Content uploaded by Luciana A M Zaina
Author content
All content in this area was uploaded by Luciana A M Zaina on Mar 21, 2018
Content may be subject to copyright.
Investigating accessibility issues of UI mobile design paerns in
online communities: a virtual ethnographic study
1st Author Name
Aliation
Address
e-mail
2nd Author Name
Aliation
Address
e-mail
3rd Author Name
Aliation
Address
e-mail
ABSTRACT
With the dissemination of mobile devices and the migration of
activities that were once executed only on desktop computers to
smartphones and tablets, the concerns related to accessibility in this
environments have increased. However, the accessibility impact
of mobile interface design paerns in the life of disabled people
has not been widely addressed in academic works. At the same
time, the community of mobile designers and developers has made
signicant advances in identifying accessibility issues with design
paerns on mobile interfaces, reporting these ndings in virtual
spaces of discussions as forums, blogs, etc. Aiming at exploring the
experiences and knowledge of professionals, this paper presents
an ethnography study in 18 virtual communities of mobile design
and development with the goal of identifying issues on the acces-
sibility of mobile UI design paerns. As a result, we analyzed 127
documents in order to propose recommendations to improve the
accessibility of mobile interfaces.
KEYWORDS
Mobile UI Design Paerns, Mobile Accessibility, Virtual Ethnogra-
phy
ACM Reference format:
1st Author Name, 2nd Author Name, and 3rd Author Name. 2017. Investigat-
ing accessibility issues of UI mobile design paerns in online communities:
a virtual ethnographic study. In Proceedings of Proceedings of the 16th Brazil-
ian Symposium on Human Factors in Computer Systems, Joinville, Santa
Catarina, Brazil, October 2017 (IHC ’17), 10 pages.
DOI: 10.1145/nnnnnnn.nnnnnnn
1 INTRODUCTION
Accessibility is dened as the quality or characteristic of something
that is possible to approach, enter or use, it refers to the design of
products were people with disabilities are able to access [
20
]. e
interest of studying the accessibility subject has been increased
recently, opening the eyes of practicals and researchers to the rele-
vance of uncovering the issues and new challenges about the theme
[14].
e latest report (2011) on world disability by the World Health
Organization (WHO) estimates that more than one billion people
currently live with a disability. In other words, approximately
Permission to make digital or hard copies of part or all of this work for personal or
classroom use is granted without fee provided that copies are not made or distributed
for prot or commercial advantage and that copies bear this notice and the full citation
on the rst page. Copyrights for third-party components of this work must be honored.
For all other uses, contact the owner/author(s).
IHC ’17, Joinville, Santa Catarina, Brazil
©2017 Copyright held by the owner/author(s). 978-x-xxxx-xxxx-x/YY/MM. .. $15.00
DOI: 10.1145/nnnnnnn.nnnnnnn
15% of the entire world population has a disability [
31
]. In Brazil
the scenario is not dierent, according to a census of 2010 by the
Brazilian Institute of Geography and Statistics (IBGE) 23,9% of the
Brazilian population declares to have some kind of visual, auditory,
cognitive or motor disability [15].
Furthermore, in the last 50 years the life expectancy in Brazil has
increased in 25 years [
14
]. is means that the population of elders
has increased and will continue to increase in the next years. As
the population ages, it’s more likely that the number of people with
disabilities will grow as elder users experience a decline in several
abilities and thus age is related to the appearance of disabilities
[
16
]. On most cases, people that have visual, cognitive, motor
or auditory deciencies require assistive devices (screen readers
or special braille keyboards, for example). However, the use of
assistive devices is not sucient to provide access to any content
on mobile devices as it’s necessary mobile applications and websites
that also delivers accessible content [27].
Mobile devices provide access to information to any user, any-
where and modern smartphones even oer more processing power
than some desktop devices. For the disabled population, mobile
devices are also an important ally as they concentrate several func-
tionalities in a single device (phone, internet browser, camera and
media player) and many of these features can be used to improve
the access to information of these users [
12
]. For example, users
with visual impairment can use mobile devices to recognize money
bills and hearing impaired users oen use video face calls to com-
municate with other users through sign language. Although the
design of user interface is essential for promoting the accessibility
to any content, in some cases problems with the interface may
lead to accessibility barriers [
7
], which are conditions that make it
dicult for people to achieve a goal when using the web on mobile
or desktop devices.
In the sense of dealing with accessibility aspects, the UI design
paerns might support the building of mobile user interfaces [
30
].
UI design paern is an interface element, that is a widely used
concept on Soware Development [
35
]. In specic situations, how-
ever, these paerns are not given much importance in terms of
user interaction with the paern and the usability and accessibility
aspects that comes along [9].
Searching for guidelines, the World Wide Web Consortium (W3C)
provides documentations with several guidelines and recommen-
dation for making accessible pages (as the WCAG -Web Content
Accessibility Guidelines- for content accessibility) that are part of
the Web Accessibility Initiative (WAI). Although there are recom-
mendations for building accessible content for the general web,
mobile guidelines on accessibility are still being made and the doc-
umentation is currently a dra [19].
Draft version - wrote up before camera ready
IHC ’17, October 2017, Joinville, Santa Catarina, Brazil blind-review
Considering the technological evolution of mobile devices, the
online spaces have been recognized as spaces of sharing practical ex-
periences, creating repositories about the theme. e exploration of
online spaces, where professionals discuss problems and solutions
of dierent topics, can reveal important practical knowledge which
trigger trends and issues to be worked by the academic community
[38].
is paper presents an investigation on accessibility of UI ele-
ments (design paerns) of mobile applications. e study consists
of a virtual ethnography [
23
] of 18 information sources that resulted
in the analysis of 127 documents from both forum discussions and
blog posts and comments in the subject of mobile accessibility. e
goal of the study was to identify accessibility issues and recom-
mendations to beer understand the problems faced by disabled
users when using the UI mobile design paerns. For the analysis
of the virtual ethnography results, the qualitative research method
Grounded eory [
40
] was used in order to identify key informa-
tion that may help to build mobile applications that comply with
the needs of disabled users.
e main contributions of this paper are the communication of
an study on UI design paerns accessibility with the use of virtual
ethnography as means to nd evidences of practical problems and
solutions on the accessibility of mobile interfaces throughout de-
sign paerns. Also, we present recommendations, to some of the
interface design paerns, for overcoming accessibility issues that
UI design paerns may cause when used.
e next sections of this paper are structured as follows: Section 2
introduce important concepts, while Section 3 present related works
to this research and Section 4 explain the planning, execution and
analysis performed in the study. Section 5 presents the ndings of
the study and, nally, Section 6 concludes this paper.
