Computerized Adaptive Testing for Ipsative Tests with Multidimensional Pairwise-Comparison Items: Algorithm...

Area of Science:

• Psychometrics
• Educational Measurement
• Psychological Assessment

Background:

• Computerized adaptive testing (CAT) is widely used for efficient assessment.
• Multidimensional pairwise-comparison (MPC) items offer rich data but pose unique psychometric challenges for CAT.
• Existing CAT solutions for MPC items are limited, especially for dichotomous and polytomous formats.

Purpose of the Study:

• To propose novel item selection and exposure control algorithms for CAT with dichotomous and polytomous MPC items.
• To evaluate the effectiveness of these new methods through simulation studies.
• To demonstrate the applicability of the proposed CAT procedures in practical assessment scenarios.

Main Methods:

• Development of new item selection strategies tailored for MPC items within a CAT framework.
• Implementation of within-person statement exposure control mechanisms to prevent item overlap.
• Simulation studies using both dichotomous and polytomous MPC item formats to assess procedural performance.

Main Results:

• The proposed item selection methods effectively identify suitable items for CAT.
• The exposure control procedures successfully limit within-person statement exposure.
• The implemented methods maintain assessment efficiency without compromising measurement accuracy.

Conclusions:

• The developed CAT procedures are effective for tests utilizing dichotomous and polytomous MPC items.
• These methods offer a viable solution for measuring complex constructs like career interest, value, and personality.
• The proposed approach has practical implications for work attitude and career interest assessments.