|
|
|
||||
|
|
|||||
Note that unless otherwise noted, we chose Estimation then RIGLS from the menu option and this estimation method more accurately reproduced the results from the book (than the default of IGLS).
Note that we accidentally mislabeled the dummy variables for time. We named the variables time1 time2 time3 and time4, whereas the book named them time0 time1 time2 and time3.
page 169, table 12.1 model 1, MLwiN model.
and the results
Page 169, table 12.1 model 2, model (note that we made 4 dummy variables from occ calling them time 1, time 2, time 3, time 4, and made cons a random effect at level 1 and 2).
and the results
Page 173, table 12.2 model 3, MLwiN model. We include occ as a random effect across teachers (but without any fixed effect).
Results
We can use a Stata program we wrote called mlcovmat to calculate the estimated covariance matrix shown on page 173. We can get this program like this.
net from http://www.ats.ucla.edu/stat/stata/ado/test net install mlcovmat
We can then run it, supplying the parameters from the multilevel model above
mlcovmat , t0( .233) t1(0.017) t01(-0.052) sigma2( 0.048 ) dim(4)
and we get the following estimated covariance matrix
c1 c2 c3 c4 r1 .281 0 0 0 r2 .181 .194 0 0 r3 .129 .111 .141 0 r4 .077 .076 .075 .122
Model 4, MLwiN model. We added created female and the interaction of female and occ (creating female.occ) using the Model then Main effects and interactions menu option.
and the results are
Table 12.3 (skipped for now)
Page 176, Table 12.4, MLwiN model.
and the results
Table 12.5, page 177, MLwiN model
and the results
Table 12.6, page 180, MLwiN model
and the results
Page 183, Table 12.7 (we do not have data for this example)
Page 185, Table 12.8 (we do not have data for this example)
Page 188, Table 12.9 (we do not have data for this example)
Page 191, Table 12.10 (we do not have data for this example)
Page 193, Figure 12.1 (we do not have data for this example)
Page 194, Table 12.11 (we do not have data for this example)
Page 197, Table 12.12 (skipped for now)
Page 198, Table 12.13 (skipped for now)
UCLA Researchers are invited to our Statistical Consulting Services
We recommend others to our list of Other Resources for Statistical Computing Help
These pages are Copyrighted (c) by UCLA Academic Technology Services