2 FUNDAMENTALS
is section introduces important concepts that supported the top-
ics of this paper (accessibity and UI mobile design paerns) and the
methodology used to conduct the study (virtual ethnography).
2.1 Accessibility and Design Patterns
e Web Accessibility Initiative (WAI) of the World Wide Web Con-
sortium (W3C) [
19
] presents guidelines, strategies and resources
related to accessibility in various contexts. However, some guide-
lines dened by W3C are not applicable or adapted to the mobile
world and, as the web become more mobile and ubiquitous, the
need for accessibility in any context and device is urgent [1].
Alongside the impairments and accessibility barriers that users
encounter while using a website, their mood is directly related
to the overall experience. e accessibility of a website or app is
benecial for everyone as it improves the usability for every user
[
32
]. Another important factor is that people with disabilities are
not the only ones that have problems using websites and mobile
devices, as there is an exponential growth of the elderly population
in the world that suers from a wide variety of disabilities developed
by the advanced age [
28
]. In the mobile context such users have
even more diculties, as, sometimes, they are not used to mobile
technology.
Universal Accessibility or Universal Design is the process of
developing a system or product that can be used by any person
with any disability (or without disabilities), in any situation or en-
vironment [33]. In Human Computer Interaction (HCI), Universal
Design is mainly related to the development of computer systems
for any person with any limitation [
21
]. It is important to point out
that many accessibility barriers faced by people with disabilities
also are witnessed by other users. However, for disabled people,
accessibility is essential in order to actually have access to informa-
tion, while for people with situational accessibility, this concern is
mere convenience [22].
e concept of Design Paern was developed with the goal of
being components of language that could be used as dialog about
organizations and development process. Design Paerns can be
understood as best practices or some sort of heuristics of well knows
design problems [25, 37].
As there are many classications and specic applications for
Design Paerns, several studies are made on individual categories
of paerns as the work presented by Sampaio [
37
] in which a tool
was developed to automatically adapt web sites navigation menu in
dierent contexts of screens, resolutions and devices according to
common menu design paerns that are organized into two orders of
navigation: simple navigation and multi-level navigation. Although
the work of Sampaio is not focused on accessibility, it discusses
aspects of the usability of navigation design paerns that have
direct impact on the accessibility of mobile interfaces as well.
ere are various paern libraries that can be consulted during
the design process of web and mobile interfaces and Ribeiro and
Carvalhais [
34
,
35
] present a comparative analysis of some of the
most known libraries in order to propose 21 Design Paerns for
general purpose derived from studies in mobile web design with the
objective of helping designers and developers to design mobile and
web interfaces. e paerns dened are later applied to three case
studies in order to ascertain about their usefulness and applicability.
2.2 Virtual Ethnography
e path to understand a social group, its culture, institutions,
interpersonal interactions and beliefs is challenging and may seen
even more inaccessible when not using the appropriate research
method [
3
]. Ethnography may help in situations where the goal is
to understand a community by analyzing it’s behavior and social
interactions as it is a qualitative research method that can combine
techniques of observation, participation, interview and document
analysis as tools to aid in this discovery [18].
While classical ethnography research focuses on the study of so-
cial interactions of communities that take place in physical spaces,
virtual ethnography transfers the research site to socials spaces in
the Internet [
24
]. With the emergence of new forms of interaction
on the Internet as online multiplayer games and social networks
as Myspace, Orkut and later Facebook, dierent seings for ethno-
graphic studies appeared with new challenges of data collection
from chat logs and dierent types of media as pictures, video and au-
dio les. is new reality demands to the development of strategies
to interact with the online communities [6].
Virtual ethnography shis the location of observation to online
spaces adapting the classical ethnography directions [
5
]. Sharp
Investigating accessibility issues of UI mobile design paerns IHC ’17, October 2017, Joinville, Santa Catarina, Brazil
et al
.
[
38
] suggest ve dimension which aid the researcher in the
planning of an ethnography: (a) participant and non-participant
observation (levels of participation between only observation and
total participation [
24
]); (b) duration of the eld study; (c) space
and location; (d) use of theoretical underpinnings to support the
conduction and the analysis of results; and (e) the ethnographers’
intent in undertaking the study. In virtual ethnography, the rst
puzzle that the researcher must face is choosing a participation
role in the research. While some researchers in the literature may
agree that some degree of participation is necessary, the nature of
the community may force the researcher to adopt an observer or
hybrid position in the study [24].
Dierent qualitative techniques can be used with the purpose of
carrying out of results analysis. One of them is Grounded eory
(GT), a method that uses a systematic set of procedures in order
to develop and to derive inductively a theory grounded about a
phenomenon based solely on the data collected [
40
]. Based on the
collection of qualitative data, the researcher reviews the data and
analyze concepts and ideas that are classied in the form of codes.
Coding is a process where data is segmented on small fractions
that can be conceptualized and integrated in order to form a theory.
is way, codes are aggregated in concepts. Later, these concepts
are organized in categories, that are the base to build the theory.
e analysis and coding process reaches its end when theoretical
saturation is reached, that is, when nothing new emerges from the
process of analysis and categorization of the data [2].
As a result of the Grounded eory method, many artifacts
can be created an one of the most important artifacts is the Memo.
Memos are notes about the data, codes, relationships between codes
or even conclusions that the researcher nds useful to beer mold
the theory [
4
]. e creation of memos is extremely important as
these annotations are the rst aempt to communicate in textual
format the results of a Grounded eory.
3 RELATED WORKS
Many eorts have been made in the challenge of improving ac-
cessibility on mobile devices for users with dierent disabilities.
D
´
ıaz-Bossini and Moreno [
16
] present a valuable set of guidelines
to achieve accessibility in mobile interfaces for elderly users as a
result of a literature review study of academic works, standards
and best practices of mobile web design. Also as people with dis-
abilities are oen excluded from the society by not having access
to information through technology (and older people are included
in this group), the guidelines were compared with a survey of three
mobile Android apps, that were advertised as accessible, where all
applications showed accessibility problems.
Some eorts have been developed aiming to solve accessibility
gaps to visually impaired users. Although mobile design can be a
resourceful tool for such users by providing portability of several
features in one device, a literature review carried on the work of
Damaceno et al
.
[
12
] mapped 65 existing accessibility problems in
the interaction with mobile devices. e accessibility issues were
classied in seven groups of problems which contributed to the
establishment of a set of recommendations to improve accessibility.
