Using an Accessibility Maturity Model to Facilitate the Inclusion of Accessibility in Design Practices

(1)

Linköping University | Department IDA Master Thesis, 120 credits | Design Spring 2020 | ISRN: LIU-IDA/LITH-EX-A--20/044--SE

Using an Accessibility Maturity Model to

Facilitate the Inclusion of Accessibility in

Design Practices

Linnea Michel

Supervisor: Jonas Löwgren Examiner: Stefan Holmlid

Linköping University SE-581 83 Linköping, Sweden +46 013 28 10 00, www.liu.se

(2)

Copyright

The publishers will keep this document online on the Internet – or its possible replacement – for a period of 25 years starting from the date of publication barring exceptional

circumstances. The online availability of the document implies permanent permission for anyone to read, to download, or to print out single copies for their own use and to use it unchanged for non-commercial research and educational purpose. Subsequent transfers of copyright cannot revoke this permission. All other uses of the document are conditional upon the consent of the copyright owner. The publisher has taken technical and administrative measures to assure authenticity, security and accessibility. According to intellectual property law the author has the right to be mentioned when her work is accessed as described above and to be protected against infringement. For additional information about the Linköping University Electronic Press and its procedures for publication and for assurance of document integrity, please refer to its www home page: http://www.ep.liu.se/.

(3)

Abstract

In an increasingly digital world, the accessibility of digital services is an important question of inclusion. Despite judicial requirements for accessibility of digital public services, the

inclusion of accessibility into design practices is a new research area.

This research project explores how to facilitate the inclusion of accessibility into design practices by creating a methodology that centres around an accessibility maturity assessment. The result of the assessment forms the basis for the selection of design methods that are mapped to the attributes of the model.

Preliminary research consists of an interview study, desk research and unorganized research activities. A case study is conducted at the Swedish Tax Agency, the result of which form the basis of the design intervention. Case study data is analysed with the help of an accessibility maturity model, which also serves as the core of the design intervention. For the intervention, adapted, proven design methods are connected to the attributes of the accessibility maturity model. A proof of concept is created in the form of a handbook and validated with designers and employees at the Tax Agency. The validation shows that there is promise in the

proposed methodology, as the use of the design practices ensure that accessibility is a part of the entire design process. Further research should investigate the use of the handbook in a case study in order to consolidate the practicalities of the methodology, as well as validate the utility of the methodology in a practical application.

Keywords: digital accessibility, digital public service, maturity model, accessibility maturity assessment, UX practices, universal design, disability

(4)

Acknowledgements

I would like to thank:

Cajsa-Tora Hermansson at Idean for believing in me, opening doors to interesting people to interview, listening and providing feedback.

Jonas Löwgren, for providing invaluable guidance and support throughout the thesis.

Every person at Idean and the Tax Agency who invited me to different activities that seemed relevant to my thesis. The workshops, inspiration sessions, brown bags and so on brightened my thesis-writing, gave me valuable insights and put my work into a context.

All persons I interviewed, discussed my work with and pressed for feedback.

My mom for enthusiastically proof-reading the handbook and finding mistakes nobody else did!

Norrköping in May 2020 Linnea Michel

(5)

1. Introduction

This chapter provides an introduction to the thesis. First, the background frames the thesis and lays out the motivation for the work. The research question is presented along with delimitations and an outline of the external contacts.

1.1 Background, motivation

The Swedish government aims to become a world-leader in terms of digitalization (Internetstiftelsen 2019b) and the digitalization of public services is an ongoing process (Myndigheten för digital förvaltning 2020a). Sweden has a higher rate of users of digital technologies than most other OECD countries (OECD 2018). Even so, 12 percent of Swedish citizens feel digitally excluded (Internetstiftelsen 2019a). As public services are being digitized with the policy of “Digital First” (OECD 2018, p.14), there is a risk for people to be excluded from accessing these services. Stefan Johansson, from the association

Begripsam, says in an interview with Swedish Public Television: “It is important to think about the fact that there are people who think that online services are difficult and avoid using them. There is a big risk that the digital divide will increase if we don’t include these people when the services are being developed.” (Mårtensson, R 2018). “Internet, smartphones, computers, and tablets can become obstacles and they can become something that makes participation possible” (Johansson, S 2018).

The Swedish government and parliament pass legislation in order to improve accessibility to digital public services (Swedish Institute, 2019). Since the 1st of January 2015, inaccessibility is regarded as a form of discrimination in the discrimination act (Myndigheten för delaktighet 2017). In 2016, the European Directive 2016/2102 about web accessibility was enforced and laid the ground for implementation decisions and national legislation. The legislation does not apply to the private sector, but to public sector bodies (Official Journal of the European Union 2018). In Sweden, government agencies play an important role in ensuring accessibility, as they have the national responsibility for accessibility in specific sectors (Swedish Institute, 2019). According to a study that made an accessibility assessment of 37 web services in 7 different European countries, including Sweden, none of the public sector websites fulfilled all requirements of the Web Content Accessibility Guidelines (WCAG) 2.0, to the regulated level of AA (Laurin et al 2014).

Compliance with web standards can be seen as a base-level effort to attain accessibility. A website might be compliant with the standard, but not be accessible and a website that has multiple validation errors might work well in terms of accessibility (Funka 2019a). The most common reasons cited from persons with disabilities for having difficulties accessing the Internet are: searching information, navigating, using services, understanding content, using passwords, the design is a disturbance, does not work with assistive technology (Johansson, S 2018). Most of these issues cannot be fully addressed by web standards. A study that examined the perception of web managers on accessibility saw that “[t]he focus of accessibility efforts [...] seems to be on technically complying with the standards or guidelines referenced in relevant policies and legislation at a specific point in the development of a website rather than optimizing and continually improving the user experience as a whole.” (Laurin et al 2014, p.14).

(8)

1.2 Research question

RQ: How can government agencies facilitate the inclusion of accessibility in the design processes of digital public services?

1.3 Delimitations

The scope of the thesis concerns digital accessibility and excludes other aspects of accessibility, such as improving the accessibility of physical spaces. The planned intervention aims at providing solutions for the public sector, not other types of organizations.

1.4 External contacts

I am writing this thesis in affiliation with Idean Stockholm, which is a design studio that is part of Capgemini Invent. My main contacts at Idean are Therese Haman (Head of Studio) and Cajsa-Tora Hermansson (Senior UX Designer). The contact with Idean has been helpful in establishing contacts with government agencies and other organizations and companies that have been part of this thesis.

