Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/gvwilson/10-accessible

Ten Quick Tips for Making a Scientific Publication More Accessible
https://github.com/gvwilson/10-accessible

Last synced: about 2 months ago
JSON representation

Ten Quick Tips for Making a Scientific Publication More Accessible

Awesome Lists containing this project

README 

        
# Ten Quick Tips for Making Scientific Publications (or Presentations?) More Accessible



[Silvia Canelón](https://silvia.rbind.io/),

[Maya Gans](https://maya.rbind.io/),

[Elizabeth (Liz) Hare](http://doggenetics.com),

[Yim Register](https://students.washington.edu/yreg/),

[JooYoung Seo](https://ischool.illinois.edu/people/jooyoung-seo),

and [Greg Wilson](http://third-bit.com).



## Tip 1: Remember that people face many different challenges.



Discussion of accessibility often focuses on vision challenges,

but people may face many other challenges.

For example,

one author of this paper has difficulty typing and using a mouse because of repetitive strain

while another is on the autistic spectrum,

which leads to difficulty picking up social communication cues.

Other people may have difficulty hearing

(which is an increasing problem as video instruction becomes more common),

and even people who can hear well may have poor English comprehension.



A corollary to this tip is that challenges can change depending on a person's physical environment:

Someone who is able to understand an audio recording while wearing headphones

may not be able to follow it if there is a lot of background noise,

while someone with chronic fatigue syndrome (CFS)

may be able to comprehend things better at some times than at others.

And as [Johnson2017](#Johnson2017) points out,

individuals' challenges usually increase over time as they age.

(In particular,

journal editors and grant reviewers are usually chosen based on experience,

which means they are frequently in the upper half of the age distribution curve in their fields…)



## Tip 2: Ask.



If you do not face a challenge yourself,

it may be hard for you to know how best to address it,

and even if you have educated yourself,

assistive technology may have changed

or the people you are trying to help may have different needs from those you have worked with before.

Asking what to do is always appropriate,

and is also basic politeness.

"Nothing about us without us" has been a political rallying cry for more than six hundred years,

and was adopted by disability activists in the 1990s to signal that

people facing challenges are both able to decide what help they need

and have a right to do so [Charlton1998](#Charlton1998).



The hard part, of course, is knowing what to ask about.

For example,

if English is your first language

you may not have noticed the assumption in the first paragraph of Tip #1

that all scientific communication uses it.

Equally,

if you do not have arthritis,

it simply might not occur to you that

selecting a link on a touch screen e-reader may be difficult or impossible.

[UKHO](#UKHO) is a good short overview of common challenges;

Appendix 1 summarizes its recommendations,

and can be used as a checklist.



## Tip 3: Design for the web but check the print version.



The paper of the future might be an interactive computational notebook,

but the paper of today is still exactly that:

paper.

Even disregarding accessibility issues,

saving an HTML page as a PDF usually doesn't produce an attractive result.

Whether you use WYSIWYG tools like Microsoft Word and Google Docs,

or document compilers like LaTeX and R Markdown,

you should save both HTML and PDF directly

rather than trying to convert one to the other.



Web pages are more typographically constrained than printed pages,

so you should limit your design to things browsers can directly support,

and you should use web markup in semantically correct ways.

For example,

you should only use heading tags for headings,

not as a way to enlarge text,

and should only use tables to display data,

not to create two-column layouts for pages.



## Tip 4: Use tools to check for accessibility issues.



Accessibility guidelines can be overwhelming,

even for people who are familiar with them,

so use software tools to check for common problems.

For example,

[WebAIM WAVE](https://wave.webaim.org/) is a browser plugin that checks things like:



- web pages have title metadata

- images have `alt` text

- tables have headers

- heading tags are used in the correct order

  (i.e., the first is level 1, there are only level 2 headings directly below it)

- text and background colors are different enough to be read easily



[Microsoft Word](https://support.microsoft.com/en-us/office/improve-accessibility-with-the-accessibility-checker-a16f6de0-2f39-4a2b-8bd8-5ad801426c7f) now comes with an accessibility checker as well.

