Speech therapy software

Portfolio | UX Testing & Improvement

 

WHO Morespeech.com web app
WHAT Teach brain-injury survivor how to speak more clearly.
Goal:  survivor use program independently.
Result Success rate improved  450%

Background

I designed software that brain-injured adults can use independently for speech & language therapy.  This Case Study focuses on one particularly challenging design problem for which we measured effectiveness.

I worked with an Aphasiology researcher (Dr. Richard Katz)  to design a program based on a standard treatment for Apraxia (articulation disorder).  Based on his explanation of the approach, I provided him with design sketches. I designed the UI based on his feedback. I tested a prototype using fivesecondtest.com.

Index

  1.  User Research
  2. User Journey
  3. Epic User Story
  4. User Interaction Sketches
  5. Usability Test & Results #1
  6. Iterate on the Design (Wireframes)
  7. Live Design
  8. Usability Test Results #2
  9. Usability Test Results #3

1 . User Research

I consulted with Dr. Katz (above) as well as interviewing patients with the Apraxia disorder this software addressed, as well as their caregivers (usually the spouse.)

User Personas

Paul the Patient

Paul is 68 years old. He had a stroke which affected his articulation (speech clarity), his word-retrieval (thinking of the right word to say)  and reading comprehension, and has right-sided paralysis. He's right-handed so he now uses the computer with his left hand, which he finds quite difficult.   Before his stroke he used the computer to manage their finances, including simple spreadsheets. He’s comfortable using the computer.


Carol the Caregiver

Carol is Paul’s wife. She’s 70. She does light computer work, searching for recipes  online. She is a little under confident about learning new things on the computer.


Tina the Therapist

Tina is their speech therapist. She’s very busy and has little spare time because she has to maintain high productivity and can only bill for hours she spends with Paul. So she has no outside time for learning new things.  She is 55 years old and comfortable with the computer. But she has never downloaded or installed software. She’s also nervous about learning something new in front of her patients (looking dumb, etc.)




2 . Describe the User Journey

 

3. Epic User Story

Carol the Caregiver sees an advertisement for MoreSpeech.com, interactive speech therapy software she and Paul can use at home.  She shows it to Tina and points out that it was created by a speech therapist. Tina is a little nervous about learning something new in front of Paul & Carol. She also doesn’t use her computer for therapy. She usually uses flash cards and worksheets.  But she can see Carol is excited because she thinks Paul would enjoy this. She’s willing to try it.


Paul wants to improve his ability to speak words clearly.  He normally watches Therapist Tina show him where to place his tongue and lips to make the right sounds.  He needs to see a mouth demonstrating the word. Tina uses a process called the Rosenbek Hierarchy which gives Paul a lot of assistance and demonstration then removes that assistance. So first she shows him a photo and a word and says the word while he watches. Once he can successfully repeat it back clearly, he proceeds to the next step for that word, and gives him less assistance, making the exercise more challenging. The goal is to get to the final step where he can say the word in response to a cue. So if the word were apple the goal might be for him to say apple when prompted with “it’s as american as baseball and ___ pie”.

There are 6 additional Steps, with less and less assistance (removing the video hint, htne removing the text, etc.). We'll only show the background sketches for the first step.


4. User Interaction Sketches

Full Assistance

 

5. Usability Testing Results

At this point the previous design was implemented and live.  I do not have a copy of that previous version.
This was our beta release version. I then performed a Usability Test on the live program

 Usability Test Results

  1. 80% of Users (n=15)  confused by Next Word button.
  2. Symptom:  80% never click the Next Step button. When asked, they indicate they didn't really notice it. After more investigation, I concluded the problem was caused by the fact that most lessons in this program have one exercise per word. So after they speak the word, they click Next Word, and miss the other 6 steps in the exercise.

 

6. Iterate on the Design

Goal:  Make UI clearer so that users go through each Step they can do, and go to the Next Word when they reach their limit or finish all the Steps.

 

Iterated again with wire-frames.

 

Ideas: Simplify the button names. Focus on User-Centric language, or even better, User-Outcomes language.

 

 

 



More iteration on Interface

  1. Once we are on the second Step, show the Back button.
    It goes back one Step. If on Step 1, it goes to previous Word and sets it to Step 1.  If at Word 1, Step 1, hide it again.
  2. De-emphasize the Back button. It's rarely used. Make it a secondary choice (link instead of button)

 

 

7. Actual Live Design

Note: the software developer did not implement the icon specified for the "Too hard"

Step 1

Final Design with buttons disabled while video is playing.
(Disable buttons prevent Paul the Patient from clicking them while the video is playing)

Step 2

Step 3

 

Step 4: Repeat the word you hear

Step 5  - Read the word aloud (no audio or video help)

Step 6: Read the word and the speak it from memory.

 


Step 7 (Final Step) - Speak the word in response to a spoken cue



8. Usability Testing Results #2

78-89% of test subjects reacted successfully.

89% of users clicked the "Said it OK" button appropriately in a 5 Second Usability Test. (Recruited participants from Facebook Stroke Survivor's group).

78% of users clicked the "Too hard" button in an appropriate scenario in a 5 second usability test I peformed.

However, I discovered a new problem:
When the first lesson starts they are not always prepared to watch the video.

Solution:  Use Progressive Disclosure:  Display only the "Said it OK"  button  with text "Start"

 

First Step, updated with Progressive Disclosure.
When they click the Start button, the first Step for the first word begins.

 

9. Usability Test Result #3


79% of test subjects were able to choose the correct button.
I performed some opportunistic Usability Tests during  Sales and Coaching calls, where I would observe the user remotely using Zoom.us without providing assistance. All users ("Patient Paul") were able to use the interface correctly, sometimes needing assistance from "Carol the Caregiver".