Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Guide users in understanding the graph #162

Open
aatishb opened this issue May 1, 2020 · 3 comments
Open

Guide users in understanding the graph #162

aatishb opened this issue May 1, 2020 · 3 comments
Milestone

Comments

@aatishb
Copy link
Owner

aatishb commented May 1, 2020

Starting a thread for some related ideas that have been sitting on my to-do list.

I think we should provide some help in interpreting the graph. Like a how to read this graph section. (Currently we point folks to the related Minute Physics video, but we can't assume folks have watched this.) This could be some kind of overlay that is shown the first time a user visits the page, which explains what the axes are, the fact that we are on a log scale by default, what the trend line means, and addresses some possible confusions.

Related: I like how these folks show 4 scenarios to guide readers into interpreting the graph.

I also like how Our World in Data adds a prominent disclaimer as a subtitle to all Covid graphs, and think we should do something similar.

Their disclaimer for deaths:

Limited testing and challenges in the attribution of the cause of death means that the number of confirmed deaths may not be an accurate count of the true number of deaths from COVID-19.

& their disclaimer for cases:

The number of confirmed COVID-19 cases is lower than the number of total cases. The main reason for this is limited testing.

We should perhaps also emphasize that number of confirmed cases is strongly dependent on the testing rate, which can differ vastly across countries, so we advise caution when making comparisons between countries, and emphasize that the data is an under-estimate. My concern is that many people are over-estimating how accurately the data reflects the ground reality, and so we need to emphasize some of the uncertainties.

Vox did a nice video explainer on understanding Covid charts. While that is more focused on a time series graph, I think it's a good example of the kind of guidance we should strive to provide through the website.

@aatishb aatishb added this to the Clarity milestone May 1, 2020
@rpkoller
Copy link
Contributor

rpkoller commented May 2, 2020

I'll add a few loose thoughts into the discussion:

  • I would opt against some sort of overlay for the explanation. I normally dislike to pack everything on a single page. make one page for the graph make one page for the explanation. one job/goal per page. is more determined and frees from further layout, usability, accessibility and alike issues. Why not just change the header. Move the Covidtrends logo from the left to the right, remove the "primer explanation text" and the youtube video and place a menu point with something like "how to read the graph" like you suggested

  • use plain language for readability purpose and the sake of easy understanding for everyone.

  • Explain in a brief primer the basic idea of the graph. Then explain the basic elements. the x axis, the y axis, how each is calculated and what they stand for briefly. then explain how to read the curve. I LOVE the 4 scenario example! you could take that to explain the different states of developmental stage using the example of particular countries. but i would even extend it to a 5th scenario like singapur who had a lockdown, then the exponential development stopped but then they reopened too fast and the exponential growth was back. with those example curves it is also a good way to explain the sense of the "days doubling time" line (especially what the current doubling time for "decreasing growth but not stabilized" is and how to find that out with the tool for the particular country). and in the end lay out the caveats of that graph. a few could be taken from the VOX video but you could also add a few from the minute physics video. additionally you should mind and mention the different response times countries have to count new cases as well as deaths (in germany it takes for example up to 60 days that 98% of total deaths found their way into the statistics - and new infections aren't usually reported on a daily basis. they arent even reported digitally but via fax most of the time ;) )

  • maybe work on the text in a google doc collaboratively?

  • and create some sort of user stories upfront to check the text against later on ( a concept from content design). and maybe also test the final text with people how easily it is comprehensible for them?

@UriGrod
Copy link

UriGrod commented May 3, 2020

First: great website. I regard this chart as the best way to view and I look mainly at this page to follow the trends.

I had an idea re: explanations, which is relevant only if it's something you'd rather save time on. Maybe one of the small companies such as HelpHero or UserPilot would like to volunteer to do this professionally for free. Alternatively, one of the many companies doing "video explanation stories". For their employees it would be great to participate I'm sure, and for the company they will probably put it as "here's what we're doing for Corona" or something.

I'm not related myself to any such company, but happy to help in contacting anyone you choose and liaising, if there's any interest. I assume they'd want some acknowledgement somewhere and you'll have to decide if that's acceptable etc.

@waldyrious
Copy link

waldyrious commented May 3, 2020

Related: I like how these folks show 4 scenarios to guide readers into interpreting the graph.

To me that graph suggests that coloring the sections of the curves based on the local slope may help conveying the severity of each part of the trajectories, moreso than eyeballing the slope as we do today, or comparing the curves to reference shapes as the images in that tweet require.

Of course this couldn't be done for all lines simultaneously, but maybe it could be done when a given line is hovered (rather than highlighting the entire line in pink as we do now).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants