Journal article
Crossvalidation of Latent Class Models of Early Substance Use Onset
Multivariate behavioral research, Vol.29(2), pp.165-183
01/01/1994
DOI: 10.1207/s15327906mbr2902_3
PMID: 26745026
Abstract
R. A. Cudeck & M. W. Browne (1983) were among the first to discuss the advantages of taking a cross-validation approach to testing of covariance structure models. Here, Cudeck's & Browne's work is expanded in two directions: (1) into testing of latent class models; & (2) into using cross-validation to examine differences among groups, where groups may be formed by gender, ethnicity, region, etc. As illustration, cross-validation is used to help select models of early substance use onset in a sample of 3,643 young adolescents. Survey results suggest that the nature of the substance use onset process & the rate of movement through the process are the same for males & females at grades 7 & 8. However, evidence was found for gender differences in substance use experience at the beginning of grade 7, with males somewhat more advanced in the onset process. It is also suggested that double cross-validation is greatly preferred over single cross-validation. Adapted from the source document.
Details
- Title: Subtitle
- Crossvalidation of Latent Class Models of Early Substance Use Onset
- Creators
- Linda M Collins - University of Southern CaliforniaJohn W Graham - University of Southern CaliforniaJeffrey D Long - University of Southern CaliforniaWilliam B Hansen - Wake Forest University
- Resource Type
- Journal article
- Publication Details
- Multivariate behavioral research, Vol.29(2), pp.165-183
- DOI
- 10.1207/s15327906mbr2902_3
- PMID
- 26745026
- ISSN
- 0027-3171
- eISSN
- 1532-7906
- Language
- English
- Date published
- 01/01/1994
- Academic Unit
- Psychiatry; Biostatistics
- Record Identifier
- 9984280840102771
Metrics
11 Record Views