Build a Categorization Quiz that is Accessible

May 28, 2021

Hi Everyone.  I am trying to build robust accessibility features into my projects. I have an instructor who wants a drag-and-drop (which is not accessible) with three drop targets. So it is essentially a categorization quiz. It is a medical symptoms question, with each symptom being categorized as Acute/Chronic/Both.  In some systems you can build a matching quiz that has more stems than targets and it becomes a categorization quiz, but I don't think Storyline does that. 

In a recent project, where an instructor wanted a drag and drop, I ended up building a parallel/equivalent question that was just a multiple answer quiz, but that question only had two categories, so I could just say "which one is NOT in category A and belongs in category B". Then I set a variable based on user input about assistive tech, and used it to route to that question instead of the drag and drop.  That won't work with 3 categories/levels. (and parallel/equivalent content for accessibility is not really best practice).

Does anyone have any suggestions?

- Scott Standifer, CFDE office, Southern California University of Health Sciences

5 Replies
EdTech Account

Hi Walt.  Thanks for that idea!  On slide A, I don't think your original version of it would be accessible for screen reader users, since they would have no way of knowing which acute/chronic/both button went with which symptom.  But I am thinking I could embed a longer title for each button in the alt text, where they would "see" it, without cluttering up the visible screen with all that extra text.  So that could work....  

Interesting way of scoring, too, with two invisible boxes acting as proxies for "correct" or "incorrect"  in a pick one quiz.  I may try using a multiple answer quiz directly on the radio buttons, to keep the programming simpler, if I can make the layout work. 

Slide B is also interesting - but I am not sure a screen reader user would be clear on how the interaction works.  I will have to think about it. 

Walt Hamilton

I guess I was envisioning text instructions of the nature of, "First select Symptom 1, Symptom 2. or Symptom 3. Then select Acute, Chronic, or Both." They were built in that order, even though Both was moved to the top of the Z-list. But you may be more experienced with this sort of thing than I.

Scott Wiley

Hi Scott,

I've had to deal with many of this type of DnD question for the same usability reasons. For me there are two ways, depending on if you need to have feedback for each question/answer selection, or overall feedback after answering each question, showing correct/incorrect answers.

Method 1:

Considering that all multi-drag/drop quizzes are answered one at a time, just like a multiple choice, I've created a series of multi-choice questions, all with the same 3 choices. This gives the opportunity to provide meaningful corrective feedback for each choice made. Also, with a little advanced variables and trigger manipulation, you can require user to retake all previously incorrect answered questions while skipping the correct ones, until all are correct to continue on in the course.

Method 2:

For the ability to show multiple questions, choices, and general correct/incorrect feedback, I created a type of grid of rows/columns, with each row having the question, then the other columns are radio buttons, each under a column header matching the original choice text. For example, and note the naming/numbering of radio button to check for completeness, etc.:

                                                            Acute                Chronic                Both

Medical symptom 1 text             radio_q1_c1     radio_q1_c2      radio_q1_c3

Medical symptom 2 text             radio_q2_c1     radio_q2_c2      radio_q2_c3

Medical symptom 3 text             radio_q3_c1     radio_q3_c2      radio_q3_c3

 

I hope that gives you some ideas. I may be able to dig up example files, if needed.

EdTech Account

Hi Scott.  Thanks for the suggestions!  I am coming around to something almost exactly like what you suggested in your second option.  Tinkering this morning, I discovered that you can apply button sets to the answer choices in multiple response questions, so you can make it so the user can only chose one of the category responses for any given row (symptom = row, in this case).  And, as you said, if you put a good name on each answer choice, that should work for screen readers. 

I also discovered that you can give a long name on the answer choices but use formatting to shrink elements of it for sighted users.  So, in your example, I might use "Symptom 1, Acute" as the label for the first answer choice, but make "Symptom 1" 6pt and "Acute" 12 pt with a manual line break in between.  Screen reader users would still hear the full title, but sighted users would probably just see "Acute" in the row for Symptom 1 and not notice the smaller first part of the label.  I will try to build an example in the next day or so and post it here for comment.

Your method one had also occurred to me but, as you point out, would take a lot of slides.

I wonder if I could give row-based feedback on Option 2 based on the answer state and embed it in the "incorrect" layer.... 

Again, I really appreciate your suggestions!  It helps to bounce ideas off of others.