Accounting for Persistence in Tests with Linear Ballistic Accumulator Models.

Area of Science:

• Psychometrics
• Educational Measurement
• Cognitive Psychology

Background:

• Low-stakes tests often show varying effort levels among test-takers.
• Understanding individual differences in capability and persistence is crucial for accurate assessment.
• Existing models may not fully capture the dynamics of disengagement during testing.

Purpose of the Study:

• To propose novel latent trait models for analyzing responses and response times in low-stakes testing.
• To incorporate individual differences in capability and persistence into measurement models.
• To model the interplay between solution processes and disengagement using the linear ballistic accumulator model.

Main Methods:

• Developed latent trait models integrating response and response time data.
• Modeled solution and disengagement processes using the linear ballistic accumulator model.
• Introduced model variants accounting for no guessing, random guessing, and informed guessing.

Main Results:

• Simulation studies indicated good recovery of parameter values and latent traits under specific conditions.
• Model variants demonstrated flexibility in handling different guessing strategies.
• Empirical data analysis validated the application of the proposed models.

Conclusions:

• The proposed latent trait models offer a robust framework for analyzing behavior on low-stakes tests.
• These models provide insights into the complex interplay of capability, persistence, and response strategies.
• The findings have implications for improving the accuracy and interpretation of educational assessments.