Contact and corrections
Rankodex takes error reports, questions and suggestions through the issue tracker of its public repository on GitHub.
How to get in touch
Open an issue at github.com/wsoule/world-data/issues. Search the open issues first, in case someone has already reported the same thing; adding a comment there helps too.
Opening an issue needs a free GitHub account, and issues are public, so do not include personal information you would not want published. There is no email address or contact form.
Reporting an error
A report that can be checked quickly includes:
- the address (URL) of the page, copied from your browser;
- the place and the statistic, and the year if the page shows one;
- the value Rankodex shows and the value you expected;
- a link to where the right figure is published, such as the original page or the agency’s release, if you have one;
- for display problems, your browser and device, and a screenshot if you can.
Each topic page credits its sources under its title, which is the best place to start checking a figure. The sources page lists them all.
How corrections are handled
A report is checked against the source the page credits. What happens next depends on where the error is:
- Rankodex misread its source, for example a wrong unit, a misplaced row, a mislabelled column or a better direction the wrong way round: the fix goes into the data pipeline or the site’s code, so it applies to every page that uses the figure, not only the one reported.
- The source itself is wrong or out of date: Rankodex does not edit individual values by hand, so the figure stays as the source publishes it until the source corrects it. The issue records the problem, and the source’s own correction channel is the place to fix it.
- A wording or calculation problem, such as a misleading sentence or a statistic computed the wrong way, is fixed in the code.
Every change is made in the public repository, so the fix, the commit that made it and its date are visible in the history. The live site is rebuilt from that repository, and a correction appears once the next build is deployed.
Suggestions and contributions
Ideas for new statistics, sources, places or features are welcome as issues too. The code is open source, so you can also propose a change directly as a pull request; the methodology describes how the data is built.