[Analyzing paired-comparison data in situations where judgment is affected by multiple factors]

Area of Science:

• Statistics
• Psychometrics
• Sensory Science

Context:

• Paired-comparison is a standard method for stimulus scaling.
• Existing methods struggle with binary data influenced by multiple factors.
• There's a need for robust statistical models in complex judgment scenarios.

Purpose:

• To propose an extension of the Bradley-Terry model for handling multi-factor binary paired-comparison data.
• To develop a method for estimating both stimulus scale values and the impact of influencing factors.
• To validate the proposed model using real-world sensory data.

Summary:

• The study presents a novel statistical method extending the Bradley-Terry model.
• This approach accommodates binary paired-comparison data influenced by multiple covariates.
• The model successfully estimates stimuli scaling and quantifies the effects of various factors.

Impact:

• Provides a more effective tool for analyzing complex paired-comparison data, especially in sensory evaluation.
• Enhances the ability to understand and quantify the influence of multiple factors on judgments.
• Offers a foundation for more generalized applications of the Bradley-Terry model in statistical analysis.