEWORK, B1                        0.933
 ----------------------------------------------------
The value of the likelihood function at iteration 11 = -8.710642E+002
 Final estimation of fixed effects:
 ----------------------------------------------------------------------------
                                       Standard             Approx.
    Fixed Effect         Coefficient   Error      T-ratio   d.f.     P-value
 ----------------------------------------------------------------------------
 For       INTRCPT1, B0
    INTRCPT2, G00          58.040447   2.944241    19.713         8    0.000
     MPUBLIC, G01         -14.660674   2.109003    -6.951         8    0.000
 For HOMEWORK slope, B1
    INTRCPT2, G10           1.952488   1.598905     1.221         9    0.253
 ----------------------------------------------------------------------------
The robust standard errors cannot be computed for this model.
 Final estimation of variance components:
 -----------------------------------------------------------------------------
 Random Effect           Standard      Variance     df    Chi-square  P-value
                         Deviation     Component
 -----------------------------------------------------------------------------
 INTRCPT1,       U0        6.77120      45.84914     8      81.98591    0.000
 HOMEWORK slope, U1        4.89507      23.96176     9     133.96327    0.000
  level-1,       R         6.55440      42.96022
 -----------------------------------------------------------------------------
 Statistics for current covariance components model
 --------------------------------------------------
 Deviance                       = 1742.128440
 Number of estimated parameters = 4

Table 3.5 on page 51.

Variable PUBLIC is included as level-2 predictor for the slope of
variable HOMEWORK, hence an interaction between the two.

c

 Sigma_squared =     42.95762
 Tau
 INTRCPT1,B0     51.85064     -36.71407 
 HOMEWORK,B1    -36.71407      27.26808 
Tau (as correlations)
 INTRCPT1,B0  1.000 -0.976
 HOMEWORK,B1 -0.976  1.000
 ----------------------------------------------------
  Random level-1 coefficient   Reliability estimate
 ----------------------------------------------------
  INTRCPT1, B0                        0.896
  HOMEWORK, B1                        0.940
 ----------------------------------------------------
The value of the likelihood function at iteration 8 = -8.693684E+002
 Final estimation of fixed effects:
 ----------------------------------------------------------------------------
                                       Standard             Approx.
    Fixed Effect         Coefficient   Error      T-ratio   d.f.     P-value
 ----------------------------------------------------------------------------
 For       INTRCPT1, B0
    INTRCPT2, G00          59.210218   7.408165     7.993         8    0.000
     MPUBLIC, G01         -15.965966   7.830068    -2.039         8    0.075
 For HOMEWORK slope, B1
    INTRCPT2, G10           1.094640   5.242788     0.209         8    0.840
     MPUBLIC, G11           0.951377   5.541947     0.172         8    0.868
 ----------------------------------------------------------------------------
The robust standard errors cannot be computed for this model.
 Final estimation of variance components:
 -----------------------------------------------------------------------------
 Random Effect           Standard      Variance     df    Chi-square  P-value
                         Deviation     Component
 -----------------------------------------------------------------------------
 INTRCPT1,       U0        7.20074      51.85064     8      81.54750    0.000
 HOMEWORK slope, U1        5.22188      27.26808     8     129.57845    0.000
  level-1,       R         6.55421      42.95762
 -----------------------------------------------------------------------------
 Statistics for current covariance components model
 --------------------------------------------------
 Deviance                       = 1738.736793
 Number of estimated parameters = 4