Relationships between acoustics, thermal, indoor air quality, and lighting conditions on student achievement in K-12 classrooms Lily M. Wang 1 Durham School of Architectural Engineering and Construction University of Nebraska - Lincoln 1110 S. 67th St. Omaha, NE 68182-0816, USA

ABSTRACT Data on acoustic, thermal, indoor air quality, and lighting conditions have been collected from 220 classrooms in the midwestern United States. Gathered acoustic data include sound levels logged every 10 seconds and room impulse responses from which reverberation times were extrapolated. K-means clustering was used to group the logged sound data into times when speech was or was not occurring; then acoustic metrics were calculated from the clustered data. When comparing the measured acoustic conditions to ANSI S12.60, 91% of the classrooms did not meet the recommended maximum background noise level for unoccupied conditions, while 15% did not meet the recommended maximum reverberation time. The field measurements also revealed that only 20% of the classrooms met ASHRAE Standard 62.1 ventilation rate requirements, while all classrooms met recommended IES illuminance level for reading and writing. Multivariate linear regression analyses between the environmental conditions and student achievement data, while controlling for student demographics, have identified a number of significant relationships. This presentation summarizes key results, describes how acoustic conditions were correlated to building mechanical systems, and considers how indoor environmental quality may be optimized to benefit occupants in educational settings. [Work supported by the United States Environmental Protection Agency Grant Number R835633.]

1. INTRODUCTION

A research team at the University of Nebraska – Lincoln has completed a measurement campaign in 220 K-12 classrooms from across 39 schools in the midwest region of the United States to understand better how indoor environmental conditions in classrooms impact student achievement. The campaign involved measuring several aspects of the indoor environmental quality, including acoustics, lighting, thermal and indoor air quality. Data were logged in occupied conditions continuously over 36 hours during two consecutive school days; these sessions were repeated three times in each classroom throughout the year. Various metrics were calculated from the gathered data, and those were then correlated to classroom-aggregated student performance on standardized achievement tests, while controlling for classroom-aggregated student demographics. 2. MEASUREMENT METHODS

Detailed descriptions of the measurement methods are provided in references [1] and [2]. Herein, only a brief summary of the acoustic measurements and demographic data are provided.

1 lilywang@unl.edu

a Shea mar ce 21-24 AUGUST SCOTTISH BENT caso

2.1. Acoustics With regards to acoustics, each classroom was outfitted with two sound level meters, one placed in a wire frame box near the instructor’s station and one that was hung from the ceiling. These logged sound levels across octave bands and other equivalent sound levels (A,B,C,Z) every ten seconds with an integration period of ten seconds. K-means clustering was then applied to the acoustic measurements to place each data point into one of two clusters: one in which speech was occurring and the other when it was not. The clustered data have been analyzed to better understand speech and noise levels in occupied K-12 classrooms [3]. Reverberation times were also calculated for each classroom from gathered impulse responses.

2.2. Demographic Data Classrooms at grades 3, 5, 8 and 11 were included in the study. Students in the lower two grade levels typically learn in the same classroom throughout the day. For grades 8 and 11, targeted classrooms were those in which either English language arts or mathematics were taught. In total, there were 144 primary school classrooms, 32 of grade 8, and 44 of grade 11 in the sample. For each school, the research team received classroom-aggregated demographic information, such as the percentage of students who receive free or reduced-price lunch, the percentage of students who speak English as a second language, the percentage of learners designated as gifted, and the percentage of learners designated as requiring special education. Figure 1 shows a sample distribution of the percentage of free and reduced lunch recipients in classrooms, listed by school district. This variable is used to control for socio-economic status which has been shown to have significant correlation to student achievement outcomes.

Figure 1: The percentage of free and reduced-price lunch recipients by school district, with the number of schools measured in each district shown in parenthesis.

Percent r r T ; 1 10 —— + Maximum 4 Hl i of , i ; 1 + <— Outliers i 1 <— 75" Percentile sof H 4 oF —_ 4 ' <— Median cof T ' 4 \ sof 1 \ 4 1 ! f sof 4 + <— 25" Percentile sok + = 4 <— Minimum, + i 20F — 4 T — L H 10F 7 1 1 of — == | District A District B District District District E (80) (65) (45) (20) (20)

3. RESULTS

Figure 2 shows (a) measured 1-minute A-weighted equivalent sound levels and (b) mid-frequency reverberation times measured in the 220 classrooms. 91% of the classrooms do not meet the recommended maximum background noise level of 35 dBA for unoccupied conditions, while 15% do not meet the recommended maximum reverberation time of 0.6 sec [4].