(9)

2. Theoretical Framework

This chapter provides an overview of the theories upon which the thesis is based. It starts off with different ways of understanding disabilities and describes the importance of digital participation, especially in the context of digital public services. The requirements of digital accessibility in the public sector are outlined. Two different ways of working towards

accessibility are described, namely conformance to the Web Content Accessibility Guidelines and the usage of Universal Design. Lastly, the topic of maturity models is introduced.

2.1 Perspectives on disability

In the 19th and 20th centuries, the medical perspective on disability was dominant in the academic discourse about disability. The main approach to persons with disabilities was to normalize what was different, for example by making pedagogic efforts to teach deaf persons how to speak and providing prosthetics regardless of how useful the affected person found them to be (Funktionsrätt Sverige 2017a). The “problem” was seen to lie with the person with the disability and therefore also the responsibility of that individual. The individual was

expected to adapt to society, not the other way around (Chöler, Erdtman & Linder 2018). In the 1970s, there was a breakthrough for the environmentally-relative view of disability. This perspective sees disability as context-dependent rather than as a personal attribute. “Put simply, disability arises in relation to the demands made by the surrounding environment on a person and their functions. It also makes disability a political issue, because political decisions on the structure of society can prevent functional impairment from becoming a disability” (Socialstyrelsen 2006, p.1). This environmentally-relative view of disability serves as the basis for the UN Convention on the Rights of Persons with Disabilities (Chöler,

Erdtman & Linder 2018). The International Classification of Functioning, Disability and Health (ICF), is a framework released by the World Health Organization (2002) that provides a shared language for the description of health. The ICF is based on the idea that all persons can be faced with decreased health, which is equated to experiencing some level of

disability. The ICF is based on the biopsychosocial model, which integrates the medical and the social models. According to the World Health Organization, neither the medical nor the social model is adequate, although “partially valid”. In the biopsychosocial model, disability is seen as a complex notion that is an outcome of the interaction of health conditions and contextual factors (World Health Organization 2002), as shown in figure 1.

(10)

Figure 1. Biopsychosocial model of disability. Graph from the World Health Organization (2002).

In the ICF, disability is seen as a spectrum, an ability continuum, upon which all persons can be placed. In the 1970’s, the Scandinavian Relational Model emerged, which drew on the concept of the Swedish welfare state. This model sees disability as a mismatch between person and environment, as well as something that is situational, contextual and relative (Johansson 2019). In some contexts, it can be important to be able to draw a line between persons with disabilities and persons without, such as in judicial contexts. The minority model of disability proposes a delimitation of disability, as they are of the opinion that persons with disabilities are a special group with rights similar to ethnic minorities (Chöler, Erdtman & Linder 2018). This disability centric approach is useful for people who have been classified as disabled, as it gives them legal recognition. For other groups of people, the disability centric perspective can be harmful, as it overlooks individuals with minor ability losses. These minor ability losses might not be labelled as disabilities, but nevertheless impact the interaction of the user with different products or interfaces. With a narrow view of what disabilities are and what is included in the term accessibility, these individuals are not taken into account in the development of new digital services and products (Hosking, Waller, Clarkson 2010).

In the field of Human-Computer Interaction (HCI), the term disability is used more liberally than the minority model of disability. Situationally induced impairments and disability (SIID), is a term used within HCI to describe difficulties that can arise in the context of use (Sears, Jacko and Xiao 2003) (Sears and Young 2003). This context describes the interplay between the user, the application and the surrounding environment (Sears, Jacko and Xiao 2003), (Dey, Abowd and Salber 2001). “When disability is conceptualized as limits on ability, then a notion of ‘situational disabilities’ is meaningful, because situations of many kinds clearly limit the expression of our abilities.” (Wobbrock 2019, p. 59). Situationally induced impairments and disability have become more important due to the rise of ubiquitous computing. As mobile devices have become more prevalent, human-computer interaction has entered more

(11)

contexts than before (Abdolrahmani, Kuber & Hurst 2016). Contextual factors that can lead to SIIDs can be categorized as follows: behavioural, environmental, attentional, affective, social and technological. Behavioural factors include walking, driving, navigating obstacles and encumbrance, which could mean carrying something heavy for example. Environmental factors comprise cold temperatures, ambient noise, dim/bright lights and rain. Attentional factors encompass distractions, diverted gaze, interruptions, multitasking, information overload and high cognitive workload. Affective contextual factors include stress, anxiety, fear, fatigue and haste. Social factors are concerned with distractions that stem from other people, such as holding conversations, being in a crowd and adhering to social norms and expectations. Finally, technological factors include small output displays, for example, small text in an application, small input areas, for example, small buttons, lack of power or

connectivity (Wobbrock 2019, Elton and Nicolle 2019). Other papers highlight factors such as mental state and mood (Sarsenbayeva et al. 2017), as well as glare (Persad & Langdon 2007), wet hands and vibrations (Persad & Waller 2007). These factors can have an impact on target acquisition, text entry, text readability (Wobbrock 2019). Research suggests that “failure to take account of the effects everyday contexts have on capability could result in the target population becoming unintentionally excluded, especially those already working to the limits of their ability” (Elton and Nicolle 2019, p.9).

Wobbrock (2019) concedes that the concept of SIIDs can be seen as controversial or even offensive, as it can be seen to diminish the experiences of persons with disabilities. The author argues that SIIDs provide a useful perspective nonetheless, as the concept visualizes that everyone is limited in their abilities sometimes, which means that accessibility is not something that is done for other people, but that it serves all users. Wobbrock (2019) states that the perspective of SIIDs adds something valuable to design thinking because it focuses the design process to find solutions that benefit multiple people, regardless of whether their limitation of ability is situational or not. This approach, of not necessarily distinguishing between different types of limitations of ability, is common in the field of interaction design. Microsoft’s Toolkit for Inclusive Design (2016) is one example where parallels are drawn between permanent, temporary or situational limitations. The toolkit illustrates the example by showing that a person with one arm, a person with a broken arm and a person who is holding a baby are facing similar challenges in the interaction with a digital interface (Microsoft 2016). This approach broadens the target audience for inclusive design.

(12)

Figure 2. Different types of limitations. Illustration based on a graph from Microsoft (2016).

