A general nonparametric classification method for multiple strategies in cognitive diagnostic assessment

Area of Science:

• Educational Measurement
• Psychometrics
• Cognitive Science

Background:

• Cognitive diagnosis models (CDMs) are vital psychometric tools for evaluating student mastery of cognitive skills.
• Existing parametric multi-strategy CDMs require large sample sizes, limiting their practical use.
• The challenge lies in developing CDMs that can handle multiple solution strategies with limited data.

Purpose of the Study:

• To propose a general nonparametric multi-strategy classification method for cognitive diagnosis.
• To address the limitations of parametric CDMs in small sample settings.
• To enhance the accuracy and applicability of CDMs in educational assessments.

Main Methods:

• Developed a general nonparametric multi-strategy classification method for dichotomous response data.
• The method incorporates flexible strategy selection and condensation rule approaches.
• Validated through simulation studies comparing performance against parametric CDMs.

Main Results:

• The proposed nonparametric method demonstrated superior classification accuracy in small samples compared to parametric CDMs.
• The method effectively accommodates multiple strategies for solving assessment items.
• Real data analysis confirmed the practical applicability of the new approach.

Conclusions:

• The nonparametric multi-strategy classification method offers a viable alternative for cognitive diagnosis with limited data.
• This approach improves the reliability and practical utility of CDMs in educational settings.
• It provides a more accessible tool for identifying student strengths and weaknesses.