Activation of auditory efferent nerve fibers alters cochlear vibration patterns. Background noise in listening environments can excite the medial olivocochlear reflex (MOCR) pathway that reduces cochlear amplification. This reflex is thought to enhance our ability to hear speech in noise and may be associated with protection from hearing loss caused by acoustic trauma. The alterations of cochlear vibration patterns caused by the reflex can be measured non-invasively, though indirectly, using OAEs. A large body of research has examined correlations between the strength of MOCR-mediated changes in OAEs and speech in noise abilities. In practice, such correlational studies have shown variable results. My work has sought to understand and reduce this variability through the use of more sophisticated measurement paradigms and signal processing techniques. I have focused on applying statistical tests of repeated measures made within individual subjects. My results helped establish the necessity of combining multiple, repeated measurements from individual ears to establish reliable baselines. These baselines then provide a meaningful comparison from which to judge changes in an individual’s MOCR strength.

My work in this area has also focused on understanding the role of the middle-ear muscle reflex (MEMR) on MOCR measurements. General thinking is that the MOCR is evoked from sounds that are low in level, while the MEMR is evoked by sounds that are high in level. Our work has shown that the two reflexes are often excited simultaneously, clouding the interpretation of measurements if care is not taken. At present, the extent of MEMR influence on OAE-based assays of MOCR is unknown, and this is a potential limitation for clinical applications of these measurements. My current work has brought new calibration and wideband reflectance techniques to bear on this problem. This work showed that there are specific frequency ranges in individual ears at which middle ear reflexes cause minimal disturbance to the intended measurements.