视听类别学习中的诊断性信息整合：“视”频增益与“听”界壁垒
Asymmetry of diagnostic information integration in audiovisual category learning
林蕴旖 1黄丹阳 1黄晖 1唐雯勤 1刘志雅1
作者信息
摘要
本研究采用家族相似性类别学习范式，探究视听双通道诊断性信息在多维度类别学习中的整合机制及其不对称性。两个实验共招募153名大学生，结合计算模型分析学习后的类别表征。实验1考察听觉信息诊断性对视觉类别学习的影响，操纵听觉诊断性为高（0.9）、一致（0.75）和低（0.6）三种水平。结果发现，当视听诊断性一致时，学习者对双通道特征维度的加工最为充分，表现出跨通道整合的增益效应。实验2则考察视觉信息诊断性对听觉类别学习的影响，发现诊断性一致条件下的整合增益没有出现，视觉诊断性对听觉特征维度的学习没有显著影响，体现视听信息加工的相对独立性。综上，多维度类别学习中的跨通道信息整合存在不对称性：视觉任务可以整合听觉信息，而听觉任务对视觉信息的整合深度有限。本研究首次揭示了多维度听觉类别学习的学习效率与表征策略，并讨论了视听类别学习在表征机制上的差异。
Abstract
Category learning research has expanded from the visual domain to auditory, olfactory, and tactile modalities. Although studies have demonstrated that individuals can learn complex visual categories by acquiring multidimensional features, whether the auditory modality has similar capabilities remains underexplored. Feature diagnosticitythe extent to which a feature predicts category membershipplays a critical role in category representation as a core construct in category structure. The present study investigated how diagnostic information from one modality influences multidimensional category learning in another modality and explored the representation strategies in this process. We hypothesized that cross-modal audiovisual integration is asymmetric due to differences in processing mechanisms and cognitive load between modalities: visual tasks are more likely to exhibit cross-modal integration, whereas auditory tasks are less likely to do so.The present study employed a family-resemblance category learning paradigm to examine the interaction between auditory and visual information and their representation mechanisms in multidimensional category learning. In Experiment 1, participants completed a visual classification task with fixed visual feature diagnosticity (0.75), while auditory feature diagnosticity was manipulated across three levels (high: 0.9; congruent: 0.75; low: 0.6) to examine its influence on visual category learning. Experiment 2 reversed the task structure to examine the influence of visual feature diagnosticity on auditory category learning. A total of 87 and 66 participants were recruited for Experiments 1 and 2, respectively. Dependent measures included accuracy and reaction time during learning, accuracy and number of acquired dimensions in dimension tests, as well as accuracy and categorization strategies in the final test.In Experiment 1, diagnosticity conditions significantly influenced the depth of processing for cross-modal feature dimensions. When auditory and visual diagnosticity levels were congruent, participants exhibited more balanced processing of information from both modalities, as reflected in the highest accuracy rates for both visual and auditory feature dimensions. This "co-frequency enhancement" effect represents the gain of audiovisual integration under conditions of aligned diagnosticity. In stark contrast, Experiment 2 revealed that visual information had no significant impact on auditory category learning. The acquisition of auditory feature dimensions did not differ across diagnostic groups, whereas the accuracy of visual feature dimensions increased with higher diagnosticity. This "auditory boundary barrier" effect indicates the relative independence of audiovisual information processing in auditory tasks. Furthermore, although auditory category learning was inferior to visual category learning in overall performance, no significant differences emerged between the two modalities in attention span or learning outcomes. Both modalities exhibited similar representational patterns, with prototype-based strategies predominating.This study revealed an asymmetric mechanism of audiovisual interaction in category learning. Visual tasks readily integrate auditory information, whereas auditory tasks exhibit limited capacity for integrating visual information, underscoring fundamental differences in visual and auditory processing. This study provided the first evidence for the learning efficiency and representation strategies of multidimensional auditory category learning, and discussed the differences in representation mechanisms between visual and auditory category learning. Future research could employ larger sample sizes with optimized experimental designs to further verify the robustness of these effects.关键词
类别学习/家族相似性/特征诊断性/视听整合/类别表征Key words
category learning/family resemblance/feature diagnosticity/audiovisual integration/categorical representations引用本文复制引用
林蕴旖,黄丹阳,黄晖,唐雯勤,刘志雅.视听类别学习中的诊断性信息整合：“视”频增益与“听”界壁垒[EB/OL].(2026-06-07)[2026-06-10].https://chinaxiv.org/abs/202606.00064.学科分类
科学、科学研究
