UD and ICT Archives - Page 2 of 10 -

Overcoming bias in AI

Artificial Intelligence (AI) is entering our everyday lives with increased speed and sometimes without our knowledge. But it is only as good as the data it is fed, and the worry about bias is a concern for marginalised groups. AI has the potential to enhance life for everyone, but that requires overcoming bias in AI development. In his article, Christopher Land argues for more advocacy and transparency in AI.

The power of machine learning comes from pattern recognition within vast quantities of data. Using statistics, AI reveals new patterns and associations that human developers might miss or lack the processing power to uncover.

A background of computer code with a female face overlaid. Overcoming AI bias.

Designing for the average is fraught with problems. Statistical averages do not translate to some kind of human average. That’s because statistics don’t measure human diversity. That’s why AI processes are at risk of leaving some people behind. But in gathering useful data there are some privacy issues.

AI shows great promise with robot assistants to assist people with disability and older people with everyday tasks. AI imaging and recognition tools help nonvisual users understand video and pictures.

Christopher Land outlines how AI and machine learning work and how bias is introduced into AI systems if not prevented. He also has some recommendations on strengthening legal protections for people with disability. The paper is not technical. Rather it explains clearly how it works, where it’s used, and what needs to be done.

The title of the article is, Disability Bias & New Frontiers in Artificial Intelligence. The “Black Box” issue is explained and the need for a “Glass Box” is presented.

From the abstract

Bias in artificial intelligence (AI) systems can cause discrimination against marginalized groups, including people with disabilities. This discrimination is most often unintentional and due to a lack of training and awareness of how to build inclusive systems.

This paper has two main objectives: 1) provide an overview of AI systems and machine learning, including disability bias, for accessibility professionals and related non-development roles; and 2) discuss methods for building accessible AI systems inclusively to mitigate bias.

Worldwide progress on establishing legal protection against AI bias is provided, with recommendations on strengthening laws to protect people with disabilities from discrimination by AI systems. When built accessibly, AI systems can promote fairness and enhance the lives of everyone, in unprecedented ways.

Diversity and inclusion in AI

An Australian book chapter takes a comprehensive and practical approach to how equity and inclusion should be considered throughout development. This should be done at both governance and development levels by applying inclusive design and human-centred design to the AI ‘ecosystem’.

The title of the chapter is Diversity and Inclusion in Artificial Intelligence.

Older people, ageism and digital design

Do stereotypes of older people affect how digital technology is designed? A team of researchers found that ageism has the potential to influence design in negative ways. However, co-design partnerships not only overcame the affect of ageism, it was also likely to produce technologies that are needed, wanted and used.

Ageism can have a detrimental role in how digital technologies are designed. Participating with older people in the design process has the additional benefit of countering stereotypes.

Image shows a group of older people on a desert camping expedition.

Photograph on a sand dune of 18 passengers and 4 drivers

Older people said the “ultimate partnership” in co-designing is to be involved from the beginning through to the end of the design process. Sharing control over design decisions was an important part of the process. They are more than informants – they are equals who have valuable contributions.

The researchers noted that although this vision of co-design is shared by designers, it is not always the case in practice.

Image shows older people working together on a workshop question.

Older people sit at round tables discussing questions. There are four round tables shown in this picture.

Older people in the study also said that ageism emerges in implicit and explicit language about ageing. And ageist images can influence the design process. Consequently, the researchers say it is important to view the diversity of older people.

It’s about co-design

How and when to involve older people in digital design is also important. Understanding co-design with older people has the potential for avoiding insufficient prototyping, biases and errors in the design process.

The title of the article is, An “ultimate partnership”: Older persons’ perspectives on age-stereotypes and intergenerational interaction in co-designing digital technologies.

From the abstract

There is a gap between the ideal of involving older persons throughout the design process of digital technology, and actual practice.

