Is the flu really worse this year? Comparing the (ongoing) 2016-17 and 2015-16 flu seasons

Written by: Josh Rosenberg

Primary Source:  Joshua M. Rosenberg – February 28, 2017


I was sick last week, and I think I might have had a mild case of the flu. Since it seems like a lot of people have been sick, I was curious whether the flu was really worse this year than last… and since the CDC makes the data available for each year, I put the data together and created a GIF. This project is just a first attempt but the code (it’s in R) is available in a GitHub repository. The measure of flu activity is from minimal (1-3) to high (8-10) and the scale in the chart is the difference from the 16-17 and 15-16 flu seasons (which range from November – March; it ends in week 7 because that’s the most recent data for this year.



Ways to Improve

This can be improved in a number of ways (if you think of others, please feel free to adapt this and create your own, and / or please let me know):

  • Inclusion of multiple years of data: Why compare only to the past year, rather than the mean of the past 10 years or so? This data is available from the CDC.
  • Adding some kind of test for whether activity is within the normal range or is really worse (or better) than usual. Maybe just a t-test for each state (with the p-value corrected?)?
  • Scale the Activity Level differently: The 1-10 scale is intuitive enough, but when focusing on the difference from one year to the last (or to the mean), the numbers are less meaningful. Maybe a standard deviation works.
  • Improve the appearance. I was still learning the gganimate package, and so some parts of the plot aren’t how I’d like them (i.e., the complete absence of a title, having some difficulty adding a caption() layer, the frame rate being too quick.)
  • Downloading the data from the CDC, rather than using the excellent cdcfluview package. I tried this and was able to view National data, but had trouble with state data. I later found out I just added the year wrong (for the 2016-17 year, the most recent year should be 2016 not 2017).

More Information

The data is from this CDC page.

I used the new (and in-development) gganimate package to create the GIF. The basic idea is to create a ggplot, and then add an aesthetic for the Frame, which represents the change between frames in the plot, which in this case is the week.


It looks like the level of flu activity was about equal during the 2016-17 and 2015-16 seasons until New Years Eve; in most of the country, flu activity was much higher during the first seven weeks of 2016-17 season compared to the 2015-16 season. So, it does look like the flu really is worse this year, although we’ll have to see how the rest of the season pans out. I’ll try to update the post then.

Of course, I think this visualization has some implications for educational research (and research in other areas). Many times (in education) we examine change, but are challenged by how to visualize change over time. The frame aesthetic can also be things other than time, like different groups to compare. I’d like to think about how a GIF could be part of a presentation.

The following two tabs change content below.
Joshua M. Rosenberg is a Ph.D. student in the Educational Psychology and Educational Technology program at Michigan State University. In his research, Joshua focuses on how social and cultural factors affect teaching and learning with technologies, in order to better understand and design learning environments that support learning for all students. Joshua currently serves as the associate chair for the Technological Pedagogical Content Knowledge (TPACK) Special Interest Group in the Society for Information Technology and Teacher Education. Joshua was previously a high school science teacher, and holds degrees in education (M.A.) and biology (B.S.).