Effects of Masker Intelligibility and Talker Sex on Speech-in-Speech Recognition by Mandarin Speakers Across the Lifespan

Objectives: Speech perception develops during childhood, matures in early adulthood, and declines in old age. Everyday listening environments often contain competing sounds that may interfere with the perception of the signal of interest. With competing speech, listeners often experience informational masking, where the intelligibility and acoustic characteristics (e.g., talker sex differences) of the maskers interfere with understanding of target speech. Across the lifespan, utilization of segregation cues in competing speech is not well understood. Furthermore, there is a dearth of research regarding speech-in-speech recognition across the lifespan in speakers of tonal languages such as Mandarin Chinese. Design: Speech recognition thresholds (SRTs) were measured in listeners with age-adjusted normal hearing; the age range of participants was 5 to 74 years old. All participants were native speakers of Mandarin Chinese. SRTs were measured in the presence of two-talker Forward or Reverse speech maskers where the masker sex was the same as or different from the target. Results: In general, SRTs were highest (poorest) with the Forward same-sex maskers and lowest (best) with the Reverse different-sex maskers. SRT data were analyzed for 5 age groups: child (5 to 9 years), youth (10 to 17 years), adult (18 to 39 years), middle-aged (40 to 59 years), and elderly (60 to 74 years). Overall, SRTs were significantly higher for the child group than for the youth, adult, middle-aged, and elderly groups ( p < 0.05), and significantly higher for the elderly than for the adult group ( p < 0.05). There was a significant interaction among age group, speech direction, and talker sex cues, where SRTs were significantly higher for Forward than for Reverse speech, and significantly higher for same-sex than for different-sex maskers for all age groups ( p < 0.05), except for the child group. Conclusions: Consistent with previous studies with non-tonal language speakers, the present SRTs with tonal language speakers were best in the adult group and poorest in the child and elderly groups. The child and youth groups demonstrated greater masking release with Reverse speech than with different-sex maskers, while the elderly group exhibited greater release with the different-sex maskers than with Reverse speech. This pattern of results may reflect developmental effects on utilization of talker sex cues in children; in older adults, enhanced top-down processes may compensate for the age-related declines in processing of temporal envelope and temporal fine structure information.