Background

Your work in the COVID testing lab has given you lots of experience collecting data from the testing facility. You recently updated your LinkedIn profile with some of this work. You also asked the Lab director if she would give you a recommendation on your profile. She obliged and wrote a generous message about your work to not just collect the data but to automatically ensure that it is valid.

The next day a recruiter sent you an update asking if you would be interested in a position helping an organization called CRISP develop its data acquisition and reporting tools. CRISP is a non-profit that works as a Health Information Exchange (HIE). A HIE aggregates health data for a region or community in order to make it available to public health officials. The State of Maryland has contracted with CRISP to collect and aggregate all the COVID test results and vaccination medical records for tracking.

After careful consideration you’ve decided to join CRISP. On your first day you are assigned to a team that is responsible for detecting anomalies in COVID related data that is being collected by CRISP from counties in Maryland. The Maryland Department of Health is using a product called ArcGIS Hub to make COVID related data that CRISP has collected available to the public in a dashboard. You can see the dashboard at https://coronavirus.maryland.gov/

Maryland’s COVID Dashboard

On December 11, 2020 the first vaccine was cleared for use in the United States by the Federal Drug Administration. [1] Your first assignment is to look at a county level vaccination dataset and analyze it to identify any problems.

Instructions

Download this CSV file for the The cumulative number of COVID-19 vaccinations within a single Maryland jurisdiction dataset that was originally published by the Maryland Department of Health at https://coronavirus.maryland.gov/datasets/md-covid19-totalvaccinationscountyfirstandseconddose but has since been removed.

When answering the following questions remember this weeks work using Python sets.

  1. What significant data quality problem do you see in this dataset? (2 points)

  2. What counties seem to be responsible for this problem? (2 points)

  3. What counties do not seem to be responsible for this problem? (2 points)

  4. In no less than 150 words please describe a scenario or scenarios where these problems could happen? Think about how data like this might get generated and speculate about what could go wrong (use your imagination). (3 points)

  5. Do you see any other problems with the dataset? (1 point)

1. Thomas, K., LaFraniere, S., Weiland, N., Goodnough, A and Haberman M. (December 11, 2020). F.D.A. Clears Pfizer Vaccine, and Millions of Doses Will Be Shipped Right Away. New York Times. https://www.nytimes.com/2020/12/11/health/pfizer-vaccine-authorized.html