|
|
|
||||
|
|
|||||
Table 8.3 on page 148 using data file meta20.ws. First we have to create a variable cons of constant 1. The other variable that we have to create is the standard error term which we call stderr (square root of variable varofd).
We will set up it as a 2-level model with the outcome variable to be the lowest level and the study level to be the next level. Variable cons will be included as a fixed parameter and random at level 2. Variable stderr is entered as random at level 1 and we will impose constraint on it later.
Now let's constrain the lowest level variance to 1. One way to do it is through Model menu. Select Constrain Parameters. A window similar to the one below will appear and we need to select random. Then enter 1 corresponding to d: stderr/stderr and also enter 1 corresponding to to equal to set the variance to be 1. Now pick a unused column in the data set, here we picked c13 to store the constraint that we just created and click on attach random constraints. We are now ready to run the model now.
The model we have built is the intercept only model:
The result is:
Now let's include variable weeks in the model and this gives the second column of Table 8.3.
The result is:
Table 8.4 on page 151 based on the same model built previously.
Part 1: Intercept only model
The result is:
Part 2: Intercept + ntot
The result is:
Part 3: Intercept + rii
The result is:
Part 4: Intercept + weeks
The result is:
Part 5: Intercept + ntot + rii + weeks
The result is:
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