Twenty-one older people participated in three focus groups. Results showed ageism was experienced by participants in their daily lives and interactions with the designers during the design process. Negative images of ageing were pointed out as a potential influencing factor on design decisions. Nevertheless, positive experiences of inclusive design pointed out the importance of “partnership” in the design process.

Participants defined the “ultimate partnership” in co-designing as processes in which they were involved from the beginning, iteratively, in a participatory approach. Such processes were perceived as leading to successful design outcomes, which they would like to use, and reduced intergenerational tension.

Microsoft’s new inclusive design toolkit

Microsoft wants designers to see beyond physical and sensory disabilities. So they have updated their popular Inclusive Design Toolkit to include cognition – the brain. Cognition is about getting, storing and retrieving information. It’s also about focusing, learning, memorising and making decisions. So how to design for people who process thoughts in different ways?

Microsoft launched it’s first inclusive design toolkit in 2015, but it only focused on physical and sensory disabilities. The second edition takes a broader approach to address cognitive exclusion.

A black and white graphic of stick people in various states of being. Microsoft's Inclusive Design Toolkit

The new toolkit has three key principles for cognition, which can be applied in many other design contexts:

Understand the user’s motivation, and the goals and tasks they are trying to complete.
Discern the cognitive load required to reduce that mismatch.
Co-create the final product with a diverse community of people across the spectrum.

A man wearing a black t-shirt holds his hand to his forehead in an act of desperation.

The toolkit is not about specific industries or specific conditions. Rather it encourages designers to collaborate with users and find out first hand how they learn and think. The Inclusive Design 101 Guidebook has the basics. The Inclusive Design Cognitive Exclusion is a separate document.

The toolkit and guides are useful for anyone who wants to learn how to design inclusively – to take a universal design approach to design.

FastCompany has an article about the Inclusive Design Toolkit’s development. Christina Mallon, Microsoft’s head of inclusive design, discloses that she has ADHD. She couldn’t complete certain tasks and felt stupid. When she learned about inclusive design she realised that she was not stupid, just designed out of products. Now she just wants her job to be just a designer, not an inclusive designer. The title of the article is, Microsoft’s new Inclusive Design Toolkit is designed for the brain.

Computing students show age and gender bias

As we know, bias sits in all of us whether we realise it or not. Some biases we are aware of, such as liking certain people and things. But what about those biases we are not aware of that sit silently in the background of our thoughts and ideas? And how much do they impact on the things we say and do? Age and gender bias has an impact on how things are designed.

Older people are thought less likely to use desktops, laptops and smartphones, and to have less expertise with them. Women, both young and old, are thought to have some experience with devices, but less expertise than men.

A group of researchers in Austria wanted to find out the perceptions of computing science students about age and gender. That’s because they are going to be designing the digital technology in the future. In a nutshell, they found several biases.

Age and gender bias

Students (aged around 21 years) started to see people as old at the average age of 57 years – several years younger than their grandparents. Older people are thought to be less likely to have experience with all types of devices. While younger people are thought to want aesthetic designs, older people are thought to need error tolerant systems.

The bias between the genders was smaller than that of age. Women were seen as less likely to use a desktop than men and to have less expertise than men overall. This fits the continuing stereotype that computing for men is ordinary, and exceptional for women. Consequently, older women were seen as less capable in using computers and laptops.

The article has a lot of statistical analysis, but the key points are in the findings, discussion and conclusions. The information is useful for teachers, and the authors recommend designing with users as a way to overcome bias. And, of course, more women should be encouraged into the computing sciences.

The title of the study is, How Age and Gender Affect the Opinions of Computing Students Regarding Computer Usage and Design Needs.

From the abstract

This study aimed to understand the perceptions of young computing science students about women and older people about computer literacy and how this may affect the design of computer-based systems. Based on photos, participants were asked how likely the person would be to use desktop computers, laptops and smartphones. And also, how much expertise they thought they would have with each technology. We asked what design aspects would be important and whether they thought an adapted technology would be required.

