Weighting for unequal selection probabilities in multilevel models.

Authors
Pfeffermann, D., Skinner, C.J., Holmes, D.J., Goldstein, H. and Rasbash, J..
Year
1998
Journal
Journal of the Royal Statistical Society: Series B, 8:2, 219-230
DOI
10.1111/1467-9868.00106
Abstract

Multi-level models for estimating conditional and unconditional longitudinal growth norms are presented. The procedure involves transforming the original growth measurements to Normality and modelling these with a two-level random coefficient model. Growth norms for any desired time interval and function can be derived. Height and weight data are used for illustration.

Number of levels
2
Model data structure
Response types
Multivariate response model?
No
Longitudinal data?
Yes
Further model keywords
Substantive discipline
Impact

Sets out formal framework for handling weights in multilevel models especially for sample surveys.

Paper submitted by
Harvey Goldstein, Graduate School of Education, University of Bristol, h.goldstein@bristol.ac.uk
Edit this page