2.2 Public Digital Services and Digital participation

According to Internetstiftelsen, all people in Sweden shall have the same right and access to the Internet (2020). In European digitalization policies, Internet usage is described as a basic precondition in order to be able to be able to fully participate in society (Helsper & van

Deursen 2015). Research has shown that citizen’s trust in the government and government bodies depends on the reliability of the public services available. In order to uphold the welfare state, public services need to be reliably made accessible to all citizens (Iacobaeus et al 2019). Sweden is ahead of the curve in terms of technological development (European Commission 2019b), but there are still many persons that are digitally excluded. As more government services become digitized, these individuals run a risk of not being able to access important functions (Internetstiftelsen 2020, Iacobaeus et al 2019). 88 per cent of the Swedish population feel that they are taking part in the digital society (Internetstiftelsen 2019a), which means that there are over a million Swedish citizens that feel excluded

(Statistiska Centralbyrån 2019). Men and women with disabilities use the Internet to a lesser extent than the rest of the population. 77 percent of men and 70 percent of women with disabilities use the Internet. (Myndigheten för delaktighet 2018). In comparison, 95 per cent of the Swedish population overall uses the Internet. (Internetstiftelsen 2019c).

Most Swedish citizens think that digital public services have led to a positive change, making it easier to take part in public services. 85 percent of Swedish citizens above the age of 16 think that digital public service has helped them. But there is also a group of people who think that digital services have made it more difficult to access public service

(Internetstiftelsen 2019b). Multiple groups of persons with disabilities express that

digitalization creates good conditions for autonomy and participation. The majority of persons with disabilities report that their participation in society has increased as a consequence of the digitalization. This is due to the increased possibility to take part of information and services in multiple and individualised ways. Nevertheless, persons who are in need of good accessibility in terms of technology, such as persons with compromised vision, cognitive difficulties and persons with difficulties to write and read, are faced with barriers. The fast pace of digitalization and technological changes create difficulties, as shortcomings in the

(13)

technical development run the risk of excluding people who previously have been able to use the technology. In order to be able to harness the positives of the technology, it is important that the development of digital products and services is guided by the concept of universal design. If accessibility fails to be considered in the fast-paced development of society, there is a risk that new barriers are created continuously (Myndigheten för delaktighet 2018). E-government has two main goals, to increase efficiency of the agency and offer benefits to citizens. In e-service development, it is beneficial to visualize differences in stakeholder groups, including differences between different groups of citizens. A more multi-faceted view of the needs and motivations of different groups will make it more likely that the goals of agency efficiency and citizens satisfaction can be met (Axelsson, Melin and Lindgren 2013). Lindgren et al. write “When digital public services mediate the interaction between citizens and public organizations, and the IT artefact plays the role of the public official, the provision and design of these systems becomes particularly important.” (2019, p. 432). In order to guarantee equal opportunities for access to digital public services, some groups of citizens require specific types of support. It is important to design digital public services with an understanding of the abilities and needs of citizens because the services need to be adapted to these needs in order to be more widely accessible (Iacobaeus et al 2019). According to Sæbø et al. (2011) the development of digital public services is often characterized by a techno-centric approach that does not satisfactorily involve citizens. Involving citizens with disabilities poses specific difficulties, as they are often excluded from census studies. The group is also often excluded from the target audience in the design processes that are conducted in the development of digital public services (Johansson 2016).

2.3 Requirements of accessibility in the public sector

In 2001, the directive 2001:526 about the implementation of disability politics was passed by the Swedish Riksdag. It ordains that government agencies should conduct their operations in a way that honours the goals of disability politics. Specific focus should be on making

facilities and information accessible (Sveriges Riksdag 2001). The European Directive

2016/2102, states a political agenda for the “accessibility of websites and mobile applications of public sector bodies'', and has been in force since December 2016. The directive states that “Member States shall ensure that public sector bodies take the necessary measures to make their websites and mobile applications more accessible by making them perceivable, operable, understandable and robust.” (Official Journal of the European Union 2016). The directive also states that all Member States of the European Union should have a body that is designated to monitor compliance with the directive (Official Journal of the European Union 2016). The body that is in charge of monitoring, reporting and enforcing this directive in Sweden is the Agency for Digital Government, called “Myndigheten för digital förvaltning” (also called DIGG) in Swedish (Myndigheten för digital förvaltning 2020b). As a way of implementing the aforementioned European Directive 2016/2102, the Commission Implementing Decision 2018/1523 was enforced. It details the requirements of an

accessibility statement to be released for every website and mobile application of a public sector’s website (Official Journal of the European Union 2018). In 2018, the European Commission released a harmonized European Standard outlining accessibility requirements for ICT products and services. This harmonized standard is in line with the latest version of the Web Content Accessibility Guidelines, called WCAG 2.1 (European Commission 2019a).

(14)

According to a report that monitored the implementation of disability politics in the Swedish public sector as of 2018, the majority of government agencies include accessibility as a perspective in at least one of the high-level regulatory documents. Half of all agencies had created an action plan with steps to take in order to reach accessibility, but only every fifth agency used goals that were measurable. This is an indication that there is an awareness and an ambition to implement disability politics, but as few agencies have measurable goals, it is also a sign of a lack of specific knowledge and concrete ambitions. The result of the report shows that there is a need for support in understanding how the disability perspective applies to the operations of the agencies and how it can be implemented practically

(Myndigheten för delaktighet 2018). Active involvement of persons with disabilities is an obligation according to the United Nation’s Convention About Rights for Persons with Disabilities, (Myndigheten för delaktighet 2018) which was adopted by the Swedish Riksdag in 2008 (Regeringskansliet 2015). According to the Convention, active involvement is central in order to secure a relevant analysis of problems and needs, as well as quality assurance of interventions that aim to strengthen the rights of persons with disabilities. The involvement should be structured and enable transparency, dialogue and co-creation. Even though ensuring active involvement is part of the public task, only half of all government agencies cooperate with relevant organisations (Myndigheten för delaktighet 2018).

2.4 Judicial action against a government agency

There has only been one judicial case about accessibility, involving a public sector body. This notable case is happened in 2016, when the Equality Ombudsman

(Diskrimineringsombudsmannen), took legal action against the Swedish Social Insurance Agency (Försäkringskassan). The Equality Ombudsman had received complaint from a blind woman who could not access the application form for parental benefit online, as the website did not support the use of her screen reader. (Chöler, Erdtman & Linder 2018, p.159). From 2007 to 2014 the website was compatible with the usage of a screen reader, but after big changes in the digital services in 2014, the service of applying for parental benefit was not accessible with a screen reader anymore. Even though public sector bodies are required to comply with the standard of WCAG 2.0, level AA, this was not taken into account in the development of the new digital services of the insurance agency (Dackmar, J 2018). In the years 2015 and 2016, the woman had been in contact with the insurance agency about this problem, receiving the answer that she could gain access to the service through a sighted person, or by visiting one of the service offices. The Equality Ombudsman argues that these alternatives weren’t adequate and that the Social Insurance Agency did not take the

appropriate steps after they received the complaint. The district court ruled that the Swedish Social Insurance Agency had subjected the woman to discrimination.