The results draw on 200 questionnaires from students in the first year of their Information and Communications Technology (ICT) studies at an Austrian university of applied sciences. Quantitative methods were used to determine if perceptions varied significantly based on the age and gender of the people depicted.

Qualitative analysis was used to evaluate the design aspects mentioned. The results show that there are biases against both older people and women with respect to their perceived expertise with computers. This is also reflected in the design aspects thought to be important for the different cohorts. This is crucial as future systems will be designed by the participants. Their biases may influence whether future systems meet the needs and wishes of all groups or increase the digital divide.

Gender and Mobile Apps

An academic paper, A Study of Gender Discussions in Mobile Apps, provides some insights into gender bias in app development.

From the abstract

Mobile software apps are one of the digital technologies that our modern life heavily depends on. A key issue in the development of apps is how to design gender-inclusive apps. Apps that do not consider gender inclusion, diversity, and equality in their design can create barriers for their diverse users.

There have been some efforts to develop gender-inclusive apps, but a lack of understanding of user perspectives on gender may prevent app developers and owners from identifying issues related to gender and proposing solutions for improvement.

Users express many different opinions about apps in their reviews, from sharing their experiences, and reporting bugs, to requesting new features. In this study, we aim at unpacking gender discussions about apps from the user perspective by analysing app reviews.

We first develop and evaluate several Machine Learning (ML) and Deep Learning (DL) classifiers that automatically detect gender reviews. We apply our ML and DL classifiers on a manually constructed dataset of 1,440 app reviews from the Google App Store, composing 620 gender reviews and 820 non-gender reviews.

Our best classifier achieves an F1-score of 90.77%. Second, our qualitative analysis of a randomly selected 388 out of 620 gender reviews shows that gender discussions in app reviews revolve around six topics: App Features, Appearance, Content, Company Policy and Censorship, Advertisement, and Community. Finally, we provide some practical implications and recommendations for developing gender-inclusive apps.

Good colour contrast for websites

How did you choose the colours for your last website update? Did you choose colours based on your brand logo and text or did you use the Web Content Accessibility Guidelines (WCAG) algorithm? But can the WCAG algorithm guarantee good legible colour contrasts for websites? Research by The University of Cambridge says it doesn’t. So they have developed an alternative algorithm for good colour contrast for websites.

Since January 2022, the Accessible Perceptual Contrast Algorithm proposes that legibility of text on websites is better with perceived difference than a mathematical contrast ratio. White text on strong coloured backgrounds are preferred over black text in almost all cases in the study.

In the examples above, the black text passes the WCAG contrast ratio but fails the white text. The Accessible Perceptual Contrast Algorithm passes the white text and fails the black text.

Sam Waller explains this more fully in his article, Does the contrast ratio actually predict the legibility of website text?

As a result of this work, an early working draft of WCAG 3 proposes using the new method for calculating contrast.

This is important information for choosing brand logos and text so it isn’t just something web designers should know. Many website designs are guided by brand colours so choose carefully. This information is also important for product labelling especially for online shopping.

Do web accessibility overlays work?

Accessibility Overlays - What are they & why are they so bad? - YouTube — First screen of the YouTube video

This site has a web accessibility overlay or add-in widget. It’s the circle icon next to our logo on the website. If you click on it, it has a dropdown accessibility toolbar. That’s because the platform, WordPress, isn’t inherently accessible. So like the tacked on ramp to a building, it is an afterthought. But really, it advertises that the website platform isn’t really accessible and there are good reasons why.

Website add-ons for accessibility go back to the 1990s with products like Browsealoud and Readspeaker. They added text to speech capabilities on the website. More products arrived in the market with similar aims. To the layperson these features seem beneficial, but their practical value is overstated. That’s because the people who need these features will already have the software on their devices to access the web and other software. The Overlay Fact Sheet by Karl Groves explains more:

From the overlay fact sheet

“It is a mistake to believe that the features provided by the overlay widget will be of much use by end users because if those features were necessary to use the website, they’d be needed for all websites that the user interacts with. Instead, the widget is —at best—redundant functionality with what the user already has.”

