A Note on the D-Scoring Method Adapted for Polytomous Test Items.

Area of Science:

• Psychometrics
• Educational Measurement
• Psychological Assessment

Background:

• The D-scoring method was previously established for binary items as an analog to basic item response theory (IRT) models.
• Many assessments utilize polytomous items, necessitating an extension of D-scoring methodologies.
• Item response theory (IRT) provides a framework for understanding item and test characteristics.

Purpose of the Study:

• To develop and present an approach for D-scoring of polytomous items.
• To compare the results of the proposed D-scoring method with those obtained using the graded response model (GRM) in IRT.
• To establish a unified framework for the psychometric analysis of tests containing both binary and polytomous items.

Main Methods:

• The study proposes generating 'virtual' binary items from polytomous items to apply the D-scoring method.
• This approach is situated within the item response theory (IRT) framework.
• The proposed method is evaluated by comparing its outcomes with the graded response model (GRM).

Main Results:

• The D-scoring approach with virtual binary items produces ability scores consistent with the graded response model (GRM).
• Item category response functions derived from this method are analogous to those obtained under the GRM.
• The proposed method offers a unified framework for D-scoring and psychometric analysis.

Conclusions:

• The developed D-scoring method effectively handles polytomous items by converting them into virtual binary items.
• This approach provides psychometric properties (ability scores, item category response functions) that align with established IRT models like GRM.
• The unified framework enhances the efficiency and applicability of D-scoring across diverse educational and psychological assessment scenarios.