(Diskrimineringsombudsmannen, 2019).

2.5 The Web Content Accessibility Guidelines

The Web Accessibility Initiative (WAI) is part of the World Wide Web Consortium (W3C), an organization that aims to “lead the World Wide Web to its full potential by developing common protocols that promote its evolution and ensure its interoperability” (World Wide Web Consortium 2019). WAI developed the Web Content Accessibility Guidelines (WCAG) through the W3C process, which means that the guidelines were developed in a

(15)

accessibility needs of individuals, organizations and governments in an international context. Web content in this context refers to any natural information on the web, such as text,

images and sounds, as well as the code that is used to present the natural information. WCAG is based on four main principles: perceivable, operable, understandable and robust. There are 13 guidelines that are sorted into the different principles. Each guideline has testable success criteria, which are structured into three levels, A, AA and AAA (World Wide Web Consortium, 2018). The WCAG have been used by governments in order to prescribe the level of accessibility that the public sector needs to comply with. The standard that is used in Swedish legislation is WCAG 2.1, AA (Laurin et al 2014). Some of the criteria specified in the guidelines are concerned with the possibility to decode information on the page, for example by using text alternatives to image-based content, minimum contrast values, minimum text sizes and not using colour as the only characteristic in order to convey information. Another category of criteria is about navigation and data input, such as being able to navigate the site through keyboard input, having sufficiently large targets and being able to pause processes and avoid timeout. There are also criteria concerned with the pedagogic layout of websites, such as making sure that navigation is consistent, providing clear headings and labels as well as meaningful sequences and labels and instructions (World Wide Web Consortium, 2018).

2.6 Universal Design

In Sweden, the association “Handikappsförbundens Centralkommitté” published a plan of action called “Ett samhälle för alla”, meaning “One Society for All”, in 1972. This plan served as the basis of the UN Convention on the Rights of Persons with Disabilities (CRPD) that was adopted in 2006 (Funktionsrätt Sverige 2017b). In the general principles of the

convention, it is declared that there should be “full and effective participation and inclusion in society” for persons with disabilities (United Nations 2006, p.5). The CRPD states that accessibility is a determinant prerequisite for upholding human rights for people with

disabilities (Myndigheten för delaktighet 2017). The United Nations Committee on the Rights of Persons with Disabilities states in General Comment No. 2 on the Convention of Rights of Persons with Disabilities that accessibility should be viewed not only in the context of equality and non-discrimination but also as a way of investing in society and as an integral part of the Sustainable Development Agenda (2014). When working towards inclusion, an approach only focused on specialist solutions is not adequate, as it hinders persons with disabilities to participate in society on equal terms with others. In order to secure inclusion, any barriers that hinder persons from participation need to be eliminated and as a complement, special solutions need to be found when they are needed (Myndigheten för delaktighet 2018). One of the general obligations for the States Parties of the UN Convention on the Rights of Persons with Disabilities is to “undertake or promote research and development of

universally designed goods, services, equipment and facilities”, as well as “promote universal design in the development of standards and guidelines.” (United Nations 2006, p.6).

The CRPD defines universal design as:

“[...] the design of products, environments, programs, and services to be usable to all people, the greatest extent as possible, without the need for adaptation or specialized design.

Universal design shall not exclude assistive devices for particular groups of persons with disabilities where this is needed.” (United Nations 2006, p.4).

(16)

2.7 Usability and Accessibility

Nielsen (1993), defines usability in the terms of learnability, efficiency, memorability, errors and satisfaction. The standard ISO 9241-11 describes usability as “the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency and satisfaction in a specified context of use” (ISO 2018). Subjective aspects also need to be taken into account in order to understand the experience of the user (Hassenzahl 2005). This makes the field of usability dynamic and context-dependant (Law et al. 2009). Following this general definition of usability, there is a need to explore the connection between usability and accessibility. “Accessibility and usability are two qualities that interact with each other and it has always been difficult to define the scope and extent of this

relationship” (Aizpurua, Harper and Vigo 2016, p.5). According to the W3C Web Accessibility Initiative, the matters of accessibility and usability have similar approaches, goals and

guidelines. The imperative of both usability and accessibility is that persons should be able to access and use digital solutions. The difference between the two fields seems to be the focus of the intended user, in the field of usability, users with disabilities are often neglected, whereas these users are the primary focus in the field of accessibility (W3C Web

Accessibility Initiative 2016). Inclusive design aims at including human diversity into the conceptual group of users that is designed for, which means that in the context of inclusive design, there is an overlap between usability and accessibility. Inclusive design highlights diversity in the population and aims to apply this understanding to the design process in order to respond to these differences and create a more satisfactory design for more people (Waller et al. 2015).

Even though the user groups targeted by usability and accessibility usually differ, studies imply that there is an overlap in usability and accessibility problems. One study showed that there was an overlap between the issues that sighted and blind users faced in user

experience testing of websites (Petrie and Kheir 2007). Another study, investigating if accessible design benefits general users of public digital services, showed that accessibility is a predictor of the perceived usability of a digital service (Bai et al. 2019), indicating that considering accessibility could lead to improved usability. High levels of accessibility have been shown, when compared with websites with low levels of accessibility, to improve general user attitudes in the aspects of perceived usability, aesthetics, workload and trustworthiness (Schmutz, Sonderegger and Sauer 2016, 2017).

Among design practitioners, there seems to be a concern that designing for accessibility necessitates overly simple and functional solutions and negate sophisticated and interesting visual design (Petrie, Hamilton and King 2004). Some studies have shown significant

correlations between UX attributes and web accessibility (Mbipom and Harper 2011, Aizpurua, Harper & Vigo 2016). Other research suggest that accessibility does not need to constrain visual design, arguing that the goal of good accessibility is only yet another objective for designers to fulfil, alongside other objectives such as visually pleasing design (Petrie, Hamilton and King 2004).