Do overlays meet compliance?

While an overlay might improve compliance in some respects, full compliance cannot be achieved using this method. That’s because the products are unable to “repair” all possible issues. In some cases, the overlay can conflict with the users software and cause problems. And ironically, some overlays are inaccessible. So that means it’s back to the programmer and designer to get it right.

The video below gives examples of overlays and graphically shows how they don’t work. You only need to look at the first three minutes to get the idea.

We all have a responsibility to make our digital information accessible. Beware any web developer who says they’ve solved the accessibility problem with an overlay or widget. Indeed, you are showing your inaccessibility by having an “accessibility” overlay and icon on your site.

Web designers might think the international web standards are sufficient. But they are not – just like the standards for access and mobility in the public domain are not enough.

By the way, CUDA uses the WordPress platform’s free version and continues to do so because we do not receive financial support for the website and want to keep it open access. As with everything universal design – it is a work in progress. “Do the best you can with what you have at the time and strive to improve next time.”

Inclusive online conference planning

Virtual and hybrid conferences have become more popular since the advent of the recent pandemic. But are they accessible and inclusive? A paper from Canada addresses the issues of inclusive online conferences. Using the recent Parks Accessibility Conference as a case study, the authors describe their experiences. As a Canadian event, they also had to consider two languages in their planning.

Some people with disability or impairment find online events less stressful than attending in person. They can avoid travel stresses and the regular access barriers. Others who find crowds and noise difficult, tuning in from home is a more comfortable. Consequently, conference planners need to take care to plan for easy access and inclusion.

And it should be for every conference, not just conferences with a disability component. However, this is a good place to begin and to learn from first hand experiences. The Parks Accessibility Conference is one such example.

Key strategies

The authors provide a list of their key strategies:

- Make visual elements accessible to attendees with vision impairment
- Make audio elements accessible to attendees who are Deaf or hard of hearing
- Avoid overstimulation for individual who are neurodiverse or with a cognitive condition
- Create ways to incorporate multi-sensory experiences remotely
- Finding the right virtual conference platform.

The planners worked with presenters to help format and organise their presentations and materials. They hosted a pre-conference session with attendees to explain how to use the various features of the online platform so they knew what to expect.

The paper reads like a story, explaining every step along the way so that others might learn from their experience. There are eight recommendations for future conferences based on what they learned. Top of the list is to include people with disability from the beginning.

The title of the paper is, Virtual Accessible Bilingual Conference Planning: The Parks Accessibility Conference.

From the Abstract

The objective of this paper is to share how our team planned and delivered a virtual conference that was fully bilingual and accessible to individuals with disabilities.

We incorporated closed captions, sign language interpretation, language interpretation (audio), regularly scheduled breaks, and a multi-sensory experience.

We describe our approaches to planning the conference, such as including individuals with disabilities in decision-making, selecting virtual conference platforms, captioners, and interpreters, and how we incorporated a multi-sensory experience.

The paper also summarizes feedback we received from our attendees using a post-conference evaluation survey and our team’s reflections on positive aspects of the conference and opportunities for improvement. We conclude by providing a set of practical recommendations that we feel may be helpful to others planning virtual accessible bilingual conferences in the future.

People with complex communication needs

See a separate paper, Supporting Communication Accessibility and Inclusion in Online Meetings for Persons with Complex Communication Access Needs. This is a Masters thesis on running online sessions for people with complex communication needs.

From the abstract

This community-based, participatory research study explores how online sessions can be designed to support complex communication access needs. The use of a community-led co-design approach resulted in a deeper understanding of the individual communication accessibility requirements, barriers, and lived experiences of persons who use AAC, within the online meeting context.

