Broad generalization of the ventriloquism aftereffect across sound frequencies

This dataset was collected to investigate the frequency generalization of the ventriloquism aftereffect (VAE)—a recalibration phenomenon where repeated exposure to spatially misaligned audiovisual stimuli shifts subsequent sound localization in the direction of the prior visual stimulus. The goal of the study was to determine whether this aftereffect is frequency-specific, cue-specific (linked to interaural time or level differences), or whether it generalizes across different frequency bands, thus reflecting adaptation at a higher-level spatial map. To test this, 12 human participants performed head-orienting sound localization tasks across three experimental sessions. Each session consisted of a pre-exposure phase (sound-only localization), an exposure phase (audiovisual localization with a consistent spatial discrepancy), and a post-exposure phase (sound-only localization to assess the VAE). Each session used a different exposure sound: a low-frequency narrowband noise, a high-frequency narrowband noise, or a broadband noise. Across all phases, participants localized up to eight sound stimuli—seven narrowband and one broadband—varying in center frequency (0.5–8 kHz), enabling analysis of frequency-dependent localization behavior and aftereffects. The collection includes the following: • Sound localization responses from all three phases. • Stimulus parameters, such as sound frequency, location, and exposure condition. • Anonymized participant id. The dataset allows for a detailed examination of how auditory spatial recalibration unfolds across frequencies and supports the hypothesis that the VAE arises within a modality-independent spatial representation. It can serve as a resource for researchers interested in multisensory integration, spatial hearing, perceptual learning, and Bayesian modelling of human behavior.