Talker is a tablet-based Augmentative, and Alternative Communication System developed to improve the communication rate of autistic children in Reyoo Paddock, a special needs school in Accra, Ghana. I completed this project in fulfillment of my undergraduate thesis requirement.
Impairment in communication for individuals who have ASD can take the form of severe language delay and limited language production. For children with autism, Augmentative and Alternative Communication (AAC) systems have traditionally been adopted to enhance communication. However, AAC systems
Do not focus on real world contexts
Have poor interface design applications
Are not designed for varying capabilities
Hence, the research question I sought to answer was:
How might we design Augmentative and Alternative Communication Systems that focus on real-world contexts as well as the motor, cognitive, and linguistic processing of autistic children?
To obtain a deeper understanding of the problem, I reviewed work on autism spectrum disorder, interface design guidelines, and augmentative and alternative communication systems.
This literature review was approached from an exploratory point of view, allowing for the development of themes and reoccurring areas of research.
UNDERSTANDING COMMUNICATION PARTNERS
Often, children with autism spectrum disorder use AAC devices with the help and intervention of their communication partners. This could be their parents, caretakers, teachers, etc., but in this study, I collaborated with teachers who also served as their caretakers.
In this study, I would be referring to them as caretakers. Two participant groups were recruited for this study: caretakers and autistic children.
Twenty autistic children were recruited from Reyo Paddock Special School, an autistic center in Madina, Ghana, that provides therapy and assistance for children with autism. The participants were diagnosed with autism spectrum disorder, aged between 4- 18 years, and used a limited number of words to communicate frequently.
The sampling method adopted was convenience sampling. The children were grouped into 3 groups based on their capabilities. The groups were:
The secondary participants in this study were the caretakers, as they are the natural communicators of the children. 6 teachers intervened during the usability and A/B test and participated in the interviews.
Due to the nature of the project, the secondary participants were interviewed. I interviewed the participants to understand the competencies and behavioral patterns of the children with complex communication needs.
The interview allowed the teachers to rate the participant's motor skills, cognitive, linguistic, receptive, and expressive language skills. In addition, I asked open-ended questions to gain insight into the child’s familiarity with technology, past experiences with AAC systems, and its impacts on the child's communication competencies.
UNDERSTANDING THE PARTICIPANTS MEANS OF COMMUNICATION
To better understand the needs of the children in this study, I needed to know how they communicated and the systems that allowed them to interact with those around them. For my proposed solution to be effective, it needed to be natural; hence my decision to conduct observational studies the participants in their natural environment.
During the observational study, I discovered that the children used PECS (Picture Exchange Communication System), a picture-based communication method. They used this system to request items when performing three significant activities: eating, outdoor, and indoor activities.
UNDERSTANDING THE CURRENT CLIMATE OF AAC SYSTEMS
The first step in my research was to observe and understand how existing AAC devices perform and their impact on the communication of autistic children with varying competencies. Hence, I conducted usability tests on Eline Speaks, a tablet-based AAC application, and document users' pain points while interacting with the technology.
My decision to conduct a usability test on existing applications was to validate a finding I derived during the literature review, which states that AAC technologies are seldom designed from a user-centered approach.
Based on the insights from the observational study, I modified the Eline Speaks interface to mimic commonly requested items by the students to create consistency within their experience.
Snippets of the usability tests is below
Participants were having difficulties using the interface due to the presence of categories.
This placed a cognitive load on the children as the internal elements are inaccessible to them.
For participants with motor deficits, selecting the requested image proved difficult, as the spacing between each element in the interface was small.
Due to their lack of motor control, the participants often made wrong selections.
Participants with limited literary capabilities, such as the inability to identify a maximum of 8 letters, had problems identifying complex texts.
Components titled “Water Play” or “Drink Water” were not understood by most of the participants.
Some participants could not match the image to text due to the font size. The font size was small; therefore, participants with visual impairments suffered significant communication difficulties.