Participants (‘co-designers’) designed and took part in collaborative design sessions aimed at developing ideas for supporting communication access and inclusion throughout the process of meeting online. Through cross-community collaboration, we co-designed an open-source communication access toolkit for online meetings.

The toolkit includes accessibility guidelines with a protocol for holding accessible and inclusive online sessions; suggested accessibility features and plugins for meeting platforms; and a template for a collaborative participant notebook.

The design outcomes provide guidance to the general population on how we might ensure that online meetings of all forms are inclusive and accessible for diverse and complex communicators, as we all have a right to communicate with dignity in ways where we understand and are understood.

Making questionnaires more readable

One area of inclusion and accessibility that often gets forgotten is readability of forms and questionnaires. Academics and marketing professionals regularly use surveys to get information from specific groups of people. Within those groups will be people with varying levels of capability in terms of being able to decipher what’s on the screen or form. And it isn’t all about literacy and reading ability. It’s about the different ways people see and interpret the information. Here are some good tips for making questionnaires more readable from Alex Haagaard in Medium.

Likert Scales

Likert scales aren’t great for screen readers because they often interpret them as tables. But much depends on the design of the survey platform. Even if they are screen-readable, Likert scales can be difficult for people who are neurodiverse. People who are autistic or dyslexic struggle with visual tracking across and between rows. This creates the need to exert more brain power to focus on getting the corresponding check box.

Instead of using a Likert scale, use a series multiple choice questions to capture the same information. Creating page breaks to separate distinct sections of the questionnaire also helps with readability for everyone.

Balancing access conflicts

As is often the case, making something more accessible for one group can create problems for another. So it’s important to identify these early and eliminate or mitigate the barriers.

One solution is to provide optional comment boxes where the participant can choose whether to reply in their own words. People who want to quickly complete the questionnaire can skip this.

Haagaard takes things a step further with a suggestion to provide detailed explanations about terms and concepts at the beginning of each section. However, this is tiresome for screen readers and others might find this overwhelming. Participants can be asked at the beginning of the survey if they would like the key information repeated for each section. Those who say no can have the concise experience.

In summary, Haagaard acknowledges that it is unrealistic to assume that anything can be fully accessible to everyone. That means that there will still be occasions where an alternative means of participating is required. This might be an interview or an email.

The title of the article is Making Your Surveys More Readable. This is the third in a series on cognitively accessible survey design.

Accessible places with Google Maps

Google Maps has two new features to help us find our way and what we are looking for. Live View uses the camera in your phone to give street view directions, and the “Accessible Places” feature marks entrances with wheelchair access.

Accessible Places feature

Google Maps has expanded its “Accessible Places” feature that shows when a place is wheelchair accessible and/or stair-free. It will be interesting to see if this will encourage more places to make their businesses accessible. As we know, when it’s good for a wheelchair, it’s good for prams, bikes and trolleys.

The feature was originally launched in 2020, but it was limited to just Australia, Japan, the U.K., and the U.S. Now, the feature is available worldwide on Android and iOS.

To use this feature, turn on the “Accessible Places” setting in the Google Maps app. You will see a wheelchair icon on the business profile if it has a wheelchair accessible entrance. You will see the same icon with a strikethrough if the location is not accessible. The feature can also check to see if the location offers accessible parking, facilities, and seating.

Steepness of slopes doesn’t appear to be covered, and Google doesn’t say if their access maps are accessible.

It’s up to business owners to update their business profile to reflect whether their business is accessible. It’s unlikely Google will check whether this is true, but user feedback should keep businesses on their toes.

Live View feature

When walking around in new surroundings, Live View helps you keep your bearings. Using the phone camera, the flat map view is transformed into the street view with arrows so you know which way to head. A great plus for tourists and visitors.

With Live View, users can walk the streets using the camera on their phone and get directions on the screen. The updated version has more information such as cafes, ATMs and transit stations.

A Gif of Live View in action showing a street view on Google Maps.

