I'm happy to announce my latest project, Data Navigator! π
I've been frustrated these past few years with all of the problems that visualization toolkits face making their visualizations more accessible. I built a tool that I hope will help.
Read more:
@TheWapplehouse
their face immediately melting into dissociated disbelief after they shook hands, eyes glazing over as their mind considers every possible interpretation of that event where they aren't a complete and utter failure
Data visualization cares disproportionately far too much about designing for colorblindness relative to other disabilities that are more common (visual impairments included).
(A thread on disability, race, and patriarchy in data visualization.)
@clapifyoulikeme
literally not even kidding but I think I know exactly who these people are and have for years, the quaker is actually the landlord and their "co-living situation" is just a boarding house
@RATFCKR
The game plays you, which makes it utterly genius. Think you're going to min-max for intellectual excellence? No. You're going to experience what everyone thinks of people who believe that is who they are supposed to be.
Who does accessibility at NASA? I legit would love introductions. I have ideas and also just would love to chat.
Their alt has been stupendous lately, to the degree I reckon there are content editors and SMEs involved in authoring it.
Here it is: (Big news + thread)
We have officially open-sourced our inclusive auditing system for data visualizations! πβΏοΈ
Introducing:
πChartability!π
Chartability is a toolkit to help people make more accessible dataviz.
You can learn more here:
Resources we could use more of (1/?):
Low vision (~30% of all people):
- High contrast text
- High contrast elements
- Using texture, shape, units
- Designing with zoom/magnification
- Using Hierarchy and Focus
- Using annotations or guides
The Nib, which brought us legendary pieces by so many creators (including this one below by
@MattBors
), is closing down.
Twitter posters in particular have so much to thank The Nib for. A serious cultural loss for the internet.
After ten years of publishing, The Nib is going to close down this summer. The new issue of our magazine, the Future issue, will be our last. Read the full statement from our editor here:
I can't believe game freak has me making a pokemon meme about weed cat. I am not apologizing to my followers or taking questions at this time.
And yes, this has alt text!
I recently shared my method for auditing the accessibility of data visualizations at a workshop in early Dec. It includes 10 critical tests that can be checked in 10 mins, plus 35 other tests.
Is there any public interest for me to open source this?
#dataVisualization
#a11y
I love the
#rstats
community.
Someone is like, "oh hey peeps, I saw a big need for this mundane but difficult task that I infrequently do, so I created a package that will literally scrape the last bits of peanut butter out of the jar for you. It's called pbplyr."
What a tribe.
I'm waiting for the renaissance of data visualization maximalism.
I want the most gorgeous, bespoke visuals you've ever seen. Intricate stories that invite you to read slowly and savor them. Visuals so dense you can revisit them for new information each time.
Resources we could use more of (3/?):
Cognitive disability (~11% of all people in US):
- Captions, summaries, clear titles, and plain text alternatives
- Reducing visual complexity
- Forgivable user interactions
- Use of hierarchy
- Assistive design (how-to-read guides, help)
Today's the day I FINALLY announce: I have accepted an offer to start a PhD from my top choice! π
@CarnegieMellon
's Human-Computer Interaction Institute (
@cmuhcii
, part of
@SCSatCMU
).
My research with
@cmudig
will venture into the uncharted territory of disability + dataviz.
Hey
#accessibility
folks, the New York Times is hiring an Accessibility Visuals Editor!
This is the first role I've ever seen explicitly devoted to accessibility and visuals like this.
It seems like an opportunity to make a huge impact!
Resources we could use more of (2/?):
Functional/motor impairment (~13% of all people in US):
- Keyboard interactivity/navigation
- UI alternatives to in-chart controls (brushing, subselecting, etc)
- Alternative data navigation schemes
- Scrollytelling alternates
@dataandpolitics
I always ask managers to tell me a story when they had to deal with a difficult employee and what they did, tell me a story of a time they had difficulty with their boss and what they did, how the work life balance is for their team, and the average tenure of employees under them
Resources we could use more of (4/?):
Attention deficit/hyperactive disorder (~9% of all people in US):
- Clear, short text summaries
- Object constancy
- Motion design and animation
- Use of breadcrumbs
- Interaction history (with undo/redo functions)
@ibogost
"I have a poor grasp of when brevity would be appropriate given the context at hand. In fact, I largely lack the ability to use brevity in any context at all."
It would be better if the band's nickname was just "fuck Brevity" though.
@ak47eating
@clapifyoulikeme
They "solve things" the quaker way (egalitarian, communal, democratic) and talk about being "in community with" each other. But only the husband/wife actually get equity and capital from the property. Their strategy rebrands landlording and makes it hard to organize/unionize.
"What are examples of accessible, effective chart descriptions?"
I get asked this question a LOT, so I am going to make a thread here to easily link to later.