Hence, the children solely relied on the image to respond to the commands of their caregiver.
Talker is a tablet-based Augmentative and Alternative Communication system that focuses on real-world contexts and varying abilities of autistic children.
Drop Down Categories
To reduce the cognitive load on the participants, I eliminated categories on the landing page. Now, caretakers have access to a drop-down that allows them to select categories, removing the responsibility from the children.
*clustering principle- interfaces should be organized to separate blocks of similar control
Increased Font Size
To address the needs of participants with visual impairments, I set the font size for the text to 60px.
To satisfy the needs of participants with motor deficits, I gave adequate spacing between each element.
The aim of increasing the spacing was to reduce selection error caused by the proximity of elements in the interface.
To cater to the everyday needs of the autistic child, I created a feature that allows caretakers to create categories and related items that fits the child's real-world contexts.
Since children's literary capabilities differ, I also incorporated an edit function for each item in the interface. With this functionality, caregivers can modify complex texts and add commonly requested items.
A/B testing was then used to find the statistical difference between the Eline Speaks interface and the prototyped design to determine if the design changes enhanced the communication rate of the participants in the study.
The two statistical tools used were the T-test and the Chi-squared test.
The t-test was conducted to determine if there was any statistical difference between the communication results of the participants when they used Eline Speaks and the prototyped solution. The t-test was used to analyze the right and wrong attempts made by the participants when communicating.
The chi-squared test was conducted to identify if there was a significant difference between the time taken to communicate when using Eline Speaks and the time taken to communicate when using the prototyped design.
Similar to the usability test, the participants in the study were presented with Eline Speaks and Talker, their communication medium.Caretakers provided commands and prompted participants to select the interface elements corresponding to the prompt.
The intervention types allowed during this study were: 1. Physical Intervention 2. Gestural Prompts 3. Physical Prompts
Time, Right and Wrong Responses were recorded during the session and compared against both designs.Below are examples of the data collection tables.
The data below shows the difference in the communication rate of the participants when using Talker compared to their communication rate when using Eline Speaks. To know more about the statistical differences observed, please refer to this report detailing an extensive quantitative analysis.
The results from the quantitative analysis showed the Talker led to a :
From conducting interviews and discussing with the caretakers after the usability and A/B testing, the following were the observations I made:
From a user interface approach, spacing, font-text, layout/ grid, and images on AAC apps should be considered when developing assistive technology for autistic children due to their varying abilities. Here are the findings:
1. Font Size
Feedback from the usability test showed that larger texts allowed the participants to see the text clearly and match it to an image, allowing them to select their desired object.
Ample spacing between elements was a helpful guideline, as it reduced the possibility of errors for participants with poor motor skills and allowed them to communicate effectively.
Adequate spacing between elements significantly contributed to the success rate of the majority of the participants in the study.
One caretaker shared," [ Participant 2 ] would find this (Talker) easier because his motor impairments can make him choose the wrong thing."
3. Visual Clutter
Reducing clutter/ elements on the screen by creating categories allowed participants to identify items quickly. Participants had a higher success rate when I reduced the items on display than when there were multiple items.
Focus on user experience not design
This is already a commonly agreed-upon concept in the HCI community, but it has often come up in my research. As a designer with an eye for visually appealing work, my research constantly reminded me that a focus on the user experience created more impact than a focus on the system's aesthetic.
By focusing on functionality and interaction processes, I could identify system elements that users will find difficult to understand or navigate and improve the user's experience. In this case, I helped improve the communication experience by making it natural and more accessible.
Inclusive design is a growing field
Even though I intended to design a system that children of varying capabilities could use and improve their communication abilities, it is evident in my research that it is impossible.
There is no one size fits all solution. However, having an inclusive approach to my work allowed me to view my research and design from the experience of users typically underserved in technology. So even though my prototype only had success for some users, I was delighted with the results.