Another disability that aects many people in the world is Color
Vision Deciency (CVD). Although approximately 8% of men and
0.5% of women in the world have some form of color blindness, this
diculty or inability recognize a certain color or to perceive color
dierences properly is not always apparent. As a result, it is not well
understood by developers and not applied on mobile development.
With the goal of improving accessibility in the mobile context, Zhou
et al
.
[
42
] proposes an adaptation of colors that are problematic to
CVD users while preserving properties of the original colors.
While other works present concerns with information access
for people with disabilities on mobile devices, Leitner et al
.
[
27
]
points out the positive motivation and business impact of providing
accessible services. In their study, the Grounded eory method
was used to analyze more than ten hours of interviews about ac-
cessibility. e main motivations found in the study are related to
economic aspects such as providing a product that is dierent from
the competition (by providing accessible content) and the possible
growth of customers.
On the work of Casadei et al
.
[
9
] an accessibility evaluation of the
Moodle Mobile application is presented in the context of Universal
Accessibility. In this work, the mobile application of the Learning
Management System Moodle was used by 21 participants without
disabilities in order to assess their emotional response to seven
interaction design paerns. Also, every user session was recorded
and analyzed in order to nd key points of diculties and problems
in the interaction of the users with the mobile application. e
results of the study shows accessibility barriers related to the design
paerns that were identied in the execution of the experiment and
the implications of these barriers in the students’ learning process.
While researching accessibility of mobile UI design paerns in
the academic literature we did nd some interesting publications
such as Meier et al
.
[
29
] that presents a comparison of ve location
search UI design paerns on mobile, but do not directly addresses
accessibility concerns. In this study, an experimentation with maps
and location information was performed in order to identify the
user’s preference in using one or another search interface design
paern in order to complete two tasks.
Also, de Barros et al
.
[
13
] aacks navigation and interaction of
older users with mobile platforms in an study that describes an
evaluation of a mobile application designed to promote the use
and prevent problems with the use of older users. e results of
usability tests of three mobile interfaces led the authors to propose
recommendations regarding the design of mobile applications for
older adults.
Similarly, Leit
˜
ao and Silva [
26
] investigates the optimal target
and spacing size between elements in mobile applications intended
for older adults. In the study, the gestures of tap and swipe were
used by 40 older users in two tasks on a mobile application. From
the data collected in the experimentation, the authors recommended
appropriate implementations of large tap and large swipe targets
for the use of mobile applications by the elderly.
e related works presented in this paper addresses various
aspects of accessibility in mobile devices in the context of older
adults, the visually impaired, Color Vision Deciency and even
universal accessibility. Although some studies use interface design
paerns, these studies are either focused on only one interface
element or on a specic deciency. However, the solutions and
recommendations proposed do not address a wide variety of UI
design paerns and do not account for what the active community
IHC ’17, October 2017, Joinville, Santa Catarina, Brazil blind-review
of designers and developers has studied and developed outside the
academic environment to address accessibility issues related to
mobile interface design paerns.
4 INVESTIGATING ACCESSIBILITY ISSUES
In this section we present the phases and steps of a virtual ethnog-
raphy study to investigate the problems and solutions involving the
use of UI design paerns in the mobile context by analyzing public
information available in forums and blogs that are not present on
academic articles. Besides the existence of few articles in the liter-
ature on accessibility of mobile UI design paerns, the choice of
using Virtual Ethnography was also made due to the importance of
knowing the practical experience of designers and developers who
have faced accessibility problems in the context of mobile interfaces
and proposed improvements to surpass these barriers.
As reported in the Section 3, Leitner et al
.
[
27
], de Barros et al
.
[
13
] and Leit
˜
ao and Silva [
26
] present contributions on mobile UI
topics. However, we’ve identied a shortage of works with current
design paerns, modern smartphones and a wider variety of design
paerns.
In this study we have considered only information related to
Android UI design paerns. us, we discarded iOS and other
mobile operating systems due to the fact that Android has 81.7%
of the smartphone market share in the world [
17
] and 93.2% of the
smartphones in Brazil are Android [8].
e virtual ethnography follows the same phases of the classical
one (planning, execution and analysis). Due to the fact that the
virtual ethnography success is directly linked with the online loca-
tions where the study is carried out, we conducted a pre-planning
phase in order to collect relevant references to be investigated.
4.1 Pre-planning
e identication and choice of sources of information with trust-
worthy and relevant data is a crucial requirement for the success of
any virtual ethnography study. Because of this, we decided to ques-
tion active designers and developers in the community about which
places they used to gather information about mobile accessibility
and design paerns.
In order to achieve this, a quick survey was created in Portuguese
and English with questions about experience working with mobile
technologies and accessibility (so that we could beer understand
the participants of the survey) and a last question where any source
found relevant to the topics should be informed.
e survey was shared online during three weeks in November
2016 and reached a total of 44 participants (designer and developers)
from 14 countries (Brazil, Colombia, Chile, Germany, USA, Ireland,
Japan, Spain, Portugal, France, England, Mexico, Peru and Italy).
Although the participants of the survey were anonymous, some
professional questions were asked and we could nd that 56% of
the 44 participants were developers (which means that the survey
reached a similar amount of designers and developers). Also, 32%
of the developers and 58% of the designers have between 5 to 10
years of experience working in their profession and 63% of all
participants ”knew” or ”knew a lot” about Mobile UI Design Paerns
and Accessibility.
e survey resulted in 72 websites, forums and blogs informed
by the participants. However a preliminary analysis was done and
all sources were accessed and searched in a process that resulted in
54 sources being eliminated due to several reasons: some addresses
or names were not found, other sources were documentations or
had focus on iOS development or nothing on mobile accessibility
and design paerns was found. Which resulted in the selection
of 18 information sources that were used in the study and will be
reported in analysis section.
Aer gathering sources where the information would be col-
lected, we proceeded with the planning step.
4.2 Planning
In the literature, Hine [
24
] presents a set of dilemmas and prac-
tices of online ethnography that includes several characteristics of
ethnographic studies in virtual communities as the participation
role of the researcher and the duration of the ethnography.
While Sharp et al
.
[
38
] compiles ve general dimensions of ethno-
graphic studies in Soware Engineering containing key topics to
help the researcher in the planning and execution of the ethnogra-
phy. ese denitions together are used to advise the researcher
about key points of an ethnographic study. However, in our study
we have identied the urge of two new dimensions that are spe-
cic for ethnographic studies in virtual communities. us, we
added these new dimensions necessary to plan and execute a virtual
ethnography: the
historical period
(Section 4.2.6) and
storage of
volatile information
(Section 4.2.7). ese new dimensions are a
contribution of this work.
