In the standard Speech Reception Threshold (SRT) test, speech perception is typically assessed using sentences masked by stationary noise. In this study, we evaluated the ability of participants to perceive speech in the presence of ambisonics recordings of either background music or foodcourt noise, to increase the ecological validity of the test. We aimed to determine whether hearing loss would influence both the SRT and listening effort as measured by pupil dilation and subjective ratings. In addition, we applied an auditory and visual version of a Paced Serial Addition test to assess if any effects of hearing loss would extend to a less language-dependent task.

Ambisonics recordings of the masker signals were presented over 8 loudspeakers arranged in the horizontal plane around the listener, while the target sentences were presented from the front. A total of 129 participants (mean age: 58 years, range 37-73 years) completed the SRT task in either the background music or the foodcourt noise condition, targeting 80% correct intelligibility. Hearing acuity ranged from normal hearing to severely impaired (mean pure tone average (0.5 – 4kHz) = 27 dB HL, range -5 to 94 dB HL), and 55 participants performed the test with their own hearing aid(s).

The results showed that increasing hearing loss was associated with poorer SRTs and SRTs were better in the foodcourt masker than in the music masker condition. These results and the pupil response, subjective, and serial addition data will be presented and discussed.