During some of their most formative years, many children go to day care centers outside their homes. While there, they require a supportive, healthy environment that includes meaningful speech and conversation. This hinges on the soundscape of the child care center.
In his presentation at the 183rd Meeting of the Acoustical Society of America, Kenton Hummel of the University of Nebraska–Lincoln will describe how soundscape research in day cares can improve child and provider outcomes and experiences. The presentation, “Applying unsupervised machine learning clustering techniques to early childcare soundscapes,” will take place Dec. 8 at 11:25 a.m. Eastern U.S. in the Summit A room, as part of the meeting running Dec. 5-9 at the Grand Hyatt Nashville Hotel.
Few studies have rigorously examined the indoor sound quality of child care centers. The scarcity of research may deprive providers and engineers from providing the highest quality of care possible. This study aims to better understand the sound environment of child care centers to pave the way toward better child care.”
Kenton Hummel, University of Nebraska–Lincoln
The goal of the research is to understand the relationship between noise and people. High noise levels and long periods of loud fluctuating sound can negatively impact children and staff by increasing the effort it takes to communicate. In contrast, a low background noise level allows for meaningful speech, which is essential for language, brain, cognitive, and social/emotional development.
Hummel is a member of the UNL Soundscape Lab led by Erica Ryherd. Their team collaborated with experts in engineering, sensing, early child care, and health to monitor three day care centers for 48-hour periods. They also asked staff to evaluate the sound in their workplace. From there, they used machine learning to characterize the acoustic environment and determine what factors influence the child and provider experience.
“Recent work in offices, hospitals, and schools has utilized machine learning to understand their respective environments in a way that goes beyond typical acoustic analyses,” said Hummel. “This work utilizes similar machine learning techniques to build and expand on that work.”