The optimal experimental design for multiple alternatives perceptual search

Area of Science:

• Cognitive Psychology
• Neuroscience
• Psychophysics

Background:

• Perceptual biases are fundamental to sensory processing, manifesting as illusions and aftereffects.
• Measuring perceptual biases is challenged by confounding nonperceptual factors like response bias and decision criteria.
• Differentiating true perceptual biases from response strategies is crucial for accurate experimental interpretation.

Purpose of the Study:

• To evaluate the robustness of Multiple Alternative Perceptual Search (MAPS) in isolating perceptual biases from confounding factors.
• To investigate the influence of response modality and feedback on MAPS estimates.
• To determine if MAPS is susceptible to decisional biases.

Main Methods:

• Utilized a four-alternative forced-choice design to measure perceptual biases.
• Assessed response biases across different response modalities (keyboard vs. mouse).
• Manipulated trial-wise feedback to observe its effect on bias estimation.

Main Results:

• Response biases were present but independent of perceptual bias estimates and response modality.
• Perceptual bias estimates decreased with trial-wise feedback, suggesting active observer correction.
• MAPS reliably detected perceptual biases and was not influenced by a middle-stimulus decision bias.

Conclusions:

• MAPS is a robust method for disentangling perceptual biases from response biases.
• Trial-wise feedback can reduce the measured magnitude of perceptual biases, impacting studies focused on bias size.
• Recommendations for MAPS experimental design include using a constant reference stimulus and considering candidate distribution mean variation.