Skip to end of metadata
Go to start of metadata
Learn more about the process of increasing accessibility in Word, PDF, and other document types. This guide reviews skill-building resources and details the most common types of barriers in Geneseo's Canvas course materials. 



Just need to improve the accessibility of particular documents? We can help with that. Use this Document Accessibility Request to get in touch with us.


Getting Started with PDF Accessibility Remediation

SUNY Geneseo offers several helpful services to help everyone feel comfortable recognizing and addressing issues of document accessibility. 

  1. Request access to Deque University for self-paced accessibility courses (CIT JIRA form)
  2. Log into Deque University. Select “My Courses.” Start with “Accessibility Fundamentals: Disabilities, Guidelines, and Laws.”
    1. Estimated time: 90 minutes
  3. Continue with Deque’s “Basic Web and Document Accessibility for Content Contributors” 
    1. Estimated time: 90 minutes
  4. Read CIT Self-Help Guide: Blackboard Accessibility Reports for Canvas
    1. Estimated time: 10 minutes
  5. Read Blackboard Ally Guides: Ally for LMS Help for Instructors
    1. Estimated time: 30 minutes
  6. Watch this video demonstrating a helpful “shortcut” to making fully accessible PDFs using Blackboard Ally: https://youtu.be/M2HJE7Y2Px0
    1. Actual time: 7 minutes

Full access to Deque University's accessibility course series is free for all SUNY employees until 2/20/21. 

Go Deeper with Document Accessibility

Depending on what documents you’re working with and the challenges you have, you may need to read further on the particular issue and how to address it. 

Additional resources can be found as needed on the following CIT Self-Help pages: 

Deque University also offers a series of courses for Document Accessibility. After registering with Deque through the steps noted above, review these courses individually as needed, or go through the entire sequence of 10 courses to earn an additional certificate. The most targeted courses to address common issues include: 

  • MS Word Accessibility Techniques (NEW version)
    • Estimated time: 75 minutes
  • MS PowerPoint Accessibility Techniques
    • Estimated time: 15 minutes
  • MS Excel Accessibility Techniques
    • Estimated time: 135 minutes
  • Basic PDF Accessibility (NEW)
    • Estimated time: 45 minutes

Why This Matters: Canvas's Accessibility Report

The Blackboard Ally tool gives us incredible insight into the level of accessibility of Canvas course materials. Ally analyzes the structure of Canvas pages and uploaded files to look for problems that might trip up students who rely on screen readers, keyboard-only navigation, text magnifiers, or other assistive technology. It provides robust guidance to improve the accessibility of course items, so that ALL students have an easier time accessing course content in the way that is most useful to them at that moment. 

At the time Blackboard Ally was first installed, Spring 2020, Geneseo Canvas courses included over 94,000 files. These include

  • 20,524 PDF files 
  • 11,377 Word files
  • 10,277 Canvas pages
  • 6692 PowerPoint and other presentations


That’s a lot of content! And the good news is that most of these items score well on Blackboard Ally’s measurements. 

But there is also room for improvement. Ally notes 37 patterns of accessibility error across all of these items, categorized as Minor, Major, or Severe. Let’s focus on the Top 10.

Ranking

Type of Problem

# of Files

Level of Severity

1

The document has contrast issues

11,727

Major

2

The document does not have any headings

10,262

Major

3

The document does not have any language set

10,232

Minor

4

The document contains images without a description

10,176

Major

5

The document is untagged

9,877

Major

6

The image does not have a description

9.740

Major

7

The document is missing a title

7,322

Minor

8

The document has tables that don’t have any headers

5,548

Major

9

The document is scanned but not OCRed

4,443

Severe

10

The image has contrast issues

986

Major


Ally offers excellent coaching about each issue within a course: it explains what the problem is, who it can potentially impact, and how to fix it. Many of the issues on this Top 10 List are fairly easy to fix, and Ally walks users through each step involved. 

Two of the items on this list are a bit more complex, however, and are worth looking at more closely. Both relate to PDFs, and require the most time to remediate of the items on this list.  

#5: The document is untagged

“Tagging” a document refers to the invisible labels associated with a document structure that indicate semantic elements like page title, headings, lists, and columns. These invisible labels help screen readers and other assistive technology determine reading order, and make sense of a document to non-visual users. 

The lack of tagging can apply to all document types, but frequently shows up as an error on PDFs in particular. Tags may be present in a Word document but get lost when that document is saved as a PDF. 

#9: The document is scanned but not OCRed

OCR stands for Optical Character Recognition, and means that a physical document like a book page has been scanned in such a way that a computer can read the type on the page. 

Some types of scanning result in image files, so that all a computer “sees” on that scan is one picture instead of readable text. Obviously, this is not ideal for users with low or no vision relying on a screen reader. 

Screenshots are one common type of non-OCR text. The screenshot below contains text from the paragraph above as an image file, which is inaccessible to users of assistive technology. If you cannot click on or highlight words with a mouse, that’s a sign that the text is not OCRed.

Example of text saved as an image file, non-OCR

Milne Library has a document scanner that reliably produces OCRed text. It’s often not possible to obtain the original item that was scanned, however, to scan it again in an accessible way. In these cases, the PDF file needs to be manipulated, or exported into Word or HTML, in order to improve its accessibility. 

Limitations of Blackboard Ally

Ally provides exceptional guidance to help us identify and address many accessibility concerns in Canvas courses. Like any other digital tool, however, it has its limitations. These are important to be aware of.

Even if a course shows a high accessibility score, it still may have significant barriers to accessibility. 

Human evaluation is required for any complete accessibility review, and this is true for Canvas courses, as well. For a sense of what the advantages and disadvantages of tools like Ally are, read “Technology Can Help Address Digital Accessibility--To an Extent” from Inside Higher Ed. 

Some limitations are described more fully below. 

Ally Can’t Identify Videos Missing Captions

Ally can note whether a YouTube video includes captions if it is embedded in a Canvas course. It has no idea how accurate those captions are, however. It also can’t see if a video uploaded from any other source, or recorded directly within Canvas, has captions and/or transcripts available, or not.  

Ally Can’t Identify Audio Files Without Transcripts

Like with video files, Ally can't "see" into the resources surrounding an audio file such as an MP3 to know if an accompanying transcript is available. All audio files should have a text equivalent option. 

Ally Assumes All Library Resources are 100% Accessible

If a library citation is associated with a PDF using Ally, that document’s score goes up to perfect, regardless of the quality of the PDF itself. Unfortunately, many journal articles and other resources that are available through Milne Library are not fully accessible, despite Ally’s confidence in them. Library staff can help improve their accessibility over time, but they will need to work with you on the documents needed for your course. 

Ally Assumes Every Image Needs Description Text

Ally can falsely flag an image as not accessible if it doesn’t have “alt text,” or descriptive text, associated with it. Not every image needs description, however, such as if it is purely used for decorative purposes.

Related articles


More Help

For questions, contact the CIT HelpDesk by calling (585) 245-5588, or visiting our online service desk.