Before starting to search the web for information regarding
mobile accessibility and interface design paerns accessibility, we
dened several aspects of our virtual ethnographic study in order
to give the directions of the investigation. In the sections below,
we present 7 dimensions that compose the ethnography protocol.
4.2.1 Participation role. e rst aspect that needs to be ad-
dressed is the role of presence in the ethnographic study. Due to
the nature of our research goal (we wished to examine public in-
formation available in the Internet), we decided to adopt the role
of observer. Also, as the role of observer does not directly inter-
acts with the studied subjects, the researcher’s identity was not
informed to the community.
Another important aspect of ethnography is the ethical position
of the researcher regarding the anonymity of the participants and
the informed consent and age. Forum and blog users, even of public
content, oen assume that anonymity is secured on the Internet
and that information and opinions posted online will not be traced
back to their real identities [39].
Although all information collected was public and shared in the
Internet without restrictions of reproduction and use, it was decided
to maintain the subject’s identity intact by not sharing citations
linked to a username. Also the use of informed consent and age
was discarded as each online platform (blogs, forums and sites)
already require its users to agree with terms of use and disclosure
of information.
4.2.2 Duration. e duration of the study cannot be foreseen
by the researcher, however, the researcher can trace a plan of work
Investigating accessibility issues of UI mobile design paerns IHC ’17, October 2017, Joinville, Santa Catarina, Brazil
that determines the time spent in the eld site and doing data
analysis. For this research we dened to take an approach where
the analysis of data would be done aer each source had been
completely scanned.
4.2.3 Location. As a result of the initial analysis of the academic
literature it was decided to search blogs, forums and other virtual
references collected from researchers, developers and designers.
4.2.4 Theoretical Underpinnings. e theoretical underpinning
adopted in this study came from our previous work on accessibility
of mobile design paerns [
9
] and from a technical report that as-
sessed a set of design paerns in terms of usability and accessibility
barriers for the web [41].
4.2.5 Ethnographers intent. e intent of this research was to
identify the practice of developers and designers regarding mobile
UI design paerns and understand issues faced in the development
of mobile application and solutions that are proposed.
4.2.6 Historical Period. is unique characteristic of virtual
ethnography studies was identied during the planning phase
where we understood that we had to dene a period in which
the information would be accepted in the research. As we oen
deal directly with technology on HCI, it’s important to have in
mind the relevance, popularity and usage curve of the technology
in order to select a space of time where information about this tech-
nology will be relevant. Because of the rapid changes on mobile
design and mobile platform in the last years, we decided to consider
articles or conversations that were published aer 2012.
4.2.7 Storage of Volatile Data. Online information is volatile
[
39
] and as ethnography in virtual communities oen deals with
historical data, it’s important to store and catalog such information
in order to preserve its integrity: a blog post that was accessed
and analyzed two weeks ago may not be available anymore or may
have been edited since then.
In this study we decided to adopt the strategy of redundancy
when storing articles and forum conversations by generating a
.png image of the entire page with the help of a plugin for Chrome
browser in order to maintain the original state of the page and two
other copies of the information in .pdf les (one generated with
the browser’s pdf saving tool and another with a special plugin
that generates ready for print .pdf les) alongside with a .txt le
containing the URL of the page.
4.3 Execution
Having our virtual ethnography planned with a well dened pro-
tocol and the 18 information sources selected (Subsection 4.1), the
next steps were (i) the exploration of the information sources search-
ing for relevant data, (ii) data selection and (iii) data storage. is
process is composed by these three steps in a cyclic iteration over
each one of the sources of information.
is process consisted of 18 cycles (one for each information
sources) where the rst cycle started on December 2016 and the
last cycle was completed only 14 weeks later.
4.3.1 Source Exploration. e rst step consisted of accessing a
source and searching for content that was valuable to the study. In
order to do this, we had to create search queries that would express
the topics we wished to nd. As many web pages, all 18 sources had
content and title based search engines with no sophisticated search
controls, which means that we had to create a set of expressions in
plain text and search for each expression.
e denition of these search strings is not random, the re-
searcher shall have previous knowledge of the topic studied and
know which expressions are oen used to refer to each specic
topic. In our study, as we reached professionals from several coun-
tries in the survey and obtained sources in three dierent languages:
Portuguese, English and Spanish, we had to adapt our search strings
to the correct expressions used in each language.
e list of search strings used in this research consisted of 10 ex-
pressions and small variations of these expressions: Mobile Design
Paerns, Mobile Accessibility, Design Paerns, Inclusive Mobile,
Inclusive Design Paerns, Android Accessibility, Android Design
Paerns, Mobile UI, Mobile UI Accessibility, UI Design Paerns.
4.3.2 Data Selection. e data selection step consisted of ac-
cessing each information source and exploring that source by using
the search engine with the search strings.
Each source was then explored by accessing every blog post or
forum discussion returned by the search query and analyzing its
content in order to evaluate its relation and relevance with the topic
and also the importance and value of the raw content, links to other
sources and comments.
As dened in our protocol on the planning phase, we selected
only information that had been published aer 2012 and that were
relevant to the research.
4.3.3 Data Storage. As we had dened on Section 4.2, any doc-
uments found relevant in the Data Selection step would be saved
in .png and .pdf formats in order to preserve the original content
from the date it was visited and collected.
However it is also necessary to establish a set of criteria to catalog
and organize all documents. Usually in a classical ethnographic
study where interview transcripts are analyzed, the researcher
organizes the transcripts by subject and date.
In our study we decided to adopt a similar approach adapted to
the nature of Internet information: we decided to create a structure
of les containing domain folder, sub-domain folder (if present) and
publication title. With this approach we were able to organize all
documents and identify its origin, also, we chose to not catalog the
publications by date because all of them were published in the last
5 years and because this information was not considered relevant
to the study.
Table 1 shows the information sources which were explored and
the number of publications gathered them.
Aer analyzing all 18 information sources, we had selected 127
publications. e distribution of publication by domain was not
uniform. Some sources as medium.com and smashingmagazine.com
presented a greater concentration of documents extracted and this is
explained by the fact that these pages publish articles from dierent
authors, thus having higher publication rate compared to other
sources.
