Psychometric Developments Related to Tests and Selection
Abstract
Recent theoretical developments in the areas of latent variable modeling, random coefficient modeling, multilevel modeling, missing data theory, and empirical Bayes estimation provide interesting research opportunities for the psychometric modeling of the relationship between job performance and test performance. These developments suggest new approaches to the assessment of predictive validity of tests as well as to problems of selection and assignment. The purpose of the project is to use these theories as a basis for developing new psychometric methods that have the potential for better understanding the predictive validity of tests like the ASVAB and improving the selection and assignment of recruits for military jobs.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1993
- Accession Number
- ADA272971
Entities
People
- Bengt O. Muthen
Organizations
- University of California, Los Angeles