Temporal aspects of auditory intensity processing

点击翻译按钮获取中文摘要

The experiments of the first project phase confirmed the hypothesis that the failure of selective attention to the target tones is a useful framework for understanding the sometimes dramatic effects of non-simultaneous masking on intensity resolution. In our experiments, we observed astonishingly large inter-individual differences (up to 20 dB) in the intensity difference limens under temporal masking. Considering that we tested normally hearing young listeners with high cognitive abilities (university students) these results are even more surprising. Similar results were reported in previous studies on intensity discrimination under non-simultaneous masking, and in experiments on informational masking, where selective attention is also assumed to play an important role. The first aim of the experiments put forth in this proposal is to test the hypothesis that a general capability of controlling selective attention is a determinant of intensity resolution under non-simultaneous masking. To this end, we will measure the spatial and temporal control of selective attention in two classical tasks from cognitive psychology. In addition, we will measure the working memory span and several basic auditory abilities (e.g., sensitivity for the auditory temporal fine structure, gap detection). We expect the intensity resolution under forward masking to be correlated with performance in the tasks assessing the control of selective attention, but only weak correlations with basic auditory processing. The second aim of the proposed project is to test the hypothesis that intensity resolution under non-simultaneous masking is a predictor for speech recognition in complex communication settings. Recent studies reported large inter-individual differences in normally hearing listeners when a target speaker was presented together with competing speakers ("cocktail party situation"). We conjecture that in this situation central mechanisms like the control of selective attention again play an important role. Using correlation analyses, we will identify those performance measures (intensity resolution under temporal masking, basic auditory processing, control of selective attention, working memory span) that best predict speech recognition in a cocktail party setting. We will simulate several cocktail party environments that differ with respect to the presence of energetic and informational masking, the importance of the temporal fine structure, and the number of competing speakers.