Inferring school district learning modalities during the COVID-19 pandemic with a hidden Markov model

• Published in: PloS one Vol. 18; no. 10; p. e0292354
• Main Authors: Panaggio, Mark J, Fang, Mike, Bang, Hyunseung, Armstrong, Paige A, Binder, Alison M, Grass, Julian E, Magid, Jake, Papazian, Marc, Shapiro-Mendoza, Carrie K, Parks, Sharyn E
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 04-10-2023; Public Library of Science (PLoS)
• Subjects: Biology and Life Sciences; Computer and Information Sciences; COVID-19; COVID-19 - epidemiology; Data sources; Disease transmission; Distance education; Earth Sciences; Education; Evaluation; Humans; Learning; Management; Markov chains; Markov processes; Medicine and Health Sciences; Modelling; Pandemics; Physical sciences; Probabilistic models; Probability; Public health; Public Health Surveillance; Public schools; School closures; School districts; Schools; Social Sciences; Students; United States; United States; United States > US
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0292354

During the COVID-19 pandemic, many public schools across the United States shifted from fully in-person learning to alternative learning modalities such as hybrid and fully remote learning. In this study, data from 14,688 unique school districts from August 2020 to June 2021 were collected to track changes in the proportion of schools offering fully in-person, hybrid and fully remote learning over time. These data were provided by Burbio, MCH Strategic Data, the American Enterprise Institute's Return to Learn Tracker and individual state dashboards. Because the modalities reported by these sources were incomplete and occasionally misaligned, a model was needed to combine and deconflict these data to provide a more comprehensive description of modalities nationwide. A hidden Markov model (HMM) was used to infer the most likely learning modality for each district on a weekly basis. This method yielded higher spatiotemporal coverage than any individual data source and higher agreement with three of the four data sources than any other single source. The model output revealed that the percentage of districts offering fully in-person learning rose from 40.3% in September 2020 to 54.7% in June of 2021 with increases across 45 states and in both urban and rural districts. This type of probabilistic model can serve as a tool for fusion of incomplete and contradictory data sources in order to obtain more reliable data in support of public health surveillance and research efforts.