ESTIMASI PARAMETER POPULASI: TEORI RESPON BUTIR UNTUK DATA LONGITUDINAL; ITEM RESPONSE THEORY FOR LONGITUDINAL DATA: POPULATION PARAMETER ESTIMATION

Ahmad Anis Abdullah, Subanar

2014 | Disertasi | PROGRAM STUDI S2 MATEMATIKA

In this thesis discussed how to estimate population parameters of item response theory for longitudinal data. Logistic model used is the three-parameter logistic model (3- PL). In a longitudinal study known some form of covariance among which compound symmetry, autoregressive, exponential, and independence. While the estimation methods used in the IRT in this thesis is the maximum likelihood method. Furthermore, we will compare the AIC value (Akaike Information Criteria) of the existing covariance models. The best model is the model of correlation that has the smallest AIC value.

Kata Kunci : Longitudinal data; Population parameters; Logistic model; Covariance structures; Maximum Likelihood