Several standards and guidelines are available in order to overcome the challenges that the technically heterogeneous web environment poses (Vollenwyder et al. 2019). One of the reasons why these standards are important for accessibility is because they are a

(17)

prerequisite for being able to develop and use assistive technology (Funka 2019a). Assistive technology is the technology that enables some people who would not otherwise be able to access the Internet, to adapt the technology in order to fit their needs. An example of assistive technology is a screen reader, which is a piece of software or hardware that

transforms visually represented information into spoken word or braille output (Funka 2019b). At the same time, accessibility standards and compliance with these standards does not automatically make a website or mobile application accessible. A validation error on a website might not pose an issue in terms of accessibility, whereas a website that complies with all standards might not be accessible at all (Funka 2019a). Nevertheless, the current trend in web accessibility is that it is compliance-driven, with the primary aim to fulfil the consensual standards (Cooper et al. 2012). An approach that only focuses on the

compliance with guidelines can neglect to address the actual needs of the users, risking that the design solutions do not prove useful for users with disabilities, even though compliance has been achieved (Power, Freire, Petrie, and Swallow 2012, Sloan and Kelly 2011). One way of increasing accessibility is by including users with different abilities in the design and development process (Vollenwyder et al. 2019). User-centred methods are not new to the field of web development, but it is not common for persons with disabilities to be included in these practices (Henry 2006). Hosking, Waller and Clarkson (2010) suggest that including boundary users in the design process is fruitful. The boundary cases are “those people on the borderline for exclusion, frustration or difficulty” (Hosking, Waller and Clarkson 2010, p.498). In conventional design methods, the user that is designed for is a placeholder for an average user. This user-selection focuses on the shared needs of average users. Assuming that the perceived difficulty of using a web service follows the principle of normal distribution, designing for the average user will mean that about half of the population will face varying degrees of difficulties in the interaction with the service. If the designer chooses to select a boundary user instead, the service will likely be more usable by a larger amount of people (Johansson 2016).

Figure 3. Using average users or boundary users as the focus of the design process. Illustration based on a graph from Johansson (2016).

(18)

2.8 Methodological maturity in organizations

Maturity models emerged from the field of quality management (Jokela et al. 2006) and were developed in order to be able to measure quality in processes. The progenitor of maturity models is Crosby’s Quality Management Maturity Grid (QMMG), which describes five maturity levels of quality management and the identifying factors displayed by the organisation at these levels (Jokela et al 2006, Crosby 1979). The two domains where maturity models appear most frequently today in academic papers are software engineering and software development. In a mapping study (Wendler 2012) it was found that maturity models have been applied in 47 domains outside of software development, which indicates that there is a “broad and general applicability of the maturity concept” (Wendler 2012 p.1332).

One branch of maturity models is aimed at assessing usability capability maturity (UCM). Although all UCM models set out to measure usability, they seem to stem from a diverse range of research traditions, as their features are quite different from each other (Jokela et al 2006). Also, few UCM models provide detailed guidelines for the application of the model and most models lack empirical validation. This shows that the field of UCM models is not mature, even though the first models that referred to the maturity of usability, in particular, are more than twenty years old (Carvajal & Moreno 2017). Even though these factors indicate that cautious use of these models is advisable, researchers argue that the models still are valuable tools to assess weaknesses and strengths in the current processes of the organization. This assessment can serve as a basis for planning and implementation of future actions for increasing the maturity of usability practices in the organisation (Jokela et al 2006, Carvajal & Moreno 2017, Capretz, Raza & Ahmed 2012, Sauro, Johnson & Meenan 2017).

(19)

3. Method

In order to reach to an understanding of how the integration of accessibility into design practices in public agencies can be strengthened, a preliminary study was conducted, followed by a case-study, upon which a design intervention was based.

The preliminary research was made up of desk research, a preliminary interview study and some unorganized research activities. These research activities were conducted in order to gain an understanding of the field of accessibility, in terms of legal, organizational and designerly aspects. Unorganized research activities, as well as the preliminary interview study, also pointed out areas of interest for the desk research. These preliminary research activities created a basis for the following research activities and informed how the main interview study was conducted. The instrumental case-study consisted of an interview study as well as a review of documentation and other materials. The data from the case-study was analysed through an accessibility maturity assessment, which was done by applying an adapted version of the Usability Maturity Model: Human-Centeredness Scale (Earthy 1998). The results of the assessment provided design openings for the intervention, which was designed as a handbook that maps design methods for accessibility to the attributes of the accessibility maturity model. The handbook was created as a prototype to work as a proof of concept and was validated formatively by persons from the design and tax agency. The timeline of the research activities is presented in Figure 4, with the caveat that there was some overlap between the case study and the unorganized research activities.

(20)

3.1 Preliminary inquiries

The aim of the preliminary inquiries was to gain an understanding of the field of accessibility in general, how it relates to design in particular, how it is regulated judiciously and how practitioners currently work with the issue. As I had little previous knowledge in the field, the initial research phase was explorative. I was keen to take advantage of any learning

opportunity that presented itself to me, which meant that I attended courses, webinars, workshops, inspiration sessions and sharing sessions that were open to me. Three unorganized research activities in particular were formative for this thesis and therefore outlined in Chapter 4: a series of workshops with Begripsam, a sharing session about accessibility at Idean and a session about design systems at the Tax Agency.

In parallel with doing desk research and unorganized research activities, I performed an interview study. To start with, I interviewed four people from Idean. These interviews provided tips for further desk research on digital accessibility, as well as tips for external contacts to involve in the research. These initial interviews helped me to test the interview guide and refine it for upcoming interviews with external persons. The same interview guide (Appendix 1) was used for all interviews.

I interviewed five people who work with digital accessibility outside of Idean. The

interviewees all have a background in User Experience but work in different roles for different organizations. The participants were chosen through purposive sampling, a non-probability sampling method that aims at selecting persons who have the relevant experience needed to answer the research questions (Etikan 2016). The purpose of the interviews was to learn about how the interviewees approached digital accessibility in their professional roles and how their organizations worked with digital accessibility.