These tools are not infallible:

for example,

they often don't point whether a page's styling makes it hard to understand

because screen readers present information linearly even when clever tricks with CSS

rearrange the apparent order of material on the screen.

However,

by catching common issues they leave you more time to look at ones

that still require human intervention.



PDFs found in the wild are notorious for being inaccessible.

In particular,

PDFs generated from PowerPoint slides are usually incomprehensible to screen readers,

and almost always lack alt-text descriptions of diagrams, pictures, and data visualizations.

If you have to submit a document as a PDF

(for example, as a supplementary document as discussed in Tip 10),

ensure that any accessibility aids you implement in the original document are not lost in the conversion process.

WebAIM provides guidelines for [converting to PDF](https://webaim.org/techniques/acrobat/converting#save).



## Tip 5: Describe visualizations---it helps everyone.



Accessibility challenges for data visualization have often been thought of in terms of blindness or low vision.

However,

the cognitive load of figuring out complex visualizations is also a barrier for people who are neurodivergent,

and the trend toward explorable (interactive) visualization in web publication

makes life difficult for people with physical or motor challenges as well.

As a rule of thumb,

if you can't figure out how to write informative alt-text for your graph

then that graph is probably too complex.

And since most publishers don't support alt-text in print,

you should include a complete written description of of your findings in the results section

[Goring2017](#Goring2017)

[Lundgard2021](#Lundgard2021).



Similarly,

the "Methods" section of every paper should have enough information

for the reader to replicate the study.

This isn't accessibility-specific,

but process descriptions that are presented as flow charts or diagrams

should also be documented textually as well.

Doing this will help search engines as well as people,

which can only help increase readership [Lin202](#Lin2020).



Consider accompanying visualizations with a

clear tabular representation of key data points, 

possibly in a supplementary document as described in tip 10. 

These supplementary tables could be referenced within the manuscript

along with the figure reference and/or in the figure caption (e.g. Figure 2, eTable 2)



## Tip 6: Provide captions and transcripts for videos.



Even before the COVID-19 lockdown,

the move toward web-first scientific publication meant that

a growing number of scientific publications referenced or included,

and recorded lessons and conference talks are clearly here to stay.

In order to be searchable as well as accessible,

these should always include descriptive text and a transcript of what is being said.

The former is needed by people who cannot see the video (or cannot see it well),

while the latter helps people who cannot hear

or might struggle to understand the language being used.



Creating transcripts is not as time-consuming as it once was.

Video hosting services like YouTube can use speech recognition to generate a starting point;

while this invariably contains mistakes (particularly for technical jargon),

editing this text is usually shorter than trying to transcribe the entire video yourself.



## Tip 7: Make publications open access.



Open access does not level the academic playing field:

reliance on article processing charges (APC) skews publication

toward scholars with greater access to resources and job security [Olejniczak2020](#Olejniczak2020).

More specifically,

"…the likelihood for a scholar to author an APC OA article increases with male gender,

employment at a prestigious institution…,

association with a STEM discipline,

greater federal research funding,

and more advanced career stage (i.e., higher professorial rank)."



However,

people with disabilities are usually less affluent than people without,

both because of the opportunities that are denied them

as because they have to pay for aids and services that able people don't need.

Paywalls are therefore an extra, unfair burden for the disabled,

just as they are for people from less affluent countries and researchers at non-elite institutions.

Simply put,

if someone cannot afford to read an article,

the presence or absence of captions and descriptive text is moot.



## Tip 8: Try not to go outside your own expertise, especially when it comes to disability or mental health.



Especially in computational social science or human computer interaction (HCI), we deal with complex topics such as mental health, disability, discrimination, cyberbullying, politics, and more. Almost certainly there will be design recommendations or analyses of mental health that are done by computer scientists without the relevant expertise. In Human Computer Interaction we seem to encourage the idea that we can be psychologists, designers, software engineers, and policymakers all from our laptops. This is usually not the hubris of an individual, but the general culture of social tech in academia: the "there's an app for that" mentality. Reviewers may even ask "what are the design recommendations?" at the end of each piece of research. Mentors may encourage students into areas they are not qualified or _certified_ to tackle. Due to lack of funding or strict deadlines, researchers may not adequately consult the experts actually necessary for the work. The culture of academia seems to be pushing us all to have an answer, an app, a social commentary, and a research plan in order to "help" others and "help" society. Our main concern should be helping people to represent themselves. As an autistic individual who has worked very hard on my own self-advocacy skills, there is no greater gift than helping others speak up for themselves. When in doubt, amplify the work of those most affected by a phenomenon. Never make design or policy recommendations outside of your own expertise.