Figure 2: The (a) one-minute background noise A-weighted equivalent sound levels (LAeq) and (b) mid-frequency averaged reverberation time (T20mid), measured in the 220 classrooms. The gathered measurements showed furthermore that as few as 20% of the classrooms in the study meet ASHRAE Std. 62.1 ventilation rate requirements [5], while all of them did meet recommended illuminance levels for reading and writing set by the Illuminating Engineering Society.

Detailed results are presented in references [3, 6-9].

3.1. Statistical Analyses Multivariate linear regression analyses between the environmental conditions and student achievement data, while controlling for student demographics, have identified a number of significant relationships. First, relationships from examining only one set of indoor environmental conditions (e.g., acoustics only) were studied; see references [6-9] for results of those. Subsequently a model with three acoustic metrics, five lighting metrics, and 23 thermal and indoor air quality metrics (including seasonal variations) was analyzed. Many of the resulting significant relationships from this combined model align with results from the individual condition models, although some metrics that were found to be significantly correlated in the individual condition models did lose significance in the combined model.

A correlation is found between the mechanical system type and resulting non-speech sound levels. In the sample, there were 41 classrooms exhibiting a centralized system with variable air volume or air handling units, 104 classrooms with heat pumps, and 59 with unit ventilator systems. Those with unit ventilators were statistically found to have higher non-speech levels than the other types of systems. 4. SUMMARY

Regression models have been run relating data on indoor environmental conditions gathered from 220 classrooms across four grade levels (3, 5, 8 and 11) with classroom-aggregated standardized achievement scores in math and reading, while controlling for classroom demographics including socioeconomic status. Results give insight into which environmental conditions correlate with student test scores. Recommendations for future work are to include other classroom demographics

(b) 839938 8 8 a 2 ‘suiooisseig jos0quinN, 10 Bs Rk Rk 2 es ‘suioossseIO Jo10qUINN,

that were not used in the current study, such as the percentage of students with hearing impairments or those who are learning in a non-native language. Additionally, research that can conclude with causational relationships is needed, rather than correlational ones shown in this exploratory study. Another idea is to study the data at a more granular level of individual student performance, rather than classroom aggregated.

5. ACKNOWLEDGEMENTS

This research was supported by the United States Environmental Protection Agency Grant Number R835633. The author would like to thank other former and current members on the project team ( https://engineering.unl.edu/healthy-schools/ ) for their work on this project. 6. REFERENCES

1. Kuhlenengel, M., Brill, L., Deng, S., Lester, H., Bovaird, J., Lau, J., L. Wang, L., & Waters, C. An investigation of school environmental effects on student achievement. 2017 AEI Conference ; Oklahoma City, OK; 12-14 April 2017. 2. Kabirikopaei, A., Kuhlenengel, M., Arthur, A., Bovaird, J., Lau, J., Waters, C. & Wang, L. M. The effects of indoor environmental factors on students' academic achievement. 2019 AEI Conferenc e, Tysons, VA; 3-5 April 2019. 3. Wang, L. M. & Brill, L. C. Speech and noise levels measured in occupied K–12 classrooms. The Journal of the Acoustical Society of America , 150(2) , 864–877 (2021). 4. ANSI. American National Standard Acoustical Performance Criteria, Design Requirements, and Guidelines for Schools, Part 1: Permanent Schools. S12.60 . American National Standards Institute, New York, 2010. 5. American National Standards Institute. Std. 62.1-2010 Ventilation for Acceptable Indoor Air Quality. ASHRAE (ANSI/ASHRAE), 2010. 6. Brill, L. C. & Wang, L. M. Higher sound levels in K-12 classrooms correlate to lower math achievement scores. Front. Built Environ . 7 , 688395 (2021). 7. Kabirikopaei, A., Lau, J., Nord, J. & Bovaird, J. Identifying the K-12 classrooms' indoor air quality factors that affect student academic performance. Sci. Tot. Envr . 786 , 147498 (2021). 8. Kuhlenengel, M., Konstantzos, I., & Waters, C. E. The Effects of the Visual Environment on K- 12 Student Achievement. Buildings 11(11) , 498 (2021). 9. Deng, S. & Lau, J. Seasonal Variations of Indoor Air Quality and Thermal Conditions and Their Correlations in 220 Classrooms in the Midwestern United States. Building and Environment, 157 , 79-88 (2019).