School Reopening Analysis


listen on castbox.fmlisten on google podcastslisten on player.fmlisten on pocketcastslisten on podcast addictlisten on tuninlisten on Amazon Musiclisten on Stitcher

--:--
--:--

School Reopening Analysis

In today’s episode, we are joined by Carly Lupton-Smith, a doctoral student/researcher in Biostatistics at Johns Hopkins Bloomberg School of Public Health. Her research interest lies in causal inference with mental health applications. Carly joins us to discuss findings from her study on the consistency between household and county measures of schools reopening.

She began with an overview of the data used for her analysis. She talked about the availability of the data, the methodology for data collection and the challenges the data had. To assess the impact of virtual school, Carly was granted access to a survey done by Delphi Group in partnership with Facebook, called the COVID-19 Trends and Impact Survey. The researcher explained how the survey questions had evolved over time until they captured the effect of school experiences.

Carly then discussed an important observation from the survey results. There had been some discrepancies between the regional-level onsite learning policies and the reality in households. She explained some of the reasons for this and the factors affecting the discrepancy. Carly further discussed how policymakers can make informed decisions from the survey findings to open up onsite learning whilst reducing the risk of COVID-19 spread.

She also highlighted other useful datasets for the research. Data such as the Burbio dataset, the U.S. School Closure and Distance Database, and the MCH Strategic data. She discussed the unique aspects of each dataset and their use cases. 

Carly went on to discuss how despite the discrepancy in the county level survey and household-level data, none are intrinsically wrong. She gave her thoughts on the validity of both data and the best situations for one over the other. Going further, she outlined potentials of the data for further analysis.

Carly rounded off with her take on how to find a sweet spot between onsite learning and minimal exposure to COVID-19. She mentioned how a research paper published by Justin Lessler and her team, showed that taking mitigative measures has a strong effect on COVID-19 spread in onsite learning. You can follow Carly on LinkedIn or Twitter to stay updated with her work. Follow CMU Delphi on Twitter for further discussion around this research work.


Thanks to our sponsors for their support

ClearML is an open-source MLOps solution users love to customize, helping you easily Track, Orchestrate, and Automate ML workflows at scale.
https://clear.ml/
Astera Centerprise is a no-code data integration platform that allows users to build ETL/ELT pipelines for modern data warehousing and analytics.
https://www.astera.com/products/centerprise-data/