💜 Being part of the session on accessibility in official statistics with so many good speakers including @statsgeekclare.bsky.social and @cararthompson.bsky.social

Haha, thanks. It looks like something our third year composition tutor would have been proud of - but we'd need to add a quick key for how to play the score. Forearms to the keyboard at Bar 5 with a fun distort+delay pedal on hooked up?

I was thinking the same - really effective in drawing the eye to what would otherwise cause the other graphs to have a scale that drowns out the patterns.

You can find my code here if you want to retrace my steps!
gist.github.com/cararthompso...

Some (rightly) can't use raw data because of important data around protecting anonymity, but we talked a lot about how visualising (or simulating) raw data can help bring nuance to visualisations which is often otherwise lost.

Humanising & anonymising - we tried to find a happy medium!

We chatted about all the stuff I normally talk about (colours, accessibility, different types of data visualisations, Gestalt principles that help us organise things visually, typography, and data-to-viz workflows). One of the key challenges for this field is the complexity of the data.

I always love receiving my copy! Such a joy to read and an invitation to slow down away from a screen.

If the main consideration is cost could there be an option of folks sponsoring some of the int. delivery to make it more accessible to others? No notion of the scale involved - worth exploring?

Always good to see a few variations on the theme! People use the style setting syntax differently so it's fun seeing the possible permutations.

Thank you! I still use it all the time and am humbled by how many other people find it useful.

I've also spotted the alignment difference. I like to control hjust *and* halign, so that the text inside the box aligns to the data point I'm highlighting.

Fully appreciate at this stage I should be opening GH issues - very happy to do that instead! Please don't feel pressure for immediate fixes!

And a final thought - I'm guessing if you want to change both the size and the colour the best approach is to set up a specific style (say, if I want the count to be big and black, and the rest of the text to be small and grey).

Really appreciate your responsiveness on this!

Here's the original with ggtext, in terms of line spacing. And the equivalent with marquee with the \n\n approach.

That was quick! Thanks. Works like a charm!

The last piece of the puzzle, if I may, is getting a line break in there too. What's the best way to do that?
- \n doesn't break the line, \n\n gives me more space than I'd like
- if I hit return twice, we get some funky formatting...

FWIW my initial "I wonder if this would work" was to try {.size=32}.

Having the option of setting size on the fly rather than needing to predefine things would be great, but I do also see the benefit of the {.big} approach for use across multiple plots.

That could be really useful! Thanks for being open to doing something like that!

Perfect! It's a bit more verbose than I'd hoped for but that's helpful, thanks!

Just keeping the <span> syntax doesn't do the trick - what's the new incantation I need?

Oops, turns out it was the N500 - this was a long time ago! This paper I published after finishing has a good few relevant references in it. Thank you for the excuse to revisit it!

www.researchgate.net/publication/...

I did that in my PhD, looking at the N400 and comparing it to a musical N400 for chords borrowed from a nearby key. Finished in 2010 so very conscious the literature has likely moved on, but I really enjoyed it! Happy to pass on some key references I built on if you don't mind a bit of a time lag 😊