Page MenuHomePhabricator

More intuitive charts for languages stats
Closed, ResolvedPublic3 Estimated Story Points

Assigned To
Authored By
Danya
May 3 2026, 9:04 PM
Referenced Files
F79674515: image.png
May 4 2026, 8:29 PM
F79674447: image.png
May 4 2026, 8:29 PM

Description

The current charts for the per-language and per-gender breakdown charts are not very clear, especially the y-axis. A more intuitive and meaningful representation should be adopted.

Related Objects

Event Timeline

Danya triaged this task as High priority.May 3 2026, 9:04 PM
Danya set the point value for this task to 3.
Danya moved this task from Backlog to Important Priority Backlog on the Tool-humaniki-2 board.
This comment was removed by Danya.

Per-language chart:

image.png (770×1 px, 67 KB)

Per-gender chart:

image.png (780×1 px, 82 KB)

Well, you could take log(wikisize) as the y-axis, which would spread your data out more. However, there doesn't appear to be much correlation between wikisize and gender proportions. As in, smaller, less developed wikis don't tend to be more male-heavy as might be expected.

Therefore, you may be better off dropping the wikisize bit altogether. Draw a stacked bar chart for each wiki, in order of the main M:F ratio, with narrow bars.

Because of Wikipedia's nature as a tertiary source, it is inevitably going to reflect gender proportions in secondary sources. Which brings us back to the question of what do you want to do with these data? Identify wikis that are outliers with a statistically significant different proportion of articles than the norm?

Hi, thank you for the feedback ^^

The general idea behind this site is to show the gender gap on different WMF wikis and how it evolves over time. There is no specific goal beyond presenting multiple indicators that may be useful to other people on- and off-wikis.
The reason why I chose this representation for this specific chart was because the original Humaniki website used it as well, so I just copied it. But as you can see, it is not a very intuitive presentation.

Your idea of a stacked bar chart is interesting, I’ll have to check whether the library we’re using allows that. One thing I fear tho, is that for Wikipedias specifically, there are almost 350 language versions, so the chart might get a bit too dense and crowded when everything is shown.

Danya changed the task status from Open to In Progress.Sat, May 9, 5:35 PM
Danya moved this task from Important Priority Backlog to Doing on the Tool-humaniki-2 board.
Danya removed a project: Patch-For-Review.
Danya moved this task from Doing to Done on the Tool-humaniki-2 board.