Quantifying accuracy and precision from continuous response data in studies of spatial perception and crossmodal...

Area of Science:

• Auditory perception
• Spatial cognition
• Psychophysics

Background:

• Quantifying sensory localization accuracy and precision is crucial in perception research.
• Existing methods include error-based and regression-based metrics, but their agreement is not fully understood.

Purpose of the Study:

• To compare error-based and regression-based metrics for estimating auditory localization accuracy and precision.
• To investigate the relationship between crossmodal spatial recalibration and baseline localization performance.
• To assess the impact of audiovisual spatial disparity on localization metrics.

Main Methods:

• 188 participants localized auditory stimuli in azimuth.
• A subsample underwent audiovisual stimuli with spatial disparity before re-testing sound localization.
• Analysis involved comparing error-based metrics (e.g., single-trial errors) and regression-based metrics (e.g., slope, bias).

Main Results:

• Accuracy and precision in sound localization are distinct but show moderate correlation.
• High correlation observed between variable error (error-based) and overestimation of eccentricity (regression-based).
• Audiovisual exposure induced a ventriloquism aftereffect (shift in bias) but did not alter precision; effect size partly due to test repetition.

Conclusions:

• Error-based and regression-based metrics offer complementary insights into spatial localization, but intercorrelations require careful consideration.
• Spatial recalibration primarily affects bias, not precision, and baseline individual differences are critical.
• Future spatial perception and learning studies must account for these metric interdependencies and individual variability.