Voice Characteristics and Speech Recognition
Children rely on voice characteristics like pitch, vocal tract length, pitch contour, and overall quality to understand speech in noisy, multitalker settings. Our work shows that younger children often need multiple cues together (for example, pitch combined with vocal tract length) to separate talkers, while older children can rely on individual cues. We also find that pitch contour matters: exaggerated contours, common in child-directed speech, do not always help children, and in some cases less variable contours make speech easier to understand. Extended high-frequency cues are especially important for children’s recognition, which helps explain why face masks or other filters that reduce these frequencies make listening more difficult. Finally, differences in voice quality, such as dysphonia, reduce word recognition, sentence comprehension, and vowel intelligibility across development. Together, this line of work shows that children’s difficulties in noise reflect their developing ability to flexibly combine and use a wide range of voice cues. Our work in this area is ongoing to better understand how children use different voice cues to support speech recognition in noisy environments.
Talker Familiarity
Children spend much of their early years surrounded by familiar voices, and this experience can shape how they understand speech in noise. Our research shows that familiarity with a target voice improves children’s recognition, while familiarity with a competing voice does not reliably help them ignore it. For example, in one study children recognized sentences more easily when their mother’s voice was the target in multitalker settings, demonstrating the benefit of long-term familiarity. In another project, short-term familiarization—just five days of interactive gameplay with a single voice—led to an 11% improvement in recognizing words spoken by that voice in classroom noise. We are continuing to study talker familiarity both in classroom contexts and as a potential intervention approach to support children’s communication in noisy environments, with the goal of clarifying how familiarity shapes children’s listening across different settings.
Bilingualism, Listening Effort, and Accessibility
Children who grow up with more than one language face unique challenges and strengths when listening in noisy environments. Our research focuses on Spanish–English bilingual children, an understudied group in pediatric hearing science. Using one of the very few validated Spanish–English word recognition tests, we study how bilingual experience, proficiency, and cognitive factors shape speech recognition in noise. This test is especially important for ensuring that Spanish-speaking children with hearing loss are not overlooked in assessment and intervention planning.
To capture not only accuracy but also the effort behind performance, we combine behavioral measures like response times, self-report, and portable child-friendly EEG. By pairing bilingual assessment tools with neural and behavioral methods, this work addresses a critical accessibility gap and provides new insights into how bilingual children understand speech in complex environments.
Our lab also examines broader issues of accessibility in hearing healthcare, including how social, cultural, and systemic factors influence families’ pathways to services. Together, this work emphasizes that improving communication outcomes requires both an understanding of children’s listening abilities and attention to the accessibility of assessment and support for diverse populations.
Firefighter Hearing and Radio Communication
Firefighters routinely operate in some of the most acoustically challenging environments imaginable, contending with engine noise, power tools, alarms, and SCBA equipment while relying on two-way radios to communicate critical information. Our lab has been investigating how this occupational listening experience shapes auditory function and speech perception over the course of a firefighting career. Using a battery of measures including pure-tone audiometry and multiple speech-in-noise tasks, we examine how years of service, noise-induced hearing loss (NIHL), and potential ototoxic exposures affect firefighters’ ability to understand speech under realistic conditions. A central focus of this work is radio-mediated communication: fire radios are designed to suppress background noise, but this filtering also distorts the speech signal itself, compounding the listening challenge on scene. Importantly, despite these acoustic demands, firefighters demonstrate superior speech-in-noise performance compared to civilian listeners with normal hearing, suggesting that extensive listening experience in complex, high-stakes acoustic environments may function as a form of real-world perceptual training. This study, initially launched two years ago and now actively resuming data collection with an expanded protocol, is conducted in partnership with the Illinois Fire Service Institute (IFSI). Future directions include real-world noise dosimetry to better characterize the listening environments firefighters encounter during live training and incident response. Our goal is to build an evidence base that ultimately informs better communication gear, hearing conservation programs, and training practices for firefighters.