The coronavirus pandemic has resulted in a deluge of data. There are statistics everywhere you look, from the number of tests administered and cases confirmed, to the jobs lost, stimulus money spent, and Peloton bikes purchased. Understanding data feels more important than ever right now.
Since the pandemic started, Quartz has been spending an even greater share of our time confronting difficult data questions. Where do we find reliable data that show how Covid-19 is impacting the global economy? How do we know if a trend is caused by the novel coronavirus, or if it would have happened anyway? How do we appropriately describe the massive economic dip and recovery caused by the pandemic without misleading readers?
We wanted to share some of the answers we’ve conjured to these questions, which have helped us get a grip on the data deluge. We hope these answers can help you, too—not just to make sense of the pandemic, but to deal with the torrent of data we all face in our work and in our lives.
Where to find good coronavirus data
Before you can start analyzing what’s happening, the first step is finding the right data. Some of the data sources our reporters have turned to are classics, like the OECD’s database and the International Trade Centre’s dashboard. But we’ve also seen a surge in new data sources in response to Covid. Most notably, timely datasets using credit card payments, job postings, and cellphone mobility data have made it so the public doesn’t have to wait months for government data. By combining the new and the old, we put together a list of the very best coronavirus-related data on the Internet.
Read: The best coronavirus data resources on the Internet
Understanding percent change
In many countries across the world, Covid-19 caused a sudden decline in economic activity, followed by a rapid, or sluggish, recovery. As a result, staggering statistics became the norm. For instance, France’s industrial production fell by more than 20.6% in April from the previous month, but then rose by 19.9% in May. This yo-yo economy has made it more important than ever to understand how to calculate and interpret percent change. We made a comprehensive guide to understanding statistical ups and downs.
Read: The challenge of percent change
The power of indexing data
From the number of hours people are working in different countries to changing company stock prices, it’s often necessary to compare numbers on very different scales. The key to doing this well is indexing. When data points are indexed, they are all compared to their level at some particular point time. We wrote an explanation of why indexing is so essential and how you can do it yourself.
Read: How to compare data that change over time
Beware of seasonality
It hasn’t even been a year since the first case of Covid-19 was confirmed. For data analysis, this means it is essential we watch out for seasonal fluctuations. Did US computer sales shoot up in the second quarter of 2020 because of Covid-19? Or was it simply because computer sales always go up this time of year? Not accounting for seasonality is one of the easiest ways to find an erroneous trend. We explain how to avoid that pitfall.
Read: Covid-19 is a reminder to beware of seasonality in data
Create a pivot table
There is no simpler way for the average person to analyze a large dataset than the pivot table. Perhaps the most powerful tool in the spreadsheet wizard’s toolbox, the pivot table allows anyone to summarize thousands of rows of data in just a few clicks. Our history and explanation of the tool will help you understand why Steve Jobs thought pivot tables were the “coolest thing ever.”
Read: The history of the pivot table, the spreadsheet’s most powerful tool
More data lessons and sources
In addition to these recent stories, over the course of Quartz’s eight year history, we’ve written a lot of stories with data advice. If you are interested in learning more, check out our stories on whether to use Microsoft Excel or Google Sheets, when it’s time to learn a statistical programming language, why you should use scatter plots and tables instead of pie charts, and the very best free books to learn statistics.
