Abstract:Educational testing intends to measure person's ability, and tests are designed for various objectives. While the ideal pure power tests and pure speed tests do not meet the practical needs in test administration, speeded tests become the primary option. The tradeoff between speed and accuracy is an important issue, especially in speeded tests, and being able to collect and analyze item response time, which is the amount of time an examinee devotes to a particular item until answering it correctly, provides an opportunity to investigate the relationship between response accuracy and response times. A few methods have been proposed to capture such relationship. This dissertation addresses the issue in the context of computer-based testing (CBT), where response times can be automatically recorded. An important component, time limits, in speeded tests is considered. The proposed mixture model incorporates both (potential) response accuracy and response times. In the spirit of speeded tests, it takes into account the possibility that some examinees are not potentially capable of solving a certain item, as in a pure power test, and decomposes response times into finite response times for those who are potentially capable and infinite response times for others. Time limits are represented in terms of censoring: some response times are censored because the problem solving procedures are terminated due to time limits. The proportion of examinees who are potentially capable of solving a particular item is modeled by a regular Item Response Theory (IRT) model. The parameter estimation method is first validated in extensive simulation studies, and the results of real data analysis indicate that the proposed mixture model outperforms the methods in which the impact of time limits is ignored.

Contributions to the statistical analysis of item response time in educational testing

A Mixture Cure-Rate Model for Responses and Response Times in Time-Limit Tests

Modeling Rapid Guessing Behaviors in Computer-Based Testlet Items

A Mixture Response Time Process Model for Aberrant Behaviors and Item Nonresponses

An Item Response Theory Model for Incorporating Response Times in Forced-Choice Measures

Modeling Not-Reached Items in Timed Tests: A Response Time Censoring Approach.

Semiparametric Factor Analysis for Item-Level Response Time Data

Differential and Functional Response Time Item Analysis: An Application to Understanding Paper versus Digital Reading Processes

A Mixture Fluency model using responses and response times with cognitive diagnosis model framework

Testing for Aberrant Behavior in Response Time Modeling

A Mixture Model for Random Responding Behavior in Forced-Choice Noncognitive Assessment: Implication and Application in Organizational Research

Applying Unidimensional and Multidimensional Item Response Theory Models in Testlet-Based Reading Assessment

Concurrent Use of Response Time and Response Accuracy for Detecting Examinees with Item Preknowledge

Test item response time and the response likelihood

Detecting Aberrant Behavior and Item Preknowledge: A Comparison of Mixture Modeling Method and Residual Method

Efficiency of computerized adaptive testing with a cognitively designed item bank

Using Item Response Theory as a Tool in Educational Measurement

Item Response Theory -- A Statistical Framework for Educational and Psychological Measurement

Item response theory in high-stakes pharmacy assessments

A Psychometric Perspective on the Associations between Response Accuracy and Response Time Residuals

Modeling Conditional Dependence of Response Accuracy and Response Time with the Diffusion Item Response Theory Model