15th Annual Conference of the International Speech Communication Association

September 14-18, 2014

Revisiting the Right-Ear Advantage for Speech: Implications for Speech Displays

Nandini Iyer (1), Eric Thompson (2), Brian Simpson (1), Griffin Romigh (1)

(2) Ball Aerospace, USA

Cherry (1953) reported that when listeners were presented with dichotic signal over headphones, they could reliably report words presented to the attended ear, while only being aware of the gross properties of the talker in the unattended ear. More recently, Gallun et al. (2007) showed that there were large differences in performance on dichotic tasks depending on ear of presentation, with significantly larger errors occurring when the target was presented to the left, rather than right ear (i.e., a right-ear advantage). In the current experiment, we explored two factors, type of signal in the non-target ear and uncertainty about the target ear, and their effects on right-ear advantage. The results indicated that the right-ear advantage was modulated by two factors: 1) nature of the speech stimuli presented in the unattended ear, and 2) target ear uncertainty. Substantial differences were observed between listeners in the tasks, leading to varying amounts of right-ear advantage across listeners for the listening conditions tested. These results and their implications for the design of multichannel speech communication displays are discussed, and the use of these methods is recommended as a useful screening tool for selection of personnel who listen to and use multichannel speech displays.

Full Paper

Bibliographic reference.  Iyer, Nandini / Thompson, Eric / Simpson, Brian / Romigh, Griffin (2014): "Revisiting the right-ear advantage for speech: implications for speech displays", In INTERSPEECH-2014, 457-461.