This Dissertation explores the unique issues related to specifying and fitting a second-order growth model to longitudinal item response data. The model examined is a hybrid of the logistic IRT model for binary item responses and the latent growth curve model for repeated measures. Attention centered around parameterization, identification, estimation, and issues related to estimation of the model such as convergence, improper solutions, bias and root mean squared error (RMSE). Two variations on the proposed model: one with correlated errors (model 2) and one without (model 1). In each model two types of estimator were examined: full and limited information estimators. Two sample size conditions were examined, one with N = 750 observations, and another with N = 3000. In addition, two item parameter sets were examined, one having a wide range of difficulty and the other, narrow. Comparing analyses stratified across model, findings indicated greater rates of improper solutions and bias for model 2 versus model 1. Limited information estimators of model 1 performed worse than full information estimators, while the opposite was true for model 2. Bias and convergence issues were greatest when difficulty had a wide range. Lastly, sample size appeared to play a negligible role in bias and RMSE, though it did affect convergence issues and improper solutions. Based on empirical results presented in this simulation the proposed model appears to be a logical statistical framework for modeling longitudinal item responses.
|Advisor:||Curran, Patrick J.|
|Commitee:||Bauer, Daniel J., MacCallum, Robert C., Thissen, David, Youngstrom, Eric|
|School:||The University of North Carolina at Chapel Hill|
|School Location:||United States -- North Carolina|
|Source:||DAI-B 71/09, Dissertation Abstracts International|
|Subjects:||Statistics, Quantitative psychology|
|Keywords:||IRT, Item response, Latent variable, Longitudinal, Longitudinal IRT, Longitudinal item response, Repeated measures, Repeated measures IRT|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be