Related Experiment Video
Updated: Oct 19, 2025
13:00
Measuring Attention and Visual Processing Speed by Model-based Analysis of Temporal-order Judgments
Published on: January 23, 2017
10.0K
Using a Response Time-Based Expected A Posteriori Estimator to Control for Differential Speededness in Computerized
1University of Illinois at Urbana-Champaign, USA.
Applied Psychological Measurement
|September 27, 2021
Summary
This study introduces a new method, the joint expected a posteriori estimator (J-EAP), to improve computerized adaptive tests (CATs) by using response times. The J-EAP enhances ability estimation and test efficiency compared to traditional methods.
Area of Science:
- Psychometrics
- Educational Measurement
- Cognitive Psychology
Background:
- Computerized adaptive tests (CATs) traditionally rely on item responses for ability estimation.
- Differential speededness, where examinees respond at different speeds, can bias ability estimates in CATs.
- Incorporating response times (RTs) offers a potential avenue to improve CAT accuracy and efficiency.
Purpose of the Study:
- To develop and evaluate an extended procedure for the joint estimation of ability and speed parameters in CATs.
- To introduce the joint expected a posteriori estimator (J-EAP) within van der Linden's hierarchical framework.
- To assess the performance of J-EAP against standard maximum likelihood estimators (MLE) in terms of accuracy and efficiency.
Main Methods:
- Developed the joint expected a posteriori estimator (J-EAP) for simultaneous estimation of ability and speed parameters.
- Utilized van der Linden's hierarchical framework for the theoretical underpinnings of the J-EAP.
- Employed simulation studies and a real data example to evaluate J-EAP's performance against MLE.
Main Results:
- The J-EAP significantly outperformed the standard maximum likelihood estimator (MLE) in correlation, root mean square error, and bias for both ability and speededness.
- Using J-EAP with the maximum information per time unit (MICT) item selection method further reduced average examinee test time and variability.
- Estimation efficiency was maintained while achieving gains in speed and reducing test time variability.
Conclusions:
- The J-EAP provides a more accurate and efficient method for ability estimation and item selection in CATs by incorporating response times.
- This approach effectively controls for differential speededness, leading to more reliable measurement.
- The findings suggest that integrating RTs into CATs using J-EAP is a promising advancement in educational and psychological measurement.
