Whole-child development losses and racial inequalities during the pandemic: Fallouts of school closure with remote learning and unprotective community
Data files
Oct 07, 2024 version files 51.60 KB
-
HELP_rev.csv
48.32 KB
-
README.md
3.28 KB
Abstract
Grounded in a strength-based (asset) model, this study explores the racial disparities in students’ learning and well-being during the pandemic. Linking the U.S. national/state databases of education and health, it examines whole-child outcomes and related factors—remote learning and protective community. It reveals race/ethnicity-stratified, state-level variations of learning and well-being losses in the midst of school accountability turnover. This data file includes aggregate state-level data derived from the NAEP and NSCH datasets, including all 50 U.S. states' pre-pandemic and post-pandemic measures of whole-child development outcomes (academic proficiency, socioemotional wellness, and physical health) as well as environmental conditions (remote learning and protective community) among school-age children.
[https://doi.org/10.5061/dryad.66t1g1k8f]
The HELP (Health-Education-Life Protection) dataset is composed of measures for state-level school-aged children’s academic proficiency, socio-emotional wellness, physical health (the whole child), and protective community.
It draws upon nationally/statewide representative data sets - the National Assessment of Educational Progress (NAEP) and the National Study of Child Health (NSCH) to examine the relationships among those measures prior to and after the COVID-19 pandemic as a whole and across varied racial/ethnic groups.
Description of the data and file structure
The dataset includes U.S. state-level 4th and 8th-grade students’ academic proficiency scores (reading and math) from 2019 and 2022 merged with state-level socioemotional wellness, physical health, and protective community derived from the National Survey of Children’s Health data (2018/19 and 2020/21) and state-level remote learning instruction/enrollment derived from the Monthly School Survey Dashboard.
The following naming conventions were used for the variables included in the dataset:
- g4m (Grade 4 math), g4r (Grade 4 reading), g8m (Grade 8 math), g8r (Grade 8 reading)
- _bas (performing at or above the NAEP Basic level)
- remotelearnpart (State-level student-reported remote learning enrollment)
- remoteinstruction (State-level School-reported remote learning instruction)
- remotelearndiff (State-level remote learning difficulties)
- gmean_ (geometric means for the constructs used)
- _wcm (Protective community measure)
- sewt (a composite measure of socioemotional wellness measure), ap (a composite measure of academic proficiency measure)
- ph (a composite measure of physical health)
- gainbas (State-level gain scores [2020/2021 - 2018/19 for socioemotional wellness and physical health measures; 2022 - 2019 for academic proficiency measures])
- wch or wcht (Whole-child measures)
- missing (blank)
Sharing/Access information
Data was derived from the following sources:
- National Assessment of Educational Progress (NAEP)
- Monthly School Survey Dashboard during the COVID-19 pandemic
- National Survey of Children’s Health data
Code/Software
- R was used for the regression analyses presented in the tables. Part of the example codes is as below:
HELP_rev <- read.csv(‘Your PATH\\HELP_rev.csv’, sep=”,”, header=T)
model_01 <- lm(ap1922gainbas ~ gmean_remote_all + gmean_wcm2021all + pctminor + pctpoor + gmean_ap19all, data= HELP_rev)
model_02 <- lm(g4r1922gainbas ~ remotelearningg4read + gmean_wcm2021all + pctminor + pctpoor + ap_all2, data= HELP_rev)
model_03 <- lm(g4m1922gainbas ~ remotelearningg4math + gmean_wcm2021all + pctminor + pctpoor + ap_all1, data= HELP_rev)
model_04 <- lm(g8r1922gainbas ~ remotelearningg8read + gmean_wcm2021all + pctminor + pctpoor + ap_all4, data= HELP_rev)
model_05 <- lm(g8m1922gainbas ~ remotelearningg8math + gmean_wcm2021all + pctminor + pctpoor + ap_all3, data= HELP_rev)
summary(model_01)
summary(model_02)
summary(model_03)
summary(model_04)
summary(model_05)
To address the research questions, this study examines repeated cross-sectional datasets with nation/state-representative samples of school-age children. For academic achievement measures, the National Assessment of Educational Progress (NAEP) 2019 and 2022 datasets are used to assess nationally representative samples of 4th-grade and 8th-grade students’ achievement in reading and math (http://www.nces.ed.gov/nationsreportcard). In 2019, the NAEP samples included: 150,600 fourth graders from 8,300 schools and 143,100 eighth graders from 6,950 schools. In 2022, the NAEP samples included: (1) for reading, 108,200 fourth graders from 5,780 schools and 111,300 eighth graders from 5,190 schools; (2) for math, 116,200 fourth graders from 5,780 schools and 111,000 eighth graders from 5,190 schools. Data are weighted to be representative of the US population of students in grades 4 and 8, each for the entire nation and every state. Results are reported as average scores on a 0 to 500 scale and as percentages of students performing at or above the NAEP achievement levels: NAEP Basic, NAEP Proficient, and NAEP Advanced. In this study, we focus on changes in the percentages of students at or above the NAEP Basic level, which is the minimum competency level expected for all students across the nation.
As a supplement to the NAEP assessment data, this study uses the NAEP School Dashboard (see https://ies.ed.gov/schoolsurvey/mss-dashboard/), which surveyed approximately 3,500 schools each month at grades 4 and 8 each during the pandemic period of January through May 2021: 46 states/jurisdictions participated, and 4,100 of 6,100 sampled schools responded. This study uses state-level information on the percentages of students who received in-person vs. remote/hybrid instructional modes. The school-reported remote learning enrollment rate is highly correlated with the NAEP survey student-reported remote learning experience (during 2021) across grades and subjects (r = .82 for grade 4 reading, r = .81 for grade 4 math, r = .79 for grade 8 reading, r = .83 for grade 8 math). These strong positive correlations provide supporting evidence for the cross-validation of remote learning measures at the state level.
For socioemotional wellness and physical health measures, the National Survey of Children’s Health (NSCH) data are used. The 2018/19 surveys involved about 356,052 households screened for age-eligible children, and 59,963 child-level questionnaires were completed. The 2020/21 surveys involved about 199,840 households screened for age-eligible children, and 93,669 child-level questionnaires were completed. Our analysis focuses on school-age children (ages 6-17) in the data. In addition, the NSCH data are also used to assess the quality of protective and nurturing environment for child development across family, school, and neighborhood settings (see Appendix).