Adaptive comparison matrix: An efficient method for psychological scaling of large stimulus sets

Area of Science:

• Cognitive Psychology
• Psychophysics
• Human-Computer Interaction

Background:

• Quantifying subjective intensity across numerous stimuli with unknown perceptual order is crucial for vision and social science research.
• Existing methods can be time-consuming and require a large number of trials.
• A need exists for efficient protocols to measure subjective responses.

Purpose of the Study:

• To introduce an easy experimental protocol for ranking and scaling subjective stimulus intensity.
• To demonstrate the protocol's efficiency in using a small number of trials.
• To provide a method for characterizing human judgments, including those without physical correlates.

Main Methods:

• An iterative comparative judgment protocol where observers select the maximum response stimulus from a subset (M) of stimuli (N).
• The selected stimulus is removed, and the process repeats until all stimuli are ranked.
• Logistic regression is used to sort perceived stimulus order, and adaptive sampling selects informative stimulus pairs for judgment.

Main Results:

• Numerical simulations show the method can accurately estimate psychological scales with minimal responses.
• Psychophysical experiments successfully estimated contrast response functions for gratings.
• The protocol efficiently determined the perceived glossiness of naturalistic objects.

Conclusions:

• The proposed comparative judgment protocol offers an efficient and effective way to quantify subjective stimulus intensity.
• This method is valuable for research in vision, social sciences, and areas involving subjective attributes like emotions.
• The protocol's adaptability makes it suitable for a wide range of perceptual and subjective judgment tasks.