Estimating and Fitting the Non-continuous category scored Polytomous Items under the Weighted Score Logistic Model...

Area of Science:

• Psychometrics
• Educational Measurement
• Statistical Modeling

Background:

• Traditional logistic models struggle with non-continuous, polytomous item scores.
• Accurate analysis of complex scoring is crucial in educational and psychological assessments.

Purpose of the Study:

• To introduce and evaluate a novel extension of the logistic model, the weighted score logistic model (WSLM).
• To assess the performance of the WSLM in analyzing polytomous items with non-continuous scores.

Main Methods:

• Developed the weighted score logistic model (WSLM) incorporating a weighted score parameter.
• Employed marginal maximum likelihood estimation for item parameters (difficulty and discrimination).
• Conducted a Monte Carlo simulation study to evaluate WSLM performance and model-data fit.

Main Results:

• The WSLM demonstrated low bias and root mean square error (RMSE) for item parameter recovery.
• Fit statistics (Q1, Q4) generally remained below critical values, indicating acceptable model-data fit.
• The model proved robust across various simulation conditions.

Conclusions:

• The WSLM is a viable and robust statistical tool for analyzing polytomous items with complex scoring schemes.
• Findings support the WSLM's applicability in practical educational and psychological assessment settings.
• The model offers an advancement over traditional dichotomous logistic models for specific item types.