Why do I get the message "weight variables not found?"
when I specify the weight() option?
| Title |
|
Specifying frequency weights in gllamm |
| Author |
Minjeong Jeon, University of California, Berkeley |
| Date |
July 2012 |
This problem has to do with specifying weight variables in the weight(stubname) option.
For example, weight(wt) specifies that variables wt1, wt2, etc. contain frequency weights.
wt is just the stub of the variable names. The variable names must begin with the same stub followed by a number.
The numbers in the variable names (wt1, wt2, etc.) determine to which level each weight applies.
For instance, wt1 is the frequency weight at level 1 and wt2 is the frequency weight at level 2, etc.
You do not need weight variables at all levels, but for the levels with no weight variables,
gllamm assumes the weights are equal to 1.
For example, suppose you have a three-level of data, but you have only myweight2 in the dataset.
Specify myweight in the weight() option in the gllamm command:
gllamm ... , weight(myweight)
It is very important to use myweight (not myweight2)
inside the parentheses for weight.
Since there are no myweight1 and myweight3 in the data set,
gllamm assumes that frequency weights at level 1 and 3
(myweight1 and myweight3) are equal to 1.
Examples and documentation
- Description of
weight() option on p.22-23,
and examples in Sections 3.2.2, 4.1.1-4.1.2, 8.3.1-8.3.2, 8.4, and 9.3
of Rabe-Hesketh, S., Skrondal, A. and Pickles, A. (2004).
GLLAMM Manual.
U.C. Berkeley Division of Biostatistics Working Paper Series. Working Paper 160.
- Skrondal, A. and Rabe-Hesketh, S. (2004).
Generalized Latent Variable Modeling: Multilevel, Longitudinal and Structural Equation Models. Chapman & Hall/CRC.
- Exercises 10.3, 10.7, 14.5, and 16.11 and p.929-930 in the book
Rabe-Hesketh, S. and Skrondal, A. (2012).
Multilevel and Longitudinal Modeling Using Stata (Third Edition).
Volume II: Categorical Responses, Counts, and Survival.
College Station, TX: Stata Press.
|