(This thread covers both simple and complex use-cases for human-authored descriptions.)
Feel free to share/bookmark!
Okay, gonna start a thread where I compare "cool concept stairs"/DIY stair fails to data visualization.
You make fancy custom charts? Guess what? They are probably an accessibility liability. Use standards.
See Fig. A (the alt on each image will be a bonus roast explanation):
Standards for accessibility exist! Techniques, suggestions, and examples exist!
Learning it is the same as any other job.
Buying a house that wasn't built to standard code is a huge liability. But the standards exist for a reason. Same with software.
Use standards!
Woah.
Just.
Woah.
This write-up by
@ChristiansenJen
is an absolute must-read if you are in the sciences, visualization, infographics, illustration, or even remotely care.
I am just blown away with how thorough, thoughtful, and provoking this is.
Pick your own dataviz midlife crisis:
- do algorithmic art
- get a pen plotter
- collect esoteric, historical dataviz books
- participate in unpaid marketing for tableau every Monday, in hopes of making zen so the void in your heart can finally be filled
@dataandpolitics
"How's the work-life balance on your team?"
*laughter*
(I have a series of questions I ask for catching red flags, and this one usually gets a diplomatic response at worst... but laughter? Immediate reject.)
So now what??
My solution is simple:
Prioritize more discourse, research, guides, palettes, and tools that help with other functional disabilities and disabled communities.
And please stop writing about colorblindness in visualization. We have enough of those blogs already!
As this new surge is happening with COVID I cannot stress enough that you MUST add alt text to your charts and graphs on here or so help me...
It is literally a human rights issue when you fail to make COVID data accessible. I know some of you know you should too. JUST WRITE IT!
Why do we have far more resources for dealing with colorblindness (<4% global rate) than low vision, which affects just over 30% of people?
I'd reckon it all started in academia...
Oh this brings me immense joy! Long before I was in data visualization, I was obsessed with the power I discovered in organizing sessions when we would start to map ourselves in a space.
Mapping became a way to see how strong we are becoming. This is so powerful for solidarity.
If youβre as excited about union Starbucks as I am, hereβs a map showing all the union campaigns! Find one near you, go give them love and support (also cash tips)
πMAP:
I just want to be clear here: not all data visualization is about "quick" or "fast" understanding. It should be about cutting through chaos and finding insight.
Sometimes a good visualization is good because it encourages *close* reading! This means it might be a slow parse.
So IDEO largely built its marketing around the idea that The Designer is a higher-level thinker who can sift through all of the mass-produced junk that was rising up as a result of automation and find quality.
This was rooted in racism and was doomed to fail... (thread)
I suppose now is as good of a time as any to announce that I have been hired by
@Apple
to do cool things! (I'm sure you could guess what.)
I am working alongside some of my absolute favorite people ever, including
@domoritz
@ldgreef
@clb5590
.
Anyway, the list goes on. These are just a few!
Black, disabled people in the south are ESPECIALLY (disproportionately) excluded from the design of equitable data systems! Include them!
If data is really so important, why is it only accessible for a select few?
Why does this matter?
Because designers, scientists, and engineers in our field continue to produce palettes, guides, research, and tools for dealing with colorblindness when visualizing data.
But where are tools and resources for all the other kinds of disabilities out there?!
~4.5% of people with northern European ancestry are colorblind. But less than half of a percent of women are.
This means that nearly 8% of men from a northern European background have some form of colorblindness.
*Colorblindness affects WHITE MEN the most.*
The state of web accessibility is pretty depressing.
- Standards have been around 20+ years, but 97% of the top 1mil sites fail to meet them.
- Standards are hard to learn but automating accessibility is terrible and makes it worse.
- Standards are still the bare minimum. (cont)
Okay this is getting a little traction already! So come to my talk on Accessibility and Data Visualization with
@LareneLg
and
@liatrisbian
at
@OutlierConf
!
@dataandpolitics
To peers I would be working with I ask them to tell me a story when they had to deal with a difficult peer and what they did, what the work life balance is for the team, why they chose to work where they are at, and what the performance review process looks like.
As a data visualization professional (who happens to work in IT), I naturally wanted to see what brilliant "automated" charts our ticketing system reported by default.
And... wow.
Just... wow.
OH! I wanted to share what I got my partner for Christmas: a commissioned drawing of our samoyed by
@momodraws
!
It features:
Pizzelle "Pizza" the Samoyed surrounded by snowflake pizzelle cookies with a slice of pizza in her mouth.
The greatest commission ever!
Do you make data visualizations? Are you a data scientist, researcher, analyst, dev, or designer?
There is a great research project trying to discover how practitioners in our field work to make visualizations accessible (if at all).
Please participate!
Academic discourse on scientific visualization and color perception (alongside computer science) is decades old.
