Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona...

42
Designing a Digital Service for Users with Low Understanding of a Language A Case Study Fredrik ¨ Ostlund VT20 Master’s Thesis in Interaction Technology and Design, 30 credits Supervisor: Ole Norberg Extern Supervisor: Mattias Hagbard Examiner: Thomas Mejtoft

Transcript of Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona...

Page 1: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

Designing a Digital Service for Userswith Low Understanding of aLanguageA Case Study

Fredrik Ostlund

VT20

Master’s Thesis in Interaction Technology and Design, 30 credits

Supervisor: Ole Norberg

Extern Supervisor: Mattias Hagbard

Examiner: Thomas Mejtoft

Page 2: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

Abstract

During the last decades Europe has seen an increase in the numberof refugees coming to the region. When arriving in a new countryone faces several challenges to overcome. In Sweden, a lot of theservices the government provides regarding immigration, social in-surance, and health care are digital. This can on some occasionscreate a situation where the user of these services can not speakor write any of the languages available in the service. This thesisaims to investigate what aspects of user experience design thatcould support a user completing tasks and understanding contextin a digital setting even though the language might not be fullyunderstood. By looking at the aspects of cross-cultural design,design for illiterate users, and understanding icon design, a proto-type was developed and tested on users not literate in the languageused. The results showed that the users were able to complete thetask given, and understand the context.

Sammanfattning

Under de senaste artionderna har Europa sett en okning nar detkommer till antalet flyktingar som kommer till regionen. Naren kommer till ett nytt land star en infor olika utmaningar. ISverige erbjduer staten olika digitala tjanster for hjalp med so-cialtjanster, sjukvard och integration. Detta kan leda till situ-ationer dar anvandaren varken kan lasa eller tala nagot av detillgangliga spraken i tjansten. Det har arbetet syftar till attundersoka vilka aspekter av anvandarcentrerad design som skullekunna hjalpa en anvandare att forsta kontext och genomfora uppgifteri en digital miljo aven fast anvandaren inte forstar spraket tillfullo. Genom att titta pa aspekter som cross-cultural design, de-sign for analfabeta anvandare och undersoka ikon-design har enprototyp skapats och testats pa anvandare som inte kunde spraketsom anvandes. Resultatet visade att anvandare kunde genomforauppgiften de tillgavs, och forstod kontexten.

Page 3: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

Contents

1 Introduction 2

1.1 Problem Background 3

1.2 Objective 3

1.3 Limitations 3

1.4 Hypothesis 4

2 Background 5

2.1 Humblebee 5

2.1.1 HIVE 5

2.1.2 Case study description 6

2.2 User Experience 6

2.3 Information design 6

3 Method 8

3.1 Empathise 9

3.1.1 Literature Study 9

3.1.2 User Research 10

3.2 Define 10

3.2.1 Persona 10

3.2.2 Design Challenge 10

3.3 Ideate 10

3.4 Prototype 11

3.4.1 Low-Fidelity 12

3.4.2 High-Fidelity 12

3.5 Test 12

3.5.1 Test Case 13

3.5.2 Test Participants 13

4 Theoretical Framework 15

4.1 Icon design 15

4.2 Cross-cultural design 15

4.2.1 Challenges in cross-cultural design 16

4.2.2 Designing in a cross-cultural setting 17

Page 4: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

4.3 Universal design 17

4.4 Communication 17

4.4.1 Handling illiteracy in communication 18

4.4.2 Second Language Communication Strategies 18

4.5 Designing for Illiterate Users 19

4.5.1 Presenting information to illiterate users 19

4.5.2 Design guidelines 20

5 Related Works 22

6 Results 23

6.1 Understanding 23

6.1.1 User Research 23

6.1.2 Takes From Literature Study 23

6.1.3 Persona 23

6.2 Exploring 25

6.2.1 Low-Fidelity 25

6.2.2 High-Fidelity 25

6.3 Materialising 28

6.3.1 User testing 28

7 Design choices 29

7.1 Icons 29

7.2 Layout 29

7.3 Interaction 29

7.4 Text 30

8 Discussion 31

8.1 Result Discussion 31

8.2 Methodology Discussion 31

8.2.1 Research methodology 31

8.2.2 Testing methodology 32

9 Conclusion 34

9.1 Future Work 34

10 Acknowledgements 35

1

Page 5: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

1 Introduction

During the last decades the world has become more and more global. Today weare connected in many different ways across the globe, and people are moving tonew countries because of love, family, and conflicts among other things. Due toconflicts around the world, Europe is one of the regions that has seen an increasein refugees in the last number of years. In 2015, Sweden had its highest number ofasylum seekers ever, leading to the highest number of immigrants in the followingyear. In Sweden, an asylum seeker is counted as an immigrant once he or she hasreceived a residence permit. In 2018, people from Syria were the largest group ofimmigrants to Sweden1. Coming to a new country with a different culture andlanguage from what one is used to creates several challenges. Housing, finance,work, education, family, and language is just a few of the challenges one must facewhen integrating into a new country. To be able to integrate successfully one mustovercome several of these challenges. It has been found that one of the key factorsfor immigrants to be able to integrate successfully into a new country and its societyis language acquisition [1]. The Swedish government is providing language educationto immigrants through ”Swedish for immigrants”(SFI) 2 which purpose is to helpimmigrants establish themselves faster in the Swedish society.

