5

Suppose we have data under an open license and in some standard machine readable format. What has to be considered on top of that with respect to making the data accessible to people with disabilities, or their tools? Do any standards or best practice guidelines exist for that?

I am roughly aware of WCAG and ISO/IEC TR 29138-X:2009 as well as some key accessibility regulations and intiatives in the US and Europe (see my notes here), but none of them seem to really cover data.

Update: I am especially interested in scientific data, much of which can not usefully be displayed in tabular form – think Magnetic Resonance images or spectra, genome sequences and their annotations, or phylogenetic trees.

Daniel Mietchen
  • 389
  • 1
  • 9

2 Answers2

2

I'm going to assume that the data is in some graphical format -- if it's numerical data, then the visualization would be a function of the software that's using it.

For images, there isn't anything that I'm aware of that you can do for blind people. I've heard of devices that allow a blind person to trace a pen-like device over a surface, and it'll vibrate depending on what it's pointing at. I've also seen specialty braille books by Noreen Grice that take a few images and apply textures to them. (one of my co-workers was involved with her 'Touch the Sun' book)

If you're serving images via HTML, you can use the 'longdesc' attribute to provide a textual description of the image, with more specificity than would be appropriate for an 'alt' attribute.

Another consideration for images are the color-blind. 'Rainbow' style color tables are particularly bad not only for the color blind, but have also been shown to influence people in unintended ways:

The last one links to a number of other resources, including alternate color tables and scholarly articles on the issue. (note that AGU's EOS had a front-page article on this topic years ago, and did absolutely nothing in their editorial process, within a couple of issues they had front-page articles with the problem).

Joe
  • 4,445
  • 1
  • 18
  • 40
1

to answer part of your question, if you're displaying your data, the same standards that apply to html, should apply here: use lean, semantic markup; apply tabkeys when applicable; etc. extra special care should be taken here when marking up tables, as they tend to get out of hand.
as far as other user agents, i'm assuming the same standards apply to their tools.
there are validation services for most formats, which you should try and implement from the early on set, such as html/css/json/csv, etc.

they may not explicitly cover "data" in the title or subject, but any w3c wcag recommendation for tabular or list data applies

albert
  • 11,885
  • 4
  • 30
  • 57
  • Thanks - that goes into the direction I'm after. I just updated the post to reflect that I am interested in scientific data, which comes in lots of varieties that do not always map usefully onto tabular or list representations. – Daniel Mietchen Mar 17 '15 at 22:12
  • can you show me some examples. images i think are ripe for using figure and figcaption, and even long description if you want to go that route. etc. – albert Mar 17 '15 at 22:34
  • 2
    Here are some: MRI data, genomic data, patient samples from a specific study – Daniel Mietchen Mar 17 '15 at 23:11
  • lol. my bad...thats pretty broad. can you point me to one example of each? i don't want to download 39gb of mri data, etc. – albert Mar 17 '15 at 23:14
  • 1
    Let's start with one of the MRI datasets then (74MB). But I guess navigation of overview pages like the ones linked from above is also an accessibility issue. – Daniel Mietchen Mar 17 '15 at 23:44
  • this is rather difficult because i have no clue what i'm looking at. i could do the research....but this isn't my task ;) i have all the tifs. i converted them to jpg. i downloaded the pdf and scraped it. and the xls and converted to csv. now what? – albert Mar 18 '15 at 06:14
  • 1
    You are looking at the MRI scan of a sea urchin (Centrostephanus longispinus) and some auxiliary data. I just created a video from the 2dseq file and am uploading it to https://vimeo.com/122591408 now.

    Your "now what?" almost comes full circle with respect to the original post, which is about accessibility beyond licensing and basic machine readability. That might include catering to people who are not from the community that normally generates or analyzes such data.

     – Daniel Mietchen Mar 18 '15 at 23:41
  • sorry if you that i was pandering to you. i wasn't. i can do the markup, i just don't know what i'm supposed to be displaying on the page. and in what context. and how the tiffs relate to the data – albert Mar 19 '15 at 00:33
  • 1
    The tiffs basically contain the same information as the 2dseq file and the video. The context would be the one described by the original paper and the associated data paper or any other context that any reuser of these openly licensed materials might choose. It is thus also not defined upfront what to display on a HTML page, or whether that is even the right way to frame it. The point I am trying to get at is that when sharing to enable reuse, how best to cater to the accessibility needs of a broad user community. – Daniel Mietchen Mar 19 '15 at 00:40
  • i hear ya man. so basically mark up the paper? or something like that. you can email me jalbertbowden gmail or there's chat on here, but i've never used it – albert Mar 19 '15 at 01:00
  • 1
    In some cases (like the paper mentioned above), some form of markup is already available. However, that may not meet the needs of assistive technology that users of the data may be relying on, and it kind of assumes that the data would be consumed on the Web (as in the Vimeo video). But what about adding some assistive stuff to the .tgz archive for offline consumption? – Daniel Mietchen Mar 19 '15 at 01:18