In the case of research about autism, the situation is particularly dire. Countless works about applications or technologies to help autistic children drown out the work of actually autistic member researchers. Many technologies suggested for autistic children are for the purpose of assimilating the child into neurotypical ways of doing things instead of celebrating their neurodiversity. Not all assistive technology for autistics is bad, but the lens through which we write about it tends to reveal problematic stances from non-autistic researchers. We do not need to be helped to behave as neurotypicals do, and we certainly do not need assistive technology to collaborate or communicate as neurotypicals do. There should be more research that teaches neurotypicals to collaborate and communicate with autistic people, amplifying our way of doing things as equally correct and feasible. Research articles will often give recommendations that simply do not apply to our autistic culture, way of life, or way of processing information. They may even cause more harm, especially if they are picked up by large corporations as evidence-based recommendations for their platforms or workplaces. In the end, it is important for all of us to stay within the realm of our expertise while also consulting those with lived experience (and compensating them appropriately for their labor).



_this is all pretty hodgepodge I need to revisit what I'm trying to say_



## Tip 9: Use formatting consistently and with numerous section headers (clearly differentiable in the text).



### Keep the Outline

As an autistic researcher and writer, the outline is the most important part of any written work for me. Nested in the various bullet points is the *structure* of the thoughts. A map of a comprehensive argument that really only needs additional evidence to bolster it's claims. At the end of the day, my brain is looking for a series of logical steps that concludes something; even if it's small or open-ended. When I read long paragraphs of prose, I tend to get distracted and confused about the main takeaways, or how it fits the overarching narrative. Having the outline still visible instead of assumed really helps me to keep reading. What I mean by "still visible" is clear headers and topics, and my personal preference is for lots of lists and tables (formatted to be screenreader accessible). Refer to Using Headers below.



### Use a Story

My brain is constantly looking out for anchors that ties one section to the next; a coherent story being told through research. In fact, I've even been known to use stories themselves to weave throughout the entire piece; using an exemplar persona or situation to carry the reader through the argument. Of course, no persona will capture a universal experience; but it will certainly explicitly tie the reader to this character's arc. Perhaps in traditional research when we make claims about people, we assume a certain level of implicit empathizing from the reader. While I am highly empathetic, my brain simply does not spontaneously think of anyone but my own perspective, as I lack skills in something called "theory of mind". Therefore, I need a story to hold on to. The persona or situation may be something I have connection to, and then it will not serve as a good narrative for me. But in my experience it is better to have some concrete example to weave throughout the research than to have open-ended possibilities.



### Using Headers

Most of us have seen "Introduction, Methods, Results, and Discussion", helpful guides for our minds as we navigate a research paper. While this may not be your preference, for me I like to have guiding headers for nearly every paragraph! What is this paragraph contributing? What is the main point? If I read through just the headers of this paper, will I know what it's talking about? I want my readers to be able to skim my work and get the right takeaways, even if I wished they'd read the entire work. Similar to Keep the Outline, the headers guide the argument. For me personally and with my specific disability, I rely on visual cues a lot to navigate a wall of text. I need it to be broken up with fonts, figures, and section headers in order to continue attending to it. I have no specific opinion on mandated page limits, but I do know that page limits have forced me to be more concise in my writing; which tends to be better for a [hyperlexic](https://www.healthline.com/health/hyperlexia#hyperlexia-and-autism) like me. With hyperlexia, I am particularly interested in structure of words and how they are organized, as opposed to the meaning or content. Generous use of headers helps to keep me both focused on the content while also organizing the content into easily processed chunks. It might result in having Section numbers like 2.3.1.2 (a) but so be it! \footnote This is mostly a joke, as more and more numbered sections can also be a large distractor. I tend to omit the Section numbers and refer to them with hyperlinks or their actual titles.