IHC ’17, October 2017, Joinville, Santa Catarina, Brazil blind-review
Table 1: Number of publications by Domain (Information
Source)
Domain No. publications
medium.com 46
smashingmagazine.com 21
nngroup.com 12
stackexchange.com 10
reddit.com 9
creativebloq.com 7
uxpin.com 6
torresburriel.com 2
grihotools.udl.cat 2
uxmovement.com 2
usabilitygeek.com 2
bradfrost.com 2
androiduipaerns.com 1
visualhierarchy.co 1
lmjabreu.com 1
mobile-ui-design.ionicthemes.com 1
linkedin.com 1
jamesarcher.me 1
4.4 Analysis
By the end of the second phase of the virtual ethnography, 127 blog
articles and forum discussions had been collected and stored as
described in Section 4.3. e data analysis method chosen for this
research was the qualitative research method Grounded eory
(GT) [40].
ere are dierent methodologies with small dierences of how
to conduct Grounded eory research as the classical denition
[
11
] and more recent and practical approaches [
10
]. In this study,
we decided to follow a more practical form of Grounded eory
introduced by Charmaz [
10
] that present a structured guide for
doing research.
For the Grounded eory approach that we have chosen for this
study, there are three main steps that are the open coding, axial
coding and selective coding. On the rst step, all data from the
127 documents was analyzed in order to assign codes to particular
chunks of text that were pertinent to the research.
ere is more than just one method for coding. One popular
coding technique is called word-by-word coding. In this technique
each word (or small groups or words) receives a code in the open
coding step and it’s usually used on extremely dense documents.
Another technique analyzes and assign a code to each line or each
sentence, called line-by-line [
10
]. e coding technique that we
adopted in this GT study is called incident-to-incident coding. In
this approach, any amount of text in any location of the document
can be coded. Incident-to-incident coding is mostly used on docu-
ments that have low density, that is, documents that are not derived
from direct interviews and may contain portions of text that are
not relevant to the research as they address other subjects.
e second step, called Axial Coding, reorganizes the data and
established initial connections or relationships between the codes.
is means that new codes can be created and also that existing
codes can be deleted or merged into new ones. While on the rst
coding step (open coding) 243 codes were created, by the end of
the second step 189 codes were le.
In the third coding step, Selective Coding, the codes created
on the two previous steps were rened and analyzed in order to
identify main subjects in which the codes would be categorized. In
this step, the 189 codes were organized in 27 categories.
Lastly, but of utmost importance, several memos were created
based on the discoveries that were made from the coding process
and relationships that were found between the codes. is resulted
in 21 memos that detail several accessibility problems cited on the
documents identied in the virtual ethnography. ese memos also
contain important thought and conclusions about possible solutions
that might improve the accessibility of content and design paerns
on mobile devices.
Table 2: Relationship connectors on Nvivo 11
Symbol Relationship Name
−→
Aone-way relationship can be
used to demonstrate a relation-
ship between items which has a
denite direction..
Leads to
An associative relationship can
be used to demonstrate that
items are in some way aliated.
Associated with
←→
Asymmetrical relationship
demonstrates a two-way
activity between the items
Is related to
In this research, the qualitative research soware Nvivo 11 [
36
]
was used. Also, in this study all three coding steps from the GT
method were used. Once each category is dened and all codes are
rened, it’s possible to determine relationships between categories
and between codes. ese relationships are dened with the use
on connectors that are provided by Nvivo 11. ese relationship
types are presented on Table 2.
Aer analyzing the data through the Grounded eory research
method with the creation of all needed codes and memos, it was
possible to beer understand the problems faced by developers
and designers in the eld of mobile accessibility of design paerns
and also learn from the creative solutions proposed to some of the
accessibility issues. e ndings of our study are detailed in next
section.
5 FINDINGS AND DISCUSSION
From the GT in the data analysis phase a total of 189 codes were
created and organized in 27 categories with 21 Memos. Table 3
presents a summary of the three most coded design paern cate-
gories ordered by number of codes. In this table, the rst column
(Category) presents the name of the category of codes, while the sec-
ond column contains the number of sources of information where
this category was coded. e third column contains the number
of times that a code from the category was coded. e column
Description of Category provides a brief description of the category
containing key information to beer understand how the subject
Investigating accessibility issues of UI mobile design paerns IHC ’17, October 2017, Joinville, Santa Catarina, Brazil
Table 3: Summary of the most frequent code categories of UI design patterns
Category Sources Codes Description of Category Accessibility Issues
Navigation 28 130
On mobile design there are basically
two UI design paerns that are most
used for mobile navigation: e Drawer
Menu (oen called hamburger menu)
and the Tab Navigation.
While the drawer menu oers the possibility of several navi-
gation options that don’t occupy screen space (only when the
navigation buon is triggered) it hurts the discoverability of
content.
Users who access applications drawer menus use less the menu
options because they either forget that the menu exists or do
not recognize that there is a hidden menu under an icon. e
tab navigation on the other hand provides navigation options
that are always available to the user, thus improving the discov-
erability. However, the navigation is limited to a maximum of 5
items, which may not be enough for some applications.
Input 14 69
Input is any kind of UI element that is
used to insert information from the user
to the application. Usually, inputs elds
are used in forms.
As there are several types of input elds, there are many issues
regarding the accessibility of this paern.
One serious concern of almost any kind of input is the use on
placeholders to display information. Many implementations use
only placeholders to inform the user about the eld’s purpose
in order to save screen space.
e problem is that placeholders are, in most scenarios, not
recognized by screen readers for example.
List and
Pagination
6 44
is category aggregates concerns with
appropriate paerns for pagination on
mobile devices.
Also, it contains concerns about the ac-
cessibility of displaying information in
form of lists.
On mobile devices, one paern that is widely used is the innite
list. is paerns replaces the pagination with a (seeming)
innite list of items where new items are loaded when the user
reaches the end of a set of items.
Although this approach may seem interesting because the users
do not need to click on any buon to load new items, it has
serious problems with engagement and location as users do not
have any notion of progress while using this paern.
aects the accessibility of mobile applications. e last column
contains text snippets of codes and relevant accessibility issues
reported and identied in the virtual ethnography study.
Although the GT analysis resulted in 27 general categories (with
12 categories being UI design paerns), there is not enough space
to detail each of those categories in this paper due to the limited
number of pages. us, the complete list with all categories is
available at BLIND REVIEW
1
. In this paper, we spotlight three
categories, which represent UI mobile design paerns, that were
most coded.
In the end, the categories that resulted from the GT method
consisted of UI Design Paerns and also other aspects related to
accessibility. For example, the most coded category is Navigation
with 130 codes where this category contains citations related to
navigation mobile UI elements such as hamburger menu and tab
navigation.