The participants were:

• A Design manager at a bank

• A Product owner at a public transport company

• An Accessibility consultant and founder of a consulting agency focused on accessibility

• Head of a team working with digital services at the Swedish Social Insurance Agency • A consultant with accessibility expertise, working at the Swedish Social Insurance

Agency

The interviews were semi-structured and lasted for around an hour each. The interviews were conducted in-person. With permission from the interviewees, the interviews were recorded and later transcribed. In order to analyse the transcribed material, the principles of phenomenological analysis of interview data (Hycner 1985) were used. According to

Creswell and Poth, a phenomenological study “describes the common meaning for several individuals of their experiences of a concept or a phenomenon” (2018, p.75). In order to do this, the researcher collects data from several participants, which serves as the basis of a summative description, rather than analysis or explanation, that aims to convey the unifying essence of the lived experience that is researched (Moustakas 1994). Hycner (1985) points out that phenomenological analysis is an approach to analysing data and that upholding the spirit of the method is more important than following the method step-by-step. I condensed the fifteen steps that Hycner (1985) describes into the following five steps:

(21)

1. Transcription of conducted interviews.

2. “Bracketing” any previously held conceptions, which means suspending

interpretations and meanings in order to approach the reality of the interview subject with openness

3. Listening to the audio recording of the interviews in order to gain an understanding of the interview as a whole

4. Noting down units of general meaning and clustering these units

5. Using the clusters of units of general meaning, then writing a composite summary, which aims to “capture the essence of the phenomenon investigated” (Hycner 1985, p. 294).

3.2 Case study

This case study was aimed at investigating how accessibility is currently integrated into work processes in a government agency. The case study is both an investigative tool, as well as the aim of the design intervention.

The Swedish Tax Agency was chosen for the case study, due to its relevance to the scope of this thesis, as well as their willingness to be studied. The case study of the Tax Agency is an instrumental case, a specific case chosen to understand and illustrate a broader issue (Stake 1995). Thomas (2015), highlights the importance of well-chosen parameters that define the chosen case, which makes it bounded (Creswell & Poth 2018). The scope of this case study is that it focuses on the digital accessibility of the Swedish Tax Agency, from February to April of 2020. The case study uses an embedded case study design instead of a holistic design (Yin 2014), in order to give a clear focus for the study. The main unit of analysis is the client meeting team, which is complemented by additional units of analysis containing

individuals who have relevant input on the topic of digital accessibility but work outside of the core team. The client meeting team is the team that is responsible for the User Experience of all digital services. This team manages the design system, as well as guidelines and

checklists that other functions in the organization should use in order to ensure a coherent User Experience, as well as digital accessibility. This means that the client meeting team has a central role to play in the area of digital accessibility, making the team relevant as a core focus in the case study.

3.2.1 Data Collection

In order to conduct a case study, evidence needs to be collected from several sources. Examples of sources of evidence are documentation, archival records, interviews, direct observations, participant observations, participant observation, physical artefacts (Yin 2014). The sources of evidence chosen for this case are documentation and interviews. Interviews can give valuable insights on the specific case study topics, although they might suffer from biases and reflexivity (Yin 2014). Through using documentation, the insights from the interviews can be triangulated. Initially, the plan was to also use direct and participant observations as two additional sources of evidence, but due to the development of the epidemic Coronavirus Covid-19 in Sweden in March of 2020, this fell short. I was able to visit the Stockholm office of the Tax Agency in February, which gave me a chance to introduce myself and get to know the client meeting team. I was also able to attend one of their weekly meetings, as well as attend two workshops with Begripsam together with members of the

(22)

team before the agency started working from home and it was recommended not to travel to Stockholm.

My contact person at the Tax Agency was the team lead of the client meeting team. The participants of the interview study were chosen through a dialogue with the contact person, who established contact with the interview candidate, ensuring that the person was willing to do the interview before giving me the contact information and green light to contact them. The contact person served as both a facilitator and a gate-keeper, which might have skewed the selection of the interviewees. I also asked the interviewees if they thought that I should contact anybody else, which led to some additional interviews.

The purpose of the interviews was to learn how accessibility currently is integrated into working processes. Some of the topics that the questions aimed at were about gaining an understanding of how the different roles of the organization work with accessibility, if the employees perceive the work to be well-structured and well-supported, as well as learning about problems and barriers in the work. The same interview guide was used for all

interviews (Appendix 2). Initially, the plan was to interview all interviewees in-person, but due to the aforementioned pandemic-related circumstances, the interviews were instead

conducted remotely. Unfortunately, the security precautions of the agency barred video calls from external parties, which meant that the interviews needed to be conducted via a

telephone connection. The lack of visual feedback made the interviews a little more difficult. With the consent of the interviewees, the calls were recorded and later transcribed. The length of the interviews ranged from fifteen minutes to more than an hour. The following table provides an overview of the participants of the interview study.

Role Responsibilities Department/unit/team Office Interview length Head of

Section (UX)

Responsible for all employees who work with UX,

content and web-analysis

Client meeting team, unit for service development

Stockholm 34:07

UX-lead Responsible for UX and digital

accessibility

Client meeting team, unit for service development

Stockholm 1:11:40

Visual designer

Visual design, UI Client meeting team, unit for service development

Stockholm 15:25

Coordinator Works with testing and involvement of users

IT-department, works with the client meeting team

Stockholm 32:31

Administrator Works with UX and usability in the

Administrative

department, works with the client meeting team

(23)

development of internal systems Graphic

Designer

Responsible for the visual identity of the agency

Communications department

Stockholm 25:11

Communicator Works with making the textual communication more understandable Communications department Stockholm 56:25

Controller Coordinator for questions about accessibility Human resource department Malmö 45:16 Head of Section (IT) Currently working on establishing the IT-department in Malmö

IT-department Malmö 34:07

Figure 4. Overview of interviews conducted for the case study

3.2.2 Data analysis

The data from the case study was analysed through an accessibility maturity assessment. The assessment aims at providing an understanding of the maturity of accessibility practices at the government agency, as well as pointing towards possible design openings for a design intervention.

According to the frame of analysis of the maturity model, the interview data was analysed in a qualitative and inductive way. The categories for the analysis of the interview data were chosen a priori on the basis of the attributes that the maturity model prescribes. The

assessment of accessibility maturity is about the client meeting team at the Tax Agency, as well as other parts of the organisation that work with the development of digital services. It is not an assessment of the general level of accessibility maturity of the entire government agency.

In order to establish a frame of reference on which to base a well-directed intervention, I chose to assess the abilities of the agency in terms of accessibility with an adapted maturity model. In my research, I could not find an empirically validated and consensual model that was specifically about accessibility. Therefore, I researched models that aimed at integrating usability and human-centeredness into organisations. The main research strategy for finding a suitable maturity model to adapt was by reviewing academic surveys of maturity models (Jokela et al. 2006) (Carvajal & Moreno 2017). Factors of scope, method of application, amount of concrete guidance for application of the model and the amount of empirical validation of the model were used as descriptors in the aforementioned surveys and important variables in the choice of the model. One of the more well-known and influential models in the field is the “Usability Maturity Model: Human-Centeredness Scale”, abbreviated

(24)

UMM-HS (Earthy 1998). This model provides detailed guidance for the application of the model and has been developed iteratively on the basis of trials (Jokela et al 2006) and validated by case-studies (Carvajal & Moreno 2017). UMM-HS is a consensual model, which means that it is based on “widespread review and agreement by experts” (Jokela et al 2006, p.269). UMM-HS is also a model that is based on all previous maturity models that were available at the time. Additionally, the model provides guidance in the application of the model and is suited for self-assessment (Jokela et al 2006).