### Making Use of Bold Text

I have been critiqued for my use of bold text in my work, but I know that it helps me to process what I am reading. I usually employ bold text for my research questions, hypotheses, and keywords present in any qualitative data. The goal of my writing, perhaps tailored to my own disability, is to be as obviously summarized as possible. The autistic brain prunes connections between neurons less than the non-autistic brain. I often joke that autistic research often comes off as a conspiracy theory in the way we "link the red thread" together between our ideas. I often use bold text to highlight the important evidence for the main claims, and enjoy when I see it used in research papers. I would try to use bold text in a consistent manner, and less generously as headers. But it can be an important tool for linking together evidence to the main claims. 



### Background, Research Questions, Methods, Main Contribution, and Ethical Considerations

I have a color coded system for how I read any and all research papers. The way that I _read_ papers may be of interest for how to _write_ papers. Below is the schema I use when reading any academic articles. As I read, I highlight according to the schema and it guides what I am looking for and how I summarize everything that I read.



* :orange_heart: **Background, Literature Review** _(important things we "already know" that are relevant to this paper's RQs)_

* :heart: **Research Questions** _(what is this paper going to try to answer?)_

* :green_heart: **Methodology** _(main methods used, interesting methods I haven't heard of before, specific and relevant details of the method)_

* :blue_heart: **Main Contribution** _(what did they determine? What are they giving us? What do we take away from this specific work?)_

* :purple_heart: **Ethical Considerations** _(did they specifically address any bias? consider marginalized identities? Or is this sentence something that hints at lack of ethical consideration?)_



Please note the use of bullet points, colors, bold text, and italic text in the above list. That is my autistic brain on a plate. Enjoy. _this is all written very informally I will be happy to edit it to sound professional, probably_



## Tip 10: Ensure that supplemental information and data are accessible.



Journals will vary in their editing and publishing practices and often publish supplementary documents "as is," which means you get to make the editorial and design decisions that readers will experience while reading. With this in mind, implement accessibility practices in the creation of any supplementary documents and treat them with the same care as you do the manuscript. Some of these practices are highlighted in tip 3, like using heading tags for headings.



Just as you take care to orient the reader in a manuscript's abstract and introduction, orient them to the information in the supplementary file on the first page of the file. Include contextual information about the manuscript and a functional table of contents. The contextual information can include details like the manuscript title, journal, and first author re-orients the reader to the proper manuscript. The table of contents can provide direct links to headings in the document that correspond to each figure and/or table, helping readers navigate the document and quickly find what they are looking for.



If you supplement visualization descriptions (see tip 5) with a tabular representation of the data, include only the key data points. An example of this approach can be found in the [WebAIM findings from screen reader user survey](https://webaim.org/projects/screenreadersurvey9/#disabilitytypes).



Use an accessible font size and font face throughout the document. Readers will likely be able to magnify or rescale text to suit their preference, but all text in your document should be a reasonable size to begin with. When it comes to accessible font faces, you will find a great resource in the [guidelines on how to make more informed font choices](https://medium.com/the-readability-group/a-guide-to-understanding-what-makes-a-typeface-accessible-and-how-to-make-informed-decisions-9e5c0b9040a0) written by Gareth Ford Williams from [The Readibility Group](https://www.thereadability.group/). The Readibility Group has also shared findings from a survey study about font preferences which included 2000+ participants. Among these were participants with dyslexia characteristics and participants with poor near vision. The talk is available online and titled [Don't Believe the Type!](https://youtu.be/h8IOqUl1zII?t=1029).



At this point you have made considerable effort to make the supplementary file as accessible as possible. The last thing you want is for your effort to be undermined by a lack of communication with the journal about accessibility. Once your paper is in the revision/submission process, ask the journal to publish the version of the supplementary file that is most accessible. The journal may not be implementing accessibility practices in all areas, so consider communicating why accessibility is important to you and to the journal's readership. You might also consider housing the supplementary information and data in HTML format in an online repository or website that can be linked somewhere in the manuscript or reference list.



