An item analysis of the Conditional Reasoning Test of Aggression

Area of Science:

• Psychological assessment
• Quantitative psychology

Background:

• The Conditional Reasoning Test of Aggression (CRT-A) is a tool for measuring aggression.
• Item response theory (IRT) provides a robust framework for evaluating psychometric properties of assessment instruments.

Purpose of the Study:

• To estimate item characteristics of the CRT-A using IRT.
• To determine the most appropriate IRT model for the CRT-A.
• To examine the dimensionality and suitability of the CRT-A for different populations.

Main Methods:

• Application of item response theory (IRT), including one-parameter (1PL), two-parameter (2PL), and three-parameter logistic (3PL) models.
• Analysis of a large sample (N=5,511) to assess latent aggression measurement.
• Exploratory analyses of polytomous IRT models and differential item functioning (DIF).

Main Results:

• The two-parameter logistic (2PL) model was identified as the most appropriate dichotomous IRT model for the CRT-A.
• CRT-A items function as intended, with aggressive response probability increasing with latent trait levels.
• The CRT-A is most effective for measuring individuals high in latent aggression.

Conclusions:

• The 2PL IRT model provides a strong psychometric foundation for the CRT-A.
• The CRT-A is a valid measure of latent aggression, particularly for high-aggression populations.
• Findings inform future research and appropriate application of the CRT-A.