TY - GEN
T1 - Acoustic conditions in occupied classrooms
AU - Wang, Lily M.
AU - Smith, Kieren H.
N1 - Funding Information:
Sound levels in 220 K-12 classrooms have been logged every 10 seconds during 36-hour periods spanning two occupied school days, at three different times throughout an academic year. The data have been parsed using k-means clustering into times when speech was occurring and times when speech was not occurring. Assorted analyses of these two data clusters are presented to understand the variation of sound levels throughout an occupied school day. Among the metrics that have been studied are equivalent levels, standard deviations, statistical levels (including L10-L90), the percent of time that the speech levels in the rooms exceeded 65 dBA, and the percent of time that the nonspeech levels in the rooms exceeded 50 dBA. Some of these metrics are found to be statistically different across the four grade levels for which classroom data were acquired (3rd, 5th, 8th and 11th grades). The sound level metrics have been correlated with classroom-aggregated student achievement data in reading and math subject areas, and results indicate that higher speech levels in rooms correlate with lower achievement scores in math. [Work supported by the United States Environmental Protection Agency Grant Number R835633.]
Publisher Copyright:
© Proceedings of the 26th International Congress on Sound and Vibration, ICSV 2019. All rights reserved.
PY - 2019
Y1 - 2019
N2 - Sound levels in 220 K-12 classrooms have been logged every 10 seconds during 36-hour periods spanning two occupied school days, at three different times throughout an academic year. The data have been parsed using k-means clustering into times when speech was occurring and times when speech was not occurring. Assorted analyses of these two data clusters are presented to understand the variation of sound levels throughout an occupied school day. Among the metrics that have been studied are equivalent levels, standard deviations, statistical levels (including L10-L90), the percent of time that the speech levels in the rooms exceeded 65 dBA, and the percent of time that the non-speech levels in the rooms exceeded 50 dBA. Some of these metrics are found to be statistically different across the four grade levels for which classroom data were acquired (3rd, 5th, 8th and 11th grades).
AB - Sound levels in 220 K-12 classrooms have been logged every 10 seconds during 36-hour periods spanning two occupied school days, at three different times throughout an academic year. The data have been parsed using k-means clustering into times when speech was occurring and times when speech was not occurring. Assorted analyses of these two data clusters are presented to understand the variation of sound levels throughout an occupied school day. Among the metrics that have been studied are equivalent levels, standard deviations, statistical levels (including L10-L90), the percent of time that the speech levels in the rooms exceeded 65 dBA, and the percent of time that the non-speech levels in the rooms exceeded 50 dBA. Some of these metrics are found to be statistically different across the four grade levels for which classroom data were acquired (3rd, 5th, 8th and 11th grades).
KW - Classroom acoustics
UR - http://www.scopus.com/inward/record.url?scp=85084012501&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85084012501&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85084012501
T3 - Proceedings of the 26th International Congress on Sound and Vibration, ICSV 2019
BT - Proceedings of the 26th International Congress on Sound and Vibration, ICSV 2019
PB - Canadian Acoustical Association
T2 - 26th International Congress on Sound and Vibration, ICSV 2019
Y2 - 7 July 2019 through 11 July 2019
ER -