Different app developers have tried their hand at creating access maps, but Google brings many features together on the one screen. However, access for wheelchair users does not guarantee access for everyone, and it doesn’t cover all disabilities. It also doesn’t say how welcome people with disability will feel. Nevertheless, it is a good start and Google will continue to improve. The next thing is finding places where you can hear each other talk.

The Metaverse: inclusive and accessible?

The concept of the Metaverse is a continuous online 3D universe that combines multiple virtual spaces. It’s the next step on from the internet. It means users can work, meet, game and socialize in these 3D spaces. We are not quite there yet, but some platforms have metaverse-like elements. Video games and Virtual Reality are two examples. So, we need to keep a careful watch on developments to make sure the Metaverse is inclusive and accessible.

Another term for the Metaverse is digital immersive environments. It sounds science fiction, but this fiction is becoming a fact. Someone is designing these environments, but are they considering equity, diversity and inclusion? Zallio and Clarkson decided to tackle this issue and did some research on where the industry is heading.

Several companies are involved in the development of digital immersive environments. So before they get too far in development it’s important to define some principles for the design of a good Metaverse. Zallio and Clarkson came up with ten principles that embrace inclusion, diversity, equity, accessibility and safety.

10 Principles for designing a good Metaverse

is open and accessible
is honest and understandable
is safe and secure
is driven by social equity and inclusion
is sustainable
values privacy, ethics and integrity
guarantees data protection and ownership
empowers diversity through self-expression
innovates responsibly
complements the physical world

A young woman is wearing a pair of virtual reality goggles and looking towards the sky.

Their paper is insightful and provides some important areas for discussion and research. We need developers to consider the essentials of inclusion, diversity and accessibility. Zallio and Clarkson advise that designers can learn from the past to reduce pitfalls in the future. As the Sustainable Development Goals say, “leave no-one behind”.

Diagram showing the 10 principles for designing a good Metaverse.

The title of the paper is Designing the Metaverse: A study on Inclusion, Diversity, Equity, Accessibility and Safety for digital immersive environments.

Synopsis of the paper

1. The Metaverse appears as the next big opportunity in the consumer electronics scenario.

2. Several companies are involved with its development.

3. It is extremely important to define principles and practices to design a good Metaverse.

4. Qualitative research pointed out to challenges and opportunities to design a safe, inclusive, accessible Metaverse that guarantees equity and diversity.

5. Ten principles for designing a good Metaverse embrace inclusion, diversity, equity, accessibility and safety.

From the abstract

The Metaverse is shaping a new way for people interact and socialise. By 2026 a quarter of the population will spend at least an hour a day in the Metaverse. This requires consideration of challenges and opportunities that will influence the design of the Metaverse.

A study was carried out with industry experts to explore the social impact of the Metaverse through the lens of Inclusion, Diversity, Equity, Accessibility and Safety (IDEAS). The goal was to identify directions business has to undertake.

The results indicated the nature of future research questions and analysis to define a first manifesto for Inclusion, Diversity, Equity, Accessibility and Safety in the Metaverse.

This manifesto is a starting point to develop a narrative, brainstorm questions and eventually provide answers for designing a Metaverse as a place for people that does not substitute the physical world but complements it.

From the abstract

Diversity and inclusion in AI

Share this:

Like this:

It’s about co-design

From the abstract

Share this:

Like this:

Share this:

Like this:

Age and gender bias

From the abstract

Gender and Mobile Apps

From the abstract

Share this:

Like this:

Share this:

Like this:

Do overlays meet compliance?

Share this:

Like this:

Key strategies

From the Abstract

People with complex communication needs

From the abstract

Share this:

Like this:

Likert Scales

Balancing access conflicts

Share this:

Like this:

Accessible Places feature

Live View feature

Share this:

Like this:

10 Principles for designing a good Metaverse

Synopsis of the paper

From the abstract

Share this:

Like this: