Literature in Variance Component Estimations, Large-Scale Assessments, and Missing Data

Abstract

Despite the efforts made in measurement research to deal with the limitations encountered when analyzing unbalanced data via G theory, the research in this area suffers from major restrictions. In this chapter, I introduce the applications of G theory in norm-referenced and criterion—referenced testing. Next, I review the advantages and disadvantages of a variety of methods for handling large and unbalanced data sets. Last, I summarize studies that provide a foundation for the proposed subdividing method.