Other categories that are not UI design paerns were also identi-
ed in the study but will not be explored in this paper. An example
of category is Color. Although color is not a design paern, it aects
the accessibility of any UI element. us, these other categories are
also important and relevant for mobile accessibility.
In the following sections we will present the ndings from the
three most frequently coded paerns that can be seen in Table 3.
Figures 1, 2 and 3 are extracted from NVivo.
1e link was omied following the recommendations of blind-review.
5.1 Navigation
Figure 1: Relationships of category Navigation
Based on the information of the paern (see Table 3), the Navi-
gation category was present on 28 documents with a total of 130
coded sentences. us, navigation is the most frequent subject
found in our research.
From the analysis of the codes identied for ”Navigation”, it was
uncovered that only two mobile navigation design paerns were
IHC ’17, October 2017, Joinville, Santa Catarina, Brazil blind-review
discussed in the 28 documents where the theme was found: the
navigation drawer (or hamburger menu) and the tab navigation.
In Figure 1 we can see that the Navigation category has relation-
ships with other ve categories that are Hamburger Menu,Tab Nav,
Action Bar,Gestures and List and Pagination. ese relationships are
important as they evidence that navigation accessibility problems
are aected or aect the interaction with mobile devices through
gestures for example.
Also, the navigation category has 7 child nodes represented by
circles lled in blue, these are some of the codes that compose this
category. Although the soware used to in the GT method (NVIVO
11) doesn’t allow to show, the child nodes Hamburger Menu and
Tab Nav are also subcategories of navigation and contain 5 and 7
child codes respectively. Similarly, navigation contains a memo
containing key information about accessibility issues on mobile
applications and the subcategories Hamburger Menu and Tab Nav
also contain memos reporting specic concerns about each design
paern that are not shown on Figure 1 due to limitations on the
qualitative data analysis soware.
From the analysis of the codes in the category Navigation and
its relationships, it’s clear that the concern with accessibility in
this category is focused on general navigation interaction with
mobile interfaces. It’s then possible to conceive recommendations
to improve or implement accessibility for general navigation in
mobile applications.
When dealing with complex processes on mobile application,
many times there is a big number of steps that the user must follow
in order to complete a task or achieve a goal. Although it may seem
more organized and appropriate to split actions in multiple steps,
on multiple screens, this segmentation of information requires
the memorization of previous actions. e problem is that people
with cognitive deciencies may struggle to remember previous
actions or information as they oen suer from short memory loss.
Also, elderly users also have problems locating themselves and
remembering something they did only moments ago. Even users
that do not suer from any disability may have diculty following
actions that are split in multiple screens when using the mobile
device on busy environments where the user must be aware of its
surroundings (in a bus for example).
One simple but eective solution for improving the general nav-
igation experience of users and provide accessible feedback to any
user is to use breadcrumbs or hints of some sort that inform the
current location. Although breadcrumbs are a common practice on
web development, it’s rarely used on mobile. With the information
of the current location, older users that oen get lost on mobile
applications and are insecure about how to undo some action for
example, will be able to understand beer where they are. However,
the developer or designer must understand that, by providing more
information, more space is occupied and less space is available for
the content. us it’s important to nd the appropriate place to
display such information, as for example an action bar or a simple
eld above the page title.
ere are also other concerns with mobile navigation accessi-
bility that are specic to implementations of mobile navigation.
e sections below provide more detailed conclusions about the
subcategories Hamburger Menu and Tab Navigation.
5.1.1 Hamburger Menu. Hamburger menu is a kind of navi-
gation UI design paern mostly used on mobile applications that
consists of presenting navigation options that are initially hidden
and that can be triggered by the click of a buon.
e problem with this approach is that the navigation is hidden
from the user that needs to have previous knowledge to understand
that the menu can be accessed by clicking a buon. e menu icon
has low information scent, and, even with a label called ”Menu”,
users may still not use the navigation as they do not know which
options are available and don’t even click the buon.
In the documents analyzed several use cases were presented
about big companies that initially adopted the hamburger menu, but
soon changed to Tab navigation. Some companies as Facebook and
BBC used the drawer navigation, but due to the low discoverability
that this type of menu provides, they identied that users had
problems locating sections and menu options.
5.1.2 Tab Navigation. e main advantage of the tab bar is that
it doesn’t hide the navigation options, they can be available all
the time and the user has easy access to the navigation bar with-
out needing to open a drawer in order to locate which navigation
options are available.
Another important contribution of tab navigation is that the
tabs communicate the user about the current location by visually
featuring the current position in the related navigation option. is
simple feature makes the user capable of visually understanding
the current location and also gives more control to visually im-
paired user who use screen readers and also to users with cognitive
deciencies who struggle with short term memory.
Although tab navigation may seen the perfect navigation ele-
ment for mobile devices, it can hold only up to 5 navigation options
in order to be able to t in the screen with appropriate font size. In
these situations, there are two approaches to still use a navigation
tab bar: the rst approach is to add a last option called “more” or
other appropriate label that will open a navigation drawer.
e second approach applies the use of a horizontally scrollable
navigation tab bar. In this approach, the navigation bar can hold
more options while keeping the optimal touch-target size. e
downside of both approaches is that they still have discoverability
issues.
5.2 Input
Inputs were probably one of the rst interface interactions that
were designed in the early days of graphical computing and they
have not changed much ever since.
In Figure 2 we can see all relationships and codes related to the
Input category. One of the accessibility concerns that were most
present in this category is the correct use of placeholder and labels
on input elds.
e use of placeholders alone saves the designer space for more
elements. However, when an input eld does not have a label, the
hit area of that element is reduced. is is harmful for older users
or users with motor disabilities that may limit they movement, for
instance.
In addition, a placeholder disappears when the user starts typing,
which instantly removes all the context of the input eld. is
problem is also harmful for people with cognitive deciencies as
Investigating accessibility issues of UI mobile design paerns IHC ’17, October 2017, Joinville, Santa Catarina, Brazil
Figure 2: Relationships of category Input
they may struggle to work with short term memory. Placeholders
are presented with low contrast to the background, which may
spoil the experience of users with visual impairments or users with
situational disabilities caused by screen reection in the screen.
ere is no solution for this problem because if the contrast is
higher, the user may confuse the placeholder with actual content
of the input, for example.
A solution for most of these problems is the appropriate use of
label and placeholder together. Labels are not help texts and thus
should be succinct, short and descriptive so that users can quickly
identify what information is required. Also, labels must present
the necessary information for the user, for example: on a login
form where the user can use a username or an email to access the
account, the label must contain “username or email”.
