Random Item MIRID Modeling and Its Application.

Area of Science:

• Cognitive Psychology
• Psychometrics
• Educational Measurement

Background:

• The Model With Internal Restrictions on Item Difficulty (MIRID) is established for cognitive behavior analysis.
• The original MIRID model has limitations, notably treating items as fixed, which restricts its application in contexts requiring random item considerations.

Purpose of the Study:

• To propose and evaluate random item approaches for the MIRID model.
• To enhance the MIRID model's flexibility and applicability in measurement contexts with random items.

Main Methods:

• Development of random item versions of the MIRID model.
• Conducting simulation studies to assess model performance under varying error conditions.
• Performing empirical studies to validate the proposed models with real-world data.

Main Results:

• The random item MIRID models demonstrated improved accuracy in parameter estimation.
• These enhanced models are particularly beneficial when substantial random errors are present in the data.
• Both simulation and empirical results support the utility of the random item approaches.

Conclusions:

• Random item approaches significantly improve the MIRID model's accuracy and utility.
• The enhanced MIRID models are recommended for cognitive behavior research involving random item sampling.
• This advancement broadens the application of MIRID in complex measurement scenarios.