Apart from the education provided by the government, it exists different kind ofdigital tools which aims to help the user to learn a new language, one example beingDuolingo3. Duolingo uses gamification [2] as a tool to help to learn a new languageto become more immersive for the user. As a refugee or immigrant, however, onemight not be familiar with using such tools. In today’s society the smartphone playsa central role, and most of us use our smartphone several times during the day fordifferent purposes and to complete different tasks. In fact, in Sweden 90% of thepopulation uses their phone to access the internet4. Even though the vast majorityof the population uses their phone to access the internet, everyone might not beable to use the essential mobile services used to access public transport, banking,etc. This could be due to a lack of understanding of the Swedish language, or thelack of experience with new technology. What could be done in terms of designand development to help these users both understand and be comfortable enoughto use these services? This thesis aims to understand and investigate what aspectsof user experience design that might help the user understand context and tasks inan application even though the language might not be understood fully.

1https://www.scb.se/hitta-statistik/sverige-i-siffror/manniskorna-i-sverige/invandring-till-sverige/ (Accessed 2020-02-06)

2https://www.skolverket.se/undervisning/vuxenutbildningen/komvux-svenska-for-invandrare-sfi/ (Accessed 2020-02-25)

3https://www.duolingo.com4Internetstiftelsen, Svenskarna och Internet 2019, https://svenskarnaochinternet.se/rapporter/svenskarna-

och-internet-2019/, (2020-03-25)

2

Page 6: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

1.1 Problem Background

As explained above, most Swedes use their smartphone every day to complete anumber of tasks. If a person for some reason is not able to use a smartphone , it canbe tough to be a part of the society we see today. If a person has come to Swedenfrom a different country, he or she might need to use Swedish digital services or toolsto complete different every day tasks. If that person is not proficient in the Swedishlanguage, it can become quite hard to complete these tasks since a lot of digitaltools mainly provide information in form of text. Could these users be helped byadapting the design of these applications so they can function even though the useris not proficient in the language in question? If so, these users could better navigatein the society, and complete the tasks that users who speak the language well can.

1.2 Objective

The objective of this Master Thesis is to investigate and analyse how user experiencedesign in a digital product could minimise a language gap for people that are newto the Swedish society.

1.3 Limitations

Some limitations have been defined to this thesis project to narrow down the scope.

Target group This thesis project will aim to target newly arrived women in Swedenliving in the outer areas of Gothenburg. The focus will be on women originatingfrom the middle east, in the ages of 18-35 years old.

Case The case will be limited to see if members from the target group can un-derstand context and complete simple tasks in an application. The designedapplication will not be complex due to the scope of the project. The designedprototype will be of a flow where the user is to book a doctors appointment fora specific symptom, and choose whether to meet in person or through a voiceor video call.

Testing The testing of the application will be made through qualitative testing ofprototypes. Since there is no coding competence available to utilize, the appli-cation will not be fully implemented in a digital environment.

General The result of the investigation will work as a general guide of how one canuse user experience design to minimise a language gap rather than going intospecific design details. Therefore it should be seen as support when designingin a cross-cultural setting.

3

Page 7: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

1.4 Hypothesis

The hypothesis for this investigation is that by the use of user experience design,icon design, and cross-cultural design users can understand context and completesimple tasks in an application regardless of language abilities.

4

Page 8: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

2 Background

This chapter will inform the reader about parts related to the thesis project, andgeneral information about the company Humblebee and the case that will be inves-tigated in collaboration with them.

2.1 Humblebee

Humblebee is a digital product and service studio based in Gothenburg, Sweden.Humblebee was founded in 2012 and are a part of Vinngroup1. They create greatservices for companies like Volvo Group, Hultafors Group, Molnlycke, Polestar andStena Metall. Humblebee evolves around a few common values. These are

• Diversity: Believing that diversity means different opinions leading to a betteratmosphere and fostering creativity.

• Equality: Everyone’s opinions is of equal importance at Humblebee and ev-eryone has the opportunity to develop.

