From theory to practice: A comprehensive toolkit for Q-matrix validation in cognitive diagnosis

Area of Science:

• Psychological assessment
• Educational measurement
• Cognitive psychology

Background:

• Cognitive diagnostic assessment (CDA) explores cognitive structures and psychological latent traits.
• The Q-matrix, essential for CDA, links items to attributes but faces accuracy and subjectivity challenges.
• Current Q-matrix validation methods are technically complex, limiting CDA's application in psychological research.

Purpose of the Study:

• To systematically review Q-matrix validation methods for saturated cognitive diagnosis models (CDMs).
• To introduce Qval, an R package designed to simplify and enhance Q-matrix validation procedures.
• To provide practical guidance for implementing Q-matrix validation in psychological research.

Main Methods:

• Systematic literature review of Q-matrix validation techniques within saturated CDMs.
• Development and implementation of the Qval R package, integrating parameter estimation, validation algorithms, and search strategies.
• Utilizing C++ code and parallel computing within Qval for improved computational efficiency.

Main Results:

• A comprehensive overview of existing Q-matrix validation methodologies is presented.
• The Qval R package offers a user-friendly, efficient, and comprehensive solution for Q-matrix validation.
• Qval facilitates the practical application of advanced validation procedures for researchers.

Conclusions:

• Effective Q-matrix validation is critical for the accurate application of CDA in psychology.
• The Qval package lowers the technical barrier, promoting wider adoption of CDA.
• This work empowers psychological researchers with accessible tools for robust cognitive diagnostic assessment.