I'm sure there is a paper waiting to be written about the demographics of peer reviewers, but I can guess they were mostly white men.
Ask: WHO was this science for?
Okay, I want to be as gentle as possible with a review of this now-award-winning work but there are some serious problems we need to make sure we don't repeat.
Our technical HCI class yesterday *just* talked about
@elizejackson
's "disability dongle" and problematic access work.
π So, our
#CHI23
paper received a Best Paper Award, which is absolutely insane!
@seongkook
@aashikazim
This work explores how hand auto-pilot can help enrich the interaction capabilities of individuals with visual impairment.
Preprint:
Getting into accessibility can be so challenging (for technical/design/knowledge reasons), but few people talk about the fact that most of us are pissed off all the time too.
We extrovert anger, even righteous anger. And I just want to let new folks to the field know this.
We tend to view accessibility as a *burden* and one that falls to some singular individual who must do the work alone.
But communication is simply the responsibility of good science. And good science is a team effort.
I'd love to know the story here.
One of the biggest misconceptions about data visualizations is that they "speak for themselves."
No they don't. We assign them meaning and value and rhetorical force.
Since my wage theft graphic exploded on reddit last night, I've gotten a lot of dms about what 'wage theft' is and what you can do if you feel you have been wronged. Here are a few resources (no means real legal advice). UCLA has a good primer:
@clapifyoulikeme
Okay.... deleted my previous thread about this because apparently there are at least 2 Quaker "coops" in the area (what??) and I absolutely don't want people to get the wrong idea about one if it isn't the one I'm thinking of. What if these people are cool? I dunno.
Either that, or the SMEs themselves are writing it and have received training or coaching.
It's the sort of artifact where you witness it and know that there was collaborative work involved to make it happen. There's a story behind it I want to know.
Throwing this out into the universe:
I want (at a bare minimum) to have the
@DataVizSociety
do a fireside chat with the NASA folks on accessibility in scientific communication. It's work that I want to celebrate but also think others could really learn from.
While the origins of the imbalance here might be conjecture (because who has asked these questions yet?), it should not be a surprise that we have more resources on accessibility for a very specific disability of white men than virtually every other major disability.
Way too many people fundamentally misunderstand what accessibility standards are and what they do for design and development.
Doing something non-standard is actually fine if you know what the standards are and why your design is actually more inclusive and does a better job.
It's official now! My first real paper publication! I did itπ€
I am also super excited for the (eventual) talk that folks will see for this too.
Here's a teaser couple of slides, to get all you academics really worked up:
Unfortunately, I have yet to see even a single map or chart about the election results in France that have alt text.
Many charts fail to have minimum contrast too.
Everyone deserves access to this information! It's a human right, after all.
I'll give 5 tips:
Well, well, well I just discovered a beta website from the US Gov for "Data Visualization Standards."
Looks like it is structured similar to a design system. I wonder what the progress on this is?
As a data visualization professional I will argue that this is a valid use of a pie chart.
Why?
Because the big slices make the little slices look like insignificant, unreadable strands of data. Puny.
This (of course) is the point.
Academic research papers are:
- Hard to learn how to write
- Hard to write even when you know how to
- Hard to read
- Hard to share
- High risk/expensive
- Largely inaccessible to both the public and people with disabilities
But entire careers depend on producing these!
It is exhausting that alt text doesn't download with images. So much careful labor is lost the second we share things. And on the internet? We share things a lot.
What a shame that our file formats can't contain textual descriptions that persist in the same way the pixels do.
@dataandpolitics
I don't always get the time to ask all of these, which can sometimes be a red flag too - if the interview is timed to only give me 5 minutes or less for questions of my own, then that can be a red flag too.
Want to embark on a research path towards accessibility and data visualization?
My
#1
paper recommendation for students/learners in this space is
@alanlundgard
and
@crystaljjlee
's "Sociotechnical Considerations."
It's so informative and approachable!
I was told not to celebrate when my paper got conditionally accepted (fair, since it is terrible).
So here is my delayed celebratory tweet:
I had my first paper (and as first author) officially (fully) accepted! π
Expect a more official announcement with details in a month.
@dataandpolitics
I view these questions as a bare minimum and typically have many more I wish to ask (based on the responses). I didn't really take this full approach until my latest job hunt and I'm glad I did.
@Sblahov
By the way, it is incredibly hard to come up with ideas that aren't actually ableist... so many things are just legitimately straight up ableist.
Accessible data visualization is going to have more in common with the humanities than a lot of tech + data companies will want to admit.
This skillset is going to be sorely overlooked as this work becomes more and more prominent (already is overlooked, to be honest).
I'm sometimes asked, "why don't you consider yourself a data scientist?"
While our skills are similar, our success looks different:
Data scientists succeed when they have taught a machine to understand data with low error.
