Related Experiment Video
Updated: Jun 29, 2026
08:05
Measuring Statistical Learning Across Modalities and Domains in School-Aged Children Via an Online Platform and Neuroimaging Techniques
Published on: June 30, 2020
A Unified Framework for Jointly modelling Response Times and Item Position Effects in Computer-Based Learning
Silvia Bacci1, Rosa Fabbricatore2, Maria Iannario3
1Department of Statistics, Computer Science, Applications "G. Parenti", University of Florence, Firenze, Italy.
Multivariate Behavioral Research
|June 27, 2026
Summary
This study introduces a new model to track changes in individual test-taker skills and speed over time. It integrates response accuracy, time, and item position for better assessment in learning environments.
Area of Science:
- Psychometrics
- Educational Measurement
- Cognitive Science
Background:
- Traditional assessment models often fail to capture individual performance dynamics, treating speed and ability variations as mere residual variances.
- The impact of item position and the ordinal nature of response accuracy are frequently overlooked in current assessment frameworks.
- Existing methods do not adequately address the dynamic changes in individual performance throughout a testing session.
Purpose of the Study:
- To develop a comprehensive modelling framework that integrates item responses, response times, and item position.
- To better understand skill acquisition and latent speed changes in individuals during assessments.
- To provide diagnostic insights into individual differences and the influence of item positioning on test performance.
Main Methods:
- Utilizing the Bivariate Generalised Linear Item Response Theory (B-GLIRT) model to capture the dual impact of ability on accuracy and the interplay of ability and speed on response times.
- Extending the B-GLIRT model by incorporating random effects for item-specific and individual-specific variations.
- Investigating the effects of item positioning on overall test performance.
Main Results:
- The proposed B-GLIRT framework effectively models the dynamic relationship between ability, speed, and accuracy.
- The model demonstrates significant individual and item-specific variations influencing performance.
- Item position was found to have a measurable effect on response accuracy and time.
Conclusions:
- The integrated modelling framework offers a more nuanced understanding of test-taker performance than traditional methods.
- This approach enhances the assessment of skill acquisition and latent speed changes in computer-based learning.
- Findings have practical implications for improving the design and interpretation of educational assessments.
