What interactives can do that articles can’t (2018)

ejb.github.io

61 points·callumlocke··4 comments

Here's my favorite recent example:

https://www.pollbludger.net/2017/11/15/sex-marriage-survey-6...

One thing I like is that it's so simple. There's a scatter plot of Australian electorates. One axis is how they voted in the same-sex marriage plebiscite, and you choose the other: average income, percentage over 60 years old, how many speak a language other than English, etc.

As you flick through the options, you get one amorphous blob after another. Then you choose "percentage of atheists", and you see the kind of straight line that usually comes from a physics experiment.

I'll be honest -- this graph gives me a headache to interpret -- I wish there was at least a single sentence elaboration of what the dropdown values indicate.

For example, selecting "Secular/No Religion" I see that Sydney scored about 83 on "Yes vote for SSM" (meaning 83% of those voting in Sydney voted in favor, right?). It scored about 44 on "Selected indicator" (in this case "secular", right?).

So -- 44% of Sydney self-reports as "Secular" -- right?

What the original blog post was referring were article length pieces of content that mix interactive and static elements to tell a story. The pollbludger has interesting data but it's not giving enough analysis or insights into selection you can pick from the selection.

Here's a great example, popular on HN just a couple of days ago:

https://www.meltingasphalt.com/interactive/going-critical/

Reply on news.ycombinator.com