Placeholders that are used with labels must be meaningful and
contain information about the format expected in the input eld or
other necessary information. Both on placeholders and on labels,
the text should never be in all caps, as it’s more dicult to read and
harder to quick scan.
5.3 List and Pagination
From Figure 3 we can see that the category ”List and Pagination” is
related to the ”Navigation” category which is explained by the fact
that the user navigates through the items in a list.
On web desktop list navigation is usually implemented with
pagination controls at the boom or top of the page. With these
controls, the user has variety of actions to choose as skipping pages,
jumping to the end or returning to the rst page.
On mobile applications, a UI design paern that has been widely
used for list navigation is the innite list. In this paern, the user
does not have to click on any buon in order to load more items,
instead, when reaching the boom of the page, a new set of results
is automatically loaded and displayed to the user. is process
continues until the end of the list is reached.
Figure 3: Relationships of category List and Pagination
Although the innite list may seem like a good alternative of min-
imalist and practical design, it raises serious accessibility concerns.
On situations of low bandwidth, that are common when using mo-
bile devices outside a Wi-Fi network, innite list may present long
waiting times while loading for new items to be displayed. is
long wait may cause the user to think that the application is not
responding or think that the list has reached its end. Also, visually
impaired users may not understand what is happening because the
screen reader may not be able to correctly inform the user that the
application is loading new information.
One of the major problems with innite scrolling is that it does
not provide any information regarding the position on the list.
e same way that the lack of information about location in the
navigation of mobile applications (Section 5.1) can cause serious
problems for people with dierent disabilities (and even people
without disabilities). It also may cause confusion and uncertainty
for many users that are not sure of their current position, to how
they got there and how to share or save the position of an item in
the list for example.
Also, on some implementations, the innite list is sorted by date
or alphabetically. is situation causes a new problem where it’s
extremely dicult and tiresome to nd an item at the end of the
list as it’s not possible to jump to the end of the list. Many times,
list implementations do not provide any kind of ltering options,
which contributes to these issues.
With these issues, it was found that, in terms of accessibility,
the beer list navigation paern is still pagination, even on mobile
devices (of course that with fewer controls due to the small view-
port). Also, it’s important to provide ltering and ordering options
in order to lower the number of items in the list.
Although in theory pagination requires more clicks and actions
from the user, it provides more control. Also, as pagination is
controlled by buons, screen readers are able to correctly parse the
elements in the page.
6 CONCLUSIONS AND FUTURE WORKS
In this paper we presented an investigation of issues on the acces-
sibility in mobile UI design paerns. Guided by an ethnographic
IHC ’17, October 2017, Joinville, Santa Catarina, Brazil blind-review
study conducted in 18 virtual communities of mobile design ac-
cessibility, we found out 127 occurrences of accessibility issues
organized in 27 categories.
From the qualitative data analysis of documents collected in
the ethnographic study using Grounded eory, many accessibility
concerns were identied with the development and design of mobile
interfaces, with special aention to the use of UI design paerns.
It was then possible to propose improvements and modications in
these paerns that should improve the accessibility of applications
by beer understanding the necessities of users who suer with
some disability and struggle with the use of mobile application.
is paper gives emphasis on the accessibility analysis of three
mobile interface design paerns that were the three paerns most
commonly identied in the qualitative data analysis.
As future work, the authors intend to extend the knowledge on
accessibility problems of mobile UI design paerns and propose
beer design for these paerns in order to improve their accessibil-
ity. In addition, these recommendations shall be tested with real
users in order to validate the accessibility of mobile applications
built with these paerns.
ACKNOWLEDGMENTS
e authors would like to thank blind for the nancial support.
REFERENCES
[1]
Shadi Abou-Zahra, Judy Brewer, and Shawn Lawton Henry. 2013. Essential
Components of Mobile Web Accessibility. In Proceedings of the 10th International
Cross-Disciplinary Conference on Web Accessibility - W4A ’13. 1–4. hps://doi.
org/10.1145/2461121.2461138
[2]
George Allan. 2003. A critique of using grounded theory as a research method.
Electronic journal of business research methods 2, 1 (2003), 1–10.
[3]
Michael Angrosino. 2007. Doing ethnographic and observational research. Sage.
[4] Melanie Birks and Jane Mills. 2015. Grounded theory: A practical guide. Sage.
[5]
Tom Boellstor. 2012. Ethnography and virtual worlds: A handbook of method.
Princeton University Press.
[6]
Danah Boyd and Jerey Heer. 2006. Proles as conversation: Networked identity
performance on Friendster. In System Sciences, 2006. HICSS’06. Proceedings of the
39th Annual Hawaii International Conference on, Vol. 3. IEEE, 59c–59c.
[7]
Giorgio Brajnik. 2008. Beyond conformance: the role of accessibility evalua-
tion methods. In Web Information Systems Engineering–WISE 2008 Workshops.
Springer, 63–80.
[8]
Lucas Carvalho. 2017. Android cresce no Brasil e aumenta distancia
para iOS e Windows Phone. (2017). hps://olhardigital.com.br/noticia/
android-cresce- mas-ios- e-windows-phone- perdem-mercado- no-brasil/68023
[9]
Vitor Casadei, Luciana Zaina, Eduardo Pinheiro, and Toni Granollers. 2016.
Accessibility Evaluation of Design Paerns on Moodle Mobile. In Brazilian
Symposium on Computers in Education (Simp
´
osio Brasileiro de Inform
´
atica na
Educac¸˜
ao-SBIE), Vol. 27. 688.
[10] Kathy Charmaz. 2014. Constructing grounded theor y. Sage.
[11]
Juliet M Corbin and Anselm Strauss. 1990. Grounded theory research: Procedures,
canons, and evaluative criteria. alitative sociology 13, 1 (1990), 3–21.
[12]
Rafael Jeferson Pezzuto Damaceno, Juliana Cristina Braga, and Jes
´
us Pascual
Mena-Chalco. 2017. Mobile device accessibility for the visually impaired: prob-
lems mapping and recommendations. Universal Access in the Information Society
(2017), 1–15. hps://doi.org/10.1007/s10209- 017-0540-1
[13]
Ana Correia de Barros, Roxanne Leit
˜
ao, and Jorge Ribeiro. 2014. Design and
Evaluation of a Mobile User Interface for Older Adults: Navigation, Interaction
and Visual Design Recommendations. Procedia Computer Science 27 (2014), 369 –
378. 5th International Conference on Soware Development and Technologies
for Enhancing Accessibility and Fighting Info-exclusion, DSAI 2013.