UMM-HS is meant to be used in order to assess how well human-centeredness has been integrated into the management of an organization and to which extent quality in use has been established. Earthy cites the benefits of human-centeredness from the standard ISO 13407:

“An approach to interactive system development that focuses specifically on making systems usable. It is a multi-disciplinary activity, which incorporates human factors and ergonomics knowledge and techniques. The application of human factors and ergonomics to interactive systems design enhances effectiveness and efficiency, improves human working conditions, and counteracts possible adverse effects of use on human health, safety and performance. Applying ergonomics to the design of systems involves taking account of human capabilities, skills, limitations and needs.” (Earthy 1998, p.8).

This approach of using design in order to cater to user needs that vary in terms of

capabilities is congruent with using universal design for accessibility. Earthy (1998) uses the term quality in use and defines it as an effective, efficient and satisfactory experience of specified users when achieving specific goals in specified environments. It is possible to use this framework to focus on accessibility by specifying the users as users with diverse

capabilities. Instead of assessing general considerations of quality in use, the focus of this maturity assessment will instead lie on accessibility, which can be seen as a subset of usability or quality in use. Due to the overlap of the areas of usability and accessibility, the model is well-suited for an adaptation to accessibility.

The UMM-HS model consists of capability levels, which consist of process attributes, that in turn consist of management practices. How these building blocks of the model are

represented visually can be seen in Figure 5.

Level -

Capability level

-.1 Process attribute

-.1.1 Management practice that should be carried out in order to achieve the process attribute.

Figure 5. Explanation of how the Usability Maturity Model: Human-Centeredness Scale (Earthy 1998) is structured.

The following model is an adaptation of the UMM-HS model that has been adapted in order to explicitly focus on concerns of maturity of the management of accessibility practices, rather than general concerns of usability. All words that have been removed in the adaptation

(25)

are presented with a strike-through, like this. All words that have been added to the model in order to refocus it on accessibility, are written in bold, like this.

Level X

Unrecognised

(no indicators)

Level A

Recognised

A.1 Problem recognition attribute

A.1.1 Problem recognition. Management and staff are aware that there is a need to improve aspects of the systems under development concerned with their use accessibility.

A.2 Performed processes attribute

A.2.1 Information collection. Information is collected which could be used to take account of user requirements of accessibility.

A.2.2 Performance of relevant practices. Practices are performed which could be used to include information about user requirements of accessibility in the system or service.

Level B Considered

B.1 Accessibility awareness attribute

B.1.1 Quality in use Accessibility training. Staff are made aware that quality in use accessibility is a particular attribute of a system which can be improved. B.1.2 Human-centred Accessibility methods training. Staff are made aware that

quality in use accessibility is achieved through the use of a series of human-centred processes during the development and support/use of a system. B.1.3 Human-systems interaction training. Staff are made aware that

human-centeredness accessibility covers the total system, not just the user interface or the physical ergonomics singular digital interfaces.

B.2 User focus attribute

B.2.1 User consideration training. Staff are made aware that the accessibility-needs of the end-users of the system should be considered when developing or supporting the system.

B.2.2 Context of use training. Staff are made aware that end user’s skills, background and motivation abilities, needs and context may differ from developers or system support staff.

(26)

C.1 User involvement attribute

C.1.1 Active involvement of users. The development process ensures understanding of user needs through user involvement in all development phases.

C.1.2 Elicitation of user experience. The design solution is shown to stakeholders boundary users and they are allowed to perform tasks (or simulated tasks). C.1.3 End users define quality-in-use in terms of accessibility. Systems are tested

using measures of quality in use derived from end-users.

C.1.4 Continuous evaluation. Early and continual testing is an essential element of the development methodology. The process is based on the necessity for feedback from users.

C.2 Human factors Accessibility technology attribute

C.2.1 Provide appropriate human-centred methods. Select and support methods for the elicitation of user input at all stages in the lifecycle.

C.2.2 Provide suitable facilities and tools. Suitable facilities and tools are provided for accessibility activities.

C.2.3 Maintain quality in use accessibility techniques. Ensure that methods and techniques are reviewed for suitability and that state-of-the-art interface technologies are used as appropriate in developing new systems. C.3 Human factors Accessibility skills attribute

C.3.1 Decide on required skills. Identify required competencies and plan how to make these available in order to facilitate multi-disciplinary design solutions. C.3.2 Develop appropriate skills. Development of appropriate skills in

human-centred staff either by training or by job experience.

C.3.3 Develop appropriate staff. Skilled staff are involved and effective in all stages of development as and when required.

Level D Integrated

D.1 Integration attribute

D.1.1 Integrate Human Factor accessibility processes. Integration of quality in use accessibility processes with quality system.

D.1.2 Facilitate interface between Human Factors teams working with

accessibility and the organisation. Ensure that the department promoting a human-centred an accessible approach understands and uses the language and working methods appropriate to successful interaction with other

(27)

D.1.3 Use appropriate representations. Representations of user requirements and changes to the system arising from user involvement should be

understandable by system developers and programmers. D.2 Improvement attribute

D.2.1 Ensure design feedback. Ensure that evaluations of accessibility take place at all stages in order to influence the system to be delivered.

D.2.2 Change based on feedback. The development process encourages design changes based on actual user experience.

D.2.3 Timing of feedback. Ensure that information on user needs and quality in use accessibility defects is fed into the design process at appropriate times and in the right format for use.

D.3 Iteration attribute

D.3.1 Minimize risks by iterations of design. Iteration of the design using prototypes etc. increases the match between the final system and user expectations. D.3.2 Manage iteration of design solutions. Information should be recorded to

manage the progress of iterative design.

D.3.3 Use design objectives to control iteration. Manage the prototyping process by setting and monitoring target quality in use accessibility levels set for

particular aspects of the system.

Level E Institutionalised

E.1 Accessibility leadership attribute

E.1.1 Manage usability accessibility programme. Management of the whole programme of human-centred accessibility processes on all projects in a department or organisation.

