A note on improving EAP trait estimation in oblique factor-analytic and item response theory models

Identificador:  imarina:9286557

Handle: https://hdl.handle.net/20.500.11797/imarina9286557

Autores:  Ferrando, Pere J; Lorenzo-Seva, Urbano

Resumen:
This note is about Bayes EAP scoring in the class of oblique linear and non-linear item response models that can be parameterized as factor analytic models. For these models, we propose an improved implementation approach that (a) provides more detailed and informative output, and (b) uses more prior information from the calibration stage. Overall, we discuss the limitations of EAP estimation as it is implemented in most available programs, provide a technical account of the basic methodology and proposed improvements, implement the proposal in the freely available FACTOR program, and illustrate how it works with FACTOR-based output.