[14]
Cleusa de F
´
atima Granao, Marynea A. P. Pallaro, and S
´
ılvia Am
´
elia Bim. 2016.
Digital Accessibility: Systematic Review of Papers from the Brazilian Symposium
on Human Factors in Computer Systems. In Proceedings of the 15th Brazilian
Symposium on Human Factors in Computer Systems (IHC ’16). ACM, New York,
NY, USA, Article 21, 10 pages. hps://doi.org/10.1145/3033701.3033722
[15] IBGE Instituto Brasileiro de Geograa e Estat´
ıstica. 2010. Censo 2010.
[16]
Jos
´
e-Manuel D
´
ıaz-Bossini and Lourdes Moreno. 2014. Accessibility to mobile
interfaces for older people. Procedia Computer Science 27 (2014), 57–66.
[17]
Laurence Goasdu and Amy Ann Forni. 2017. Gartner Says Worldwide Sales
of Smartphones Grew 7 Percent in the Fourth arter of 2016. (2017). hp:
//www.gartner.com/newsroom/id/3609817
[18]
Martyn Hammersley and Paul Atkinson. 2007. Ethnography: Principles in practice.
Routledge.
[19]
Simon Harper, Peter iessen, and Yeliz Yesilada. 2014. Research Report on Mobile
Web Accessibility. Technical Report. W3C. hp://www.w3.org/WAI/RD/2012/
mobile/note/ED-mobile
[20]
Shawn Lawton Henry. 2006. Understanding Web Accessibility. Apress, Berkeley,
CA, 1–51. hps://doi.org/10.1007/978- 1-4302-0188- 5 1
[21]
Shawn Lawton Henry, Shadi Abou-Zahra, and Judy Brewer. 2014. e Role of
Accessibility in a Universal Web. In Proceedings of the 11th Web for All Conference
(W4A ’14). ACM, New York, NY, USA, Article 17, 4 pages. hps://doi.org/10.
1145/2596695.2596719
[22]
J Heumann. 1998. Keynote address to Microso employees and experts on
disabilities and technologyl. (1998). Redmond, Washington, February 19, 1998.
[23] Christine Hine. 2000. Virtual ethnography. Sage.
[24]
Christine Hine. 2008. Virtual ethnography: Modes, varieties, aordances. e
SAGE handbook of online research methods (2008), 257–270.
[25]
Steven Hoober and Eric Berkman. 2011. Designing mobile interfaces. O’Reilly
Media, Inc.
[26]
Roxanne Leit
˜
ao and Paula Alexandra Silva. 2012. Target and Spacing Sizes
for Smartphone User Interfaces for Older Adults: Design Paerns Based on an
Evaluation with Users. In Proceedings of the 19th Conference on Paern Languages
of Programs (PLoP ’12). e Hillside Group, USA, Article 5, 13 pages.
[27]
Marie-Luise Leitner, Christine Strauss, and Christian Stummer. 2016. Web acces-
sibility implementation in private sector organizations: motivations and business
impact. Universal Access in the Information Society 15, 2 (2016), 249–260.
[28]
Ricardo Leme, Luciana Zaina, and Vitor Casadei. 2015. Um estudo da intera
c¸˜
ao do
p
´
ublico da terceira idade com o Facebook em dispositivos m
´
oveis: identica
c¸˜
ao e
deni
c¸˜
ao das personas. In Proceeding of the 16th Brazilian Symposium on Human
Factors in Computing Systems IHC ’15. Brazilian Computer Society, 53–62.
[29]
Sebastian Meier, Frank Heidmann, and Andreas om. 2014. A Comparison
of Location Search UI Paerns on Mobile Devices. In Proceedings of the 16th
International Conference on Human-computer Interaction with Mobile Devices
& Services (MobileHCI ’14). ACM, New York, NY, USA, 465–470. hps://doi.
org/10.1145/2628363.2634216
[30]
Erik G. Nilsson. 2009. Design paerns for user interface for mobile applications.
Advances in Engineering Soware 40, 12 (2009), 1318 – 1328. Designing, modelling
and implementing interactive systems.
[31] World Health Organization. 2011. World report on disability. 19–46 pages.
[32]
Afra Pascual, Mireia Ribera, Toni Granollers, and Jordi L. Coiduras. 2013. Impact
of accessibility barriers on the mood of blind, low-vision and sighted users.
Procedia Computer Science 27, 2013 (2013), 431–440. hps://doi.org/10.1016/j.
procs.2014.02.047
[33]
Hans Persson, Henrik
˚
Ahman, Alexander Arvei Yngling, and Jan Gulliksen. 2015.
Universal design, inclusive design, accessible design, design for all: dierent
concepts—one goal? On the concept of accessibility—historical, methodological
and philosophical aspects. Universal Access in the Information Society 14, 4 (2015),
505–526. hps://doi.org/10.1007/s10209-014- 0358-z
[34]
Jorge Ribeiro. 2012. Web Design Paerns for Mobile Devices. Master’s thesis.
Faculdade de Engenharia da Universidade do Porto. hp://repositorio- aberto.up.
pt/handle/10216/64860
[35]
Jorge Ribeiro and Miguel Carvalhais. 2012. Web Design Paerns for Mobile
Devices. In Proceedings of the 19th Conference on Paern Languages of Programs
(PLoP ’12). e Hillside Group, USA, Article 13, 48 pages.
[36] Lyn Richards. 1999. Using NVivo in qualitative research. Sage.
[37]
Ana Isabel Sampaio. 2013. Responsive Web Design. Master’s thesis. Escola de
Engenharia, Universidade do Minho.
[38]
H. Sharp, Y. Dirich, and C. R. B. de Souza. 2016. e Role of Ethnographic Studies
in Empirical Soware Engineering. IEEE Transactions on Soware Engineering
42, 8 (Aug 2016), 786–804.
[39]
Kevin F. Steinmetz. 2012. Message Received: Virtual Ethnography in Online
Message Boards. International Journal of alitative Methods 11, 1 (2012), 26–39.
[40] Anselm Strauss and Juliet M Corbin. 1997. Grounded theory in practice. Sage.
[41]
Yeliz Yesilada, Tianyi Chen, and Simon Harper. 2009. D3: Mobile Web Barriers for
the Barrier Walkthrough Method. Technical Report. School of Computer Science,
University of Manchester.
[42]
L. Zhou, V. Bensal, and D. Zhang. 2014. Color adaptation for improving mobile
web accessibility. In 2014 IEEE/ACIS 13th International Conference on Computer
and Information Science (ICIS). 291–296. hps://doi.org/10.1109/ICIS.2014.6912149