E.1.2 Systematic improvement of quality in use accessibility. Use quality in use accessibility defects to analyse and improve problems with the organisation’s processes, thereby reducing quality in use accessibility defects.

E.1.3 Human-centred Improvement of organisation. The approaches used to ensure that systems are human-centred accessible are also used within the

organisation to improve its own processes and systems. E.2 Organisational accessibility attribute

E.2.1 Organisational implementation of user-centred accessibility practices. Assist the organisation in the establishment and use of human-centred accessibility tools and methods, and in maintaining a focus on the considerations of user issues.

(28)

E.2.2 Acceptance of human-centred accessibility skills. Recognition of the pivotal role by human-centred accessibility skills in an integrated development team. Management of the resources available - human and others.

Figure 6. Accessibility Maturity Model. Adaption of the Usability Maturity Model: Human-Centeredness Scale (Earthy 1998).

The documentation of the maturity model suggests that one or two interviews should be conducted with the explicit focus of asking questions surrounding the process attributes from the maturity model (1998). The interviews from the case study were not conducted to

explicitly follow the framework of the model but aimed to create a broad understanding of the case. Instead of only interviewing up to two employees, the assessment is based on nine interviews. The analysis of the data, the maturity assessment, was conducted according to the documentation of the model. Each practice attribute of the model was rated for each respondent, based on evidence from the interview transcripts. When there were multiple attributes, they were summarized by a combined rating for each respondent. Then, these combined ratings were in turn combined in order to reach one rating for each process attribute. The process attributes were combined in order to reach a single rating for each level. How this assessment works is illustrated in Figure 7.

(29)

Earthy’s (1998) proposed rating scale was used in the assessment. The scale has four steps and ranges from “not achieved” to “fully achieved”.

N Not achieved There is no evidence of the defined practice P Partially achieved There is some evidence of the defined practice

L Largely achieved There is significant achievement of the defined practice F Fully achieved There is full achievement of the defined practice

Figure 8. Rating scale used in the maturity assessment, based on the documentation of the Usability Maturity Model: Human Centeredness Scale (Earthy 1998).

According to Earthy (1998), when it isn’t clear to what level of achievement a practice is achieved, the analyst should give the benefit of the doubt and rate to the higher level. The rating process should be repeated, going up the levels of the model, until there is no

evidence of achievement of the defined practice. In the methodology that Earthy (1998) lays out, respondents are asked to answer questions for the overall organization. In the interviews conducted as part of the case study, the interviewees mainly talked about how accessibility pertains to their own domain. This has implications on the individual-level test results, as it is not feasible that all roles should conduct user tests with boundary users for example. In the assessment, respondents who are not involved in the area that the attribute describes are not included in the rating for that particular attribute.

3.3 Design intervention

The initial aim of the design intervention was to find solutions to the design openings that the accessibility maturity assessment pointed towards. The accessibility maturity model was intended as a method that made it possible to analyse the case study data in a structured way. Upon finishing the assessment, it became clear how useful the methodology was in order to identify the necessary changes of accessibility practices. It seemed important, both for the creator and the receiver of the design intervention, to have an understanding of the level of maturity of the organisation and the areas of weaknesses. With the realization that the accessibility maturity assessment itself was important for the design intervention, it changed from being a methodology to arrive at a design intervention, to becoming a part of the design intervention itself. The aim for the intervention became to find a way of integrating the accessibility maturity assessment into design processes. The core idea was to connect design practices to the specific attributes of the maturity model. The goal of the design intervention was to package this methodological idea in a format that would be useful for practitioners, in order to be able to communicate the idea and receive formative feedback on it.

Ease of prototyping as well as suitability as a proof of concept were the two main parameters in choosing the format. Idean had previously used the format of a handbook in order to communicate design methodology to consultants and clients. It is a proven way of inspiring and instructing persons with different backgrounds in new ways of designing. The prototype is meant to be a proof of concept, focusing on how the methodology of assessment and related design activities could be implemented in organisations. Therefore, the overall visual

(30)

design of the prototype was of lesser priority. As there already were handbooks available that used the Idean design libraries and had established visual design patterns, it allowed me to create a proof of concept without having to spend too much time on the visual design. This made it possible to create a more extensive and developed proof of concept for the

validation.

The design methods and other content in the handbook were created through a design process with a long phase of incubation. The methods are based on previous knowledge of design methods and process, that I have acquired through education and practice of design. From the research phase, insights from the preliminary study, case study, general desk research and specific research about inclusive design methods were combined. Some ideas for design interventions arose throughout the thesis work, but there were also planned phases of ideation where design interventions for all attributes of the maturity model were brainstormed.

(31)

3.4 Validation

The prototype was validated formatively with persons from the design agency and the Tax Agency. The participants in the validation received a digital copy of the prototype as well as a question that the discussion centred around. The participants in the validation can be split into three different groups, designers from the design agency, designers from the Tax Agency and non-designers from the Tax Agency. The designers from the design agency were asked about how they could see the handbook being used and how the handbook could be changed in order to be useful in practice. The same question was asked to the designers in the Tax Agency. To the non-designers in the Tax Agency, the question was a more open call for feedback, as they were not the target audience for a practitioner’s use of the handbook.

Based on the nature of the work and context, a formative validation was likely to yield more rich and useful feedback than a summative validation. Asking a close-ended question such as whether the methodology of connecting design practices to a maturity assessment is useful for including accessibility into design practices would likely lead to speculative answers of little analytical value. Because people in general want to be kind and the

participants in the validation mostly are people I have had previous contact with, it would be unlikely to receive a negative answer to the question. A formative validation of the prototype allows for a more nuanced and rich discussion and can lead to an analysis of what should be changed in further iterations.

The validation was conducted with five people from the design agency. Four of them were designers, one was a front-end developer. Two of them were from the Idean Stockholm office and were not experts in accessibility. The remaining three had an expressed interest in accessibility, two belonged to the Idean Oslo office and one belonged to the Idean

Amsterdam office. Other than that, three people from the Tax Agency were part of the

validation. I received feedback from the UX-lead, Visual designer and Head of the IT-section, who were also part of the case study. The validation was conducted through one-on-one discussions. The format for the discussions ranged from video call, to telephone call and chatting on messaging platforms. The differences in format were due to practical differences in the availability of the respondents.

Using an Accessibility Maturity Model to Facilitate the Inclusion of Accessibility in Design Practices