• Respect: Working in collaborative teams and respecting colleagues and part-ners at all times.

• Co-creation: Believing that co-creation is the key to success. Humblebee putsthe right experts in a room and great things happen.

• Speed: Humblebee knows speed to get services launched is a differentiator forsmart companies and brings competetive advantage.

• No bullsh*t: Always aiming to build what’s needed, not always what’s askedfor.

Through these values, Humblebee harness technology to jumpstart growth, speed-up time to market and foster innovation2.

This Master Thesis will be written in collaboration with Humblebee at theiroffice in Gothenburg.

2.1.1 HIVE

HIVE is a student-driven lab at Humblebee, where students with different compe-tencies get together and solve a real problem. This Master Thesis is written inparallel with the HIVE project running during the spring of 2020.

1https://www.vinngroup.com2https://www.humblebee.se

5

Page 9: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

2.1.2 Case study description

Through HIVE, Humblebee has partnered with the organisation ”Tidigt Foraldrastod3”which helps newly arrived women to Sweden that either is, or are becoming a motherwith everything from being a support in the maternity care to contact with author-ities. During the spring of 2020, five students in the HIVE team will develop adigital product that is supposed to help these women. This master thesis is writtenin parallel with the HIVE project, which means that the target group and user re-search are shared. The case study aims to investigate how user experience designcan be used to help users understand the context and simple tasks in an applicationregardless of language abilities.

Through the HIVE project, some research has been made regarding the targetgroup of this thesis project. The research finds that the members of the target groupsee being able to talk the Swedish language as the number one aspect to be able tobecome a part of the Swedish society. The members, and persons that work withintegration and meet a lot of these women, mention that to be able to truly developin one’s language, physical meetings need to occur where one can talk Swedish. Ifthis aspect is not fulfilled, the learning of a new language will become harder.

2.2 User Experience

Norman and Nielsen [3] defines User Experience(UX) as:

”The first requirement for an exemplary user experience is to meet the exact needsof the customer, without fuss or bother. Next comes simplicity and elegance that

produce products that are a joy to own, a joy to use. True user experience goes farbeyond giving customers what they say they want, or providing checklist features.

In order to achieve high-quality user experience in a company’s offerings theremust be a seamless merging of the services of multiple disciplines, including

engineering, marketing, graphical and industrial design, and interface design.”

Hassenzahl [4] explains user experience design so that it stands for technology thatsuggests meaningful, engaging, valuable, and aesthetically pleasing experiences. Theshift from thinking communication experiences instead of mobile devices is a shifttowards working with user experience design, and it opens up a huge design spacefor possible devices to unfold from.

2.3 Information design

According to Sless, information designers create and manage the relationship be-tween people and information so that the information is accessible and usable bypeople, and they provide evidence that the information is accessible and usable to anagreed high standard [5]. The main goal of information design is clarity of communi-cation. By accurately designing, producing, and distributing a message, a designercan fulfil this goal. In the end, the message should be interpreted and understoodby members of the intended audience. These processes in information design are

3https://tidigtforaldrastod.se

6

Page 10: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

guided by principles, which are performed with the help of tools and influenced bythe social context in a specific situation [6].

7

Page 11: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

3 Method

This section will present the chosen method for investigating current research anddeveloping a design proposal for the given case.

This thesis has used the process of design thinking to produce a design proposal.The concept of design thinking does not come with a distinct definition. The conceptis used in lots of different situations, both theoretical and practical. Tom Kelleywrites about IDEO’s way of working with design. Through his literature, we get aview of how the company works with lessons in creativity and their design practicemeaning that everyone can work with design if one follows the practices of theirmodels [7].

For this thesis, the process which Gibbons [8] describes has been used. She de-scribes the process as an overall flow of understanding, exploring, and materialising,see figure 1. These three phases can be split up into the following phases:

Empathise In the first part of the process the objective is to understand the usersin terms of what they do, how they think, and what they want. This includesboth physical and emotional needs.

Define In the define phase the research made in the previous section is combined toget an understanding of where the users’ problems exist. The goal is to definethe challenges that lie ahead.

Ideate The ideate phase is where the idea generation takes place, based on theknowledge from previous parts of the process. The aim is to generate a widerange of solutions to the identified problems.

Prototype The intention of the prototype phase is to get answers to what willcreate a good solution and what will not. This is done by iteratively generatingdesign proposals starting with low-fidelity prototypes and ending up with high-fidelity prototypes that are as close to the solution as possible.

Test The testing phase means to return to the users to get feedback. Questionsto get answers to are ”Does the solution meet the users’ needs? ” and ”Hasit improved how they feel and think?”. This phase is also an iterative processgoing back and forth to the prototype phase.