## Bibliography



Charlton1998

James Charlton:

Nothing About Us Without Us: Disability Oppression and Empowerment.

University of California Press, 1998, 978-0520207950.




Goring2017

Simon Goring, Kaitlin Stack Whitney, Aerin Jacob, Emilio Bruna, and Timothée Poisot:

"Making Scientific Content More Accessible."

Authorea, Inc.,

2017,

https://doi.org/10.22541/au.150844289.92609826.




Johnson2017

Jeff Johnson and Kate Finn:

Designing User Interfaces for an Aging Population: Towards Universal Design.

Morgan Kaufmann, 2017, 978-0128044674.




Lin2020

Sarah Lin, Ibraheem Ali, and Greg Wilson:

"Ten quick tips for making things findable".

PLOS Computational Biology, 16(12), 2020,

10.1371/journal.pcbi.1008469.




Lundgard2021

Alan Lundgard and Arvind Satyanarayan:

"Accessible visualization via natural language descriptions: a four-level model of semantic content".

IEEE Transactions on Visualization and Computer Graphics, 2021,

10.1109/tvcg.2021.3114770.




Olejniczak2020

Anthony J. Olejniczak and Molly J. Wilson:

"Who's writing open access (OA) articles? Characteristics of {OA} authors at Ph.D.-granting institutions in the United States".

Quantitative Science Studies, 4(1), 2020,

10.1162/qss_a_00091.




UKHO

UK Home Office:

"Designing for accessibility".




## Appendix 1: UKHO Checklist



  

    For people on the autistic spectrum

  

  

    Do

    Don't

  

  

    use simple colors

    use bright contrasting colors

  

  

    write in plain language

    use figures of speech and idioms

  

  

    use simple sentences and bullets

    create a wall of text

  

  

    make buttons descriptive

    make buttons' meaning vague and unpredictable

  

  

    build simple and consistent layouts

    build complex and cluttered layouts

  

  

    For users of screen readers

  

  

    Do

    Don't

  

  

    describe images and provide transcripts for video

    only show information in an image or video

  

  

    follow a logical linear layout

    spread content all over the page

  

  

    structure content using HTML5

    indicate structure with text size and placement

  

  

    build for keyboard-only use

    require mouse or touch-screen use

  

  

    write descriptive links and headings

    write uninformative text like "click here"

  

  

    For users with low vision

  

  

    Do

    Don't

  

  

    use good color contrast and a readable font size

    use low color contrast and small fonts

  

  

    publish all information on web pages

    bury information in downloads (particularly PDFs)

  

  

    use shape and text to indicate meaning as well as color

    rely only color alone to convey meaning

  

  

    put buttons and notifications in context

    separate controls and actions from their context

  

  

    For users with dyslexia

  

  

    Do

    Don't

  

  

    use images and diagrams to support text

    create a wall of text

  

  

    left-align text and use a consistent layout

    underline words, use italics, and write in ALL CAPS

  

  

    give reminders and prompts

    require people to remember things from previous pages

  

  

    offer autocorrection hints

    require exact spelling

  

  

    For users with physical or motor challenges

  

  

    Do

    Don't

  

  

    make large clickable actions

    require precision

  

  

    space out clickable items

    bunch clickable items together

  

  

    design for keyboard-only or speech-only use

    make dynamic content that requires mouse interaction

  

  

    design for mobile and touch-screen devices

    have short time-out windows

  

  

    provide shortcuts and auto-fill

    require a lot of typing and scrolling

  

  

    For users who are deaf or hard of hearing

  

  

    Do

    Don't

  

  

    write in plain language

    use complicated words or figures of speech

  

  

    provide subtitles and transcripts for videos

    put content in audio or video only

  

  

    follow a logical linear layout

    spread content all over the page

  

  

    break up content with sub-headings, images, and videos

    create a wall of text

  

  

    For users with anxiety or post-traumatic stress

  

  

    Do

    Don't

  

  

    give users lots of time to complete actions

    have short time-outs or prod users

  

  

    explain what will happen after completing an action

    require users to guess what will happen after taking action

  

  

    highlight the most important information

    leave users guessing about whether they know what they need to know