Individual Differences in Pitch Perception

This thesis reports a number of experiments investigating individual differences in pitch perception. Experiment 1 identified otherwise normally hearing adult listeners who were relatively insensitive to the direction of small but detectable frequency changes between sequential pure tones. Following previous research, an important feature of the stimuli used in Experiment 1 was that the standard
frequency was varied randomly—or ‘roved’—over a wide range (400–2000.1 Hz) over trials. Subsequent psychoacoustical experiments (Experiments 2–6) revealed that the insensitivity to pitch-change direction observed in some individuals appeared to depend critically on the use of a relatively wide frequency-roving range, and the possible origins of this effect were investigated in detail. A compelling hypothesis that emerged was that roving introduced random, irrelevant frequency changes
to the stimulus ensemble that interfered with pitch-direction identification (but not pitch-change detection) in some listeners. The results of Experiments 1–6 were considered under the framework of signal detection theory, and were compared to the predictions of several phenomenological models of frequency discrimination. The pre-existing models from the literature were not able to account for many aspects of the present findings without modification. The listeners taking part in the psychoacoustical experiments completed questionnaires assessing their everyday hearing experiences and musical ability. Listeners’ thresholds were weakly or moderately correlated with some aspects of their self-reports, including sound segregation and musical experience/perception, and uncorrelated with other aspects, such as sound localisation and speech intelligibility. Experiment 7 used magnetoencephalography to test the hypothesis that insensitivity to pitch-change direction is associated with a greater-than-normal involvement of left-hemisphere cortical areas when listening to and identifying the direction of frequency changes. The results suggested that the extent to which the two hemispheres contributed to the recorded signals varied considerably across individuals, and contrary to the results of an earlier study, poor pitch-discrimination ability was associated with stronger right-hemisphere auditory-evoked fields.