#DataViz
peeps succeed when they can do that with people
@KellyLepo
@NASAWebb
@stsci
@timrhueii
Oh, so glad you found my post! Your team's work is just stupendous. I really feel like it is some of the best alt text I've seen for scientific graphics. There really was a celebration of the science in every one.
One of the most beautiful things about Geralt's story is that it mirrors mine.
My surgeries took *most* of my pain away and saved me from imminent death. But the experience I am left with is *still painful* and very much, 100%, still a disability.
Mark Thompson π³οΈββ§οΈ @mustangsart.bsky.social
If I get one more angry gamer telling me "Fringilla Vigo healed Geralt actually" when she only managed to heal Geralt enough to alleviate the worst of the pain, I am going to lose my mind.
It still doesn't detract from Geralt's experience of disability and the fact he's still-
Lots of alt text authoring just isn't done at all. But when it is, it is either done by experts who don't adequately communicate clearly or by non SMEs who don't adequately communicate deeply.
Ah, it is hard to put into one short tweet how much I appreciate this article by
@ChanceyFleet
(so I'll do it in 3).
Her framing around "image poverty" for blind and low vision folks is important to think about, and of course, do something about.
π¨Oh my GOD it is FINALLY OUT so I can talk about it!π¨
They identified a pretty huge problem in most auto-/manual captioning and descriptions for charts: nobody asked what *kind* of information is actually important to convey.
So much work failed to ask this basic question!
Descriptions, captions, and alt text can help make visualizations more accessible. But what should they say, or not say? My new paper with
@arvindsatya1
contributes a model for evaluating textual descriptions of charts.
Accessible HTML Paper:
#a11y
/Thread
Hi everyone. Big news: I am leaving academia. Northwestern University has been great. I couldn't have asked for a better staff position at a top tier research institution. But I have been picked up by Visa Inc, to help build out their data visualization. Wish me luck!
β¬οΈ Thread of thoughts on this topic β¬οΈ
I've taught
#D3js
to nearly 300 faculty, staff, and students at primarily the grad/PhD level across various workshops and bootcamps.
Some key reasons behind why D3 is hard, I'll outline below (in case anyone is learning or curious)
The trouble with D3 is to build a visualization you must also have a deep understanding of SVG, DOM, JavaScript, geometry, color spaces, data structures, the standard model, and quantum physics
Can someone please tell me why it is a ubiquitous design for pasted text to retain its original formatting by default?
Who came up with this? What were they solving? I want names. I promise I just want to talk.
Someone asked me once, "what would make data visualization work more accessible for someone with chronic illness, like yours?"
And guess what? The answer isn't technical lmao.
My answer is being free to work in my own time and place, strong healthcare, fair pay.
One of the biggest problems in data visualization is that the practitioner experts in this field still expect every graphic to be a hand-crafted masterpiece by a chef who intimately understands every ingredient.
A grammar is a useful set of ingredients. But people need recipes.
As a disabled person, I did not "chase my dreams"
I chased employer-provided healthcare instead.
Inspirations like mine die because disabled people don't have their needs met. The effects of universal healthcare would be massive. It would bring a renaissance.
Every now and then I go back to this series on accessible color because someone brings it up (or mentions it was helpful).
It means so much to me to hear from members of the community that came across these notebooks and it had an impact in their work.
Exhausting. While this might be a cause for celebration for some, my preliminary audit shows that this is yet again another tool for easily creating inaccessible data experiences.
No SR or keyboard access, no semantic control of marks or their relationship to one another, etc
My favorite paper of the year just dropped at ASSETS: "Cripping Data Visualizations."
An incredible piece that frames and re-orients what accessibility is and should be. A great paper to read for accessibility, even if you aren't into visualization.
Eyyyy our team has officially been renamed from "Data Visualization Team" to "Data Experience Team" - more inclusive language for the work that we actually do:
Making data interfaces and systems that work for more than just the eyes. Disability inclusion is good.
Lots of good points have been made already but I want to add that COVID was a mass-disabling event.
There *are* a lot more disabled people now than pre-pandemic.
All this "the number of disabled students has risen, so it has to be cons" discourse is partly because some people cannot imagine anything other than someone hiding their disabled status in shame and secrecy for all of eternity. They're horrified by the thought of anything else.
I cannot stress enough how people's mental model of design in data visualization needs to mature beyond the expectation that one experience is the end result.
For example, EVERY time I bring up textures on bars, someone crawls from the depths to tell me they hate them.
Hey, not to potentially beat
@DataVizSociety
to a
#GAAD2023
announcement, but the next issue of the magnificent Nightingale Magazine will have a 12 page section on accessibility and visualization, featuring myself,
@tanner_jaime
,
@johnycassidy
and more!
@ditz_carlton
literally not even kidding but I think I know exactly who these people are and have for years, the quaker is actually the landlord and their co-living situation is just a boarding house