Implement In the implementation phase the solution will come to life and end upas a real product that the users can use.

8

Page 12: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

Figure 1: Visualisation of the design thinking framework based on [8]

Below it will be explained what has been done in all of these phases throughoutthe project.

3.1 Empathise

This section explains what has been done in the empathise part of the design process.

3.1.1 Literature Study

The first phase in this study was to gain information and knowledge from existingliterature and previous work in the field. To achieve this an extensive literaturestudy was conducted. The study aimed to find information about what aspects totake into consideration when designing in a cross-cultural setting and what specifictypes of tools or methods that could be used. The target group has high writtenliteracy in some language, but for the most part their Swedish writing and readingskills are not so developed. Therefore research has been made into the aspect of de-signing for illiterate users. Several relevant articles, books, and reports were foundin the areas of user experience design, designing icons for digital products, designingfor illiterate users, and cultural aspects of design. This was accomplished by usingsources as the Umea University library 1 and Google Scholar 2. To find relevantinformation searches was done with phrases such as ”cross-cultural design”, ”lan-guage design”, ”designing icons”, ”understand context UX”, ”design illiteracy”, and”culture interaction”. The literature study resulted in the theoretical framework,see chapter 4.

1https://www.umu.se/bibliotek/2https://scholar.google.com

9

Page 13: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

3.1.2 User Research

To get a deeper understanding of the users a user research phase was performed. Theoverall goal of this research was to understand what pain points and experiences withdigital services the user might experience. The research was done by investigatingthe previous work the team at HIVE, explained in section 2.1.1 had done. The teamhad made interviews with women in the ages between 18-35 years old about theirlife in Sweden and how they are using digital services. The interviews were madeboth in a group and one to one.

Furthermore, experts within the field had been interviewed. This includedpersons that have been working with helping women establish in Swedish societythrough different initiatives focusing on language acquisition, cultural fit, and muchmore. The purpose of these interviews was to get an understanding of their expe-riences working with services and projects which will be used by women who havenewly arrived to Sweden.

3.2 Define

During the define phase, the goal was to define the design challenge rising from theresearch findings from the previous phase. The data from the empathise phase wereanalysed and clustered to be able to present them as insights. The insights providedknowledge about the users’ needs, wants, challenges, and opportunities. This wasboiled down into a persona.

3.2.1 Persona

To set a specific target group for the design to be evolved around, a persona wascreated. This was done by analysing the research done in the empathise phase,specifically looking into the most common background, living situation, and digitalexperience. Personas are a good tool to use in the design process since it offers afictional character that still is realistic [9].

3.2.2 Design Challenge

After creating a persona, all theories gained in the literature study were collectedinto a theoretical framework that was used to create a design that fulfills the needsof the target group.

To investigate if the theories gained could be applied properly in a designedprototype, a context was to be decided. The prototype to test was a service wherethe user would be able to book an appointment with a doctor. This was chosenbecause it is a limited scope, and something the target group could find beneficial.

3.3 Ideate

Based on the knowledge gained from the previous step, the objective of the ideatephase was to generate ideas that would solve the problems defined in the definephase.

10

Page 14: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

Figure 2: Early sketch of persona using pen and paper

When generating ideas in this process, the aim was to generate as many ideasas possible, preferably ideas that are quite differentiated from each other [8]. Inorder to achieve this, a set of association cards from the company Pollen3 were usedto inspire the mind to be more creative and able to come up with a set of moredifferentiated ideas, see figure 3.

3.4 Prototype

When a better understanding of how the users behave and think, and how onemight use design to overcome language disabilities, the solution could be designed.To present the design two different prototypes were created; low-fidelity and high-fidelity. Prototypes are often used as a tool in the design process to simulate thefinished product in order to examine how the content, aesthetics, and interactionaffects the user. Both low-fidelity prototypes and high-fidelity prototypes are a goodway to find usability problems early in the development process [10].

3https://www.pollenideas.com

11

Page 15: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

Figure 3: Examples of association cards from the company Pollen.

3.4.1 Low-Fidelity

To be able to visualise the ideas generated in the ideate phase, simple low-fidelitysketches were created using pen and paper. These sketches were used as a groundto choose which ideas that could work best in the high-fidelity prototype. Since thelow-fidelity sketches only served as a foundation for ideas and visualisation, no usertests were performed using these prototypes.

3.4.2 High-Fidelity

To be able to investigate the ideas created in the previous steps, a high-fidelity pro-totype was created. This prototype contained graphical user interface (UI) elementsand interaction schemas. This to be able to test the prototype in a setting as similarto a real usage situation as possible. Both the design and interaction schema of theprototype was created in Figma4, which is an online tool for creating interactivegