The CoronaNet Research Project is a joint initiative between the Hochschule für Politik at Technical University of Munich, New York University – Abu Dhabi and University of Southern California. We are running a large-scale data collection effort with more than 600 scholars worldwide to comprehensively understand government responses to the COVID-19 pandemic. We already have a large database of over 60,000 policies, and are looking to recruit volunteers/research assistants for the coming months to continue the work. 

The position is an unpaid, volunteer position, but working hours are flexible. It will primarily be of interest to anyone who enjoys policy research and has interest in international relations/comparative politics but also senior scholars are needed for evaluating and curating the data. Research assistants are able to work virtually from home and the project coordination takes place in Slack.

We have additional positions open for volunteers with additional skills in data visualisation and research with opportunities for publications open after you have joined the project.

Our initial Paper was published by Nature: Human Behaviour. More information can be found on our website at https://www.coronanet-project.org/.

If you have any questions please contact admin@coronanet-project.org.

We are looking forward to your application via https://www.coronanet-project.org/join.html !

OPENCOVID19 Grant Review 5

Problem and Background  

Governments everywhere have implemented an astonishing variety of policies in response to the COVID-19 pandemic. These measures range from travel restrictions to school closures. Moreover, the targets of these policies have not only been multifaceted (e.g. residents, travelers, countries), but the degree and timing with which such policies have been enforced has also been incredibly diverse. As the global community debates and compares different governmental approaches, comprehensive, fine-grained and validated data is crucial for not only understanding policies and their social, political, economic and health effects, but also for enabling timely and effective policy responses to the COVID-19 pandemic. 

Since late March 2020, the CoronaNet Research Project, a crowdsourced, open-science research endeavor with over 600 international volunteers, has been systematically documenting such policies. With more than 60,000 policies to date, it is the largest such dataset of its kind. However, in order to ensure that the data CoronaNet provides for research on the drivers and effects is as reliable and valid as possible, it requires additional funding support for cleaning and validating its data. 

Solution summary in simple terms  

In order to help forward the ability to analyze the drivers and effects of COVID-19 policies, the CoronaNet dataset is publicly available, updated on a daily basis and can be downloaded from our website. Aside from the data, we also provide users with a customizable, interactive overview of the data via the CoronaNet Research Project’s Public Data Dashboard, COVID-19 country policy reports and research notes by our volunteer experts as well as open science academic research from the principal investigators of the project (https://www.coronanet-project.org/publications.html). 

Any such analyses however, is only as good as the quality of the data it is based on. With this 2000 euro microgrant, we seek to further invest in the CoronaNet’s cleaning and validation teams by employing two paid research assistants for two months each; one to lead the cleaning effort and the other to lead the validation effort. 

Solution summary in technical terms 

As researchers learn more about the various health, economic, and social effects of the COVID-19 pandemic, it is crucial that to the greatest extent possible, they have access to data that is reliable, valid, and timely.

With funds made possible through a JOGL micro-grant, CoronaNet will be able to employ a paid research assistant to bolster its ongoing data cleaning efforts. This research assistant would be tasked with i) identifying and recruiting new data cleaners ii) helping to manage the existing data cleaning team and iii) writing new computer code to identify clear and systematic errors in the CoronaNet dataset and providing the identified errors as feedback to data collectors more generally. 

CoronaNet will also be able to employ a paid research assistant to bolster its ongoing data validation efforts. Under this scheme, 10% of the data is sampled to be triple coded by independent coders in order to assess the reliability of the CoronaNet data. This research assistant would be tasked with i) identifying and recruiting new data validators ii) helping to manage the existing data validation team.

State of advancement of the project  

To date, CoronaNet has documented more than 60,000 from over 195 countries, not only at the national level but also at the subnational level. Indeed, for 12 countries (Brazil, China, France, Germany, India, Italy, Japan, Nigeria, Russia, Spain, Switzerland, the United States), we are also systematically collecting this information for more than 400 ISO-2 level units. 

Until now, CoronaNet has cleaned about 4,400 of these entries and achieved an inter-coder reliability of 76.53 percent (for more details, see section 2.2). CoronaNet could greatly benefit from additional support in cleaning and validating the data. 

Project Timeline

• Week 1: Data Validation and Data Cleaning Overview. With the help of the required funds, we will be able to employ two paid research assistants for two months each who will lead and coordinate our ongoing data cleaning and data validation efforts. In the first week, they will focus on getting an overview of the current state of the cleaning and validation efforts (i.e. produce summary statistics, data gaps, etc.) and identify steps to move forward. The data cleaning RA will also create code which can identify systematic errors in the data during this week. 
• Week 2: Team Formation. After assessing the current situation, they will actively recruit and monitor additional research assistants to help out in cleaning and validating the CoronaNet data up until October 1st, 2020 capturing the first wave of the COVID-19 pandemic. They will lead the team by scheduling weekly meetings, setting up internal competitions for cleaning and validation and installing other motivational tools to boost the overall engagement.  
• Week 3-7: Step up Cleaning and Validation Efforts. In the following weeks, both teams will work on improving the overall quality of data entries in the CoronaNet dataset. On the one hand, the Data Cleaning team will utilize R to identify and communicate common smaller errors such as typographical mistakes as well as systematic errors found with the help of the code the paid RA will have created. On the other hand, the Data Validation team will check for the inter-coder reliability of the existing entries by recoding, with the help of two independent coders, a sample of 10% of the collected data. In coding one policy multiple times, the team’s efforts will be able to better assess the quality and reliability of the CoronaNet taxonomy and data. 
• Week 8: Re-Assessment of the CoronaNet Dataset. To finalize the grant period, the heads of the Data Validation and the Data Cleaning team will assess the progress made in improving the quality of the CoronaNet dataset. 

Slack Channel:  #proj-coronanet 

Website: https://www.coronanet-project.org/ 

LinkedIn: https://www.linkedin.com/company/coronanet-research-project/ 

Twitter: https://twitter.com/coronanet_org

Github: https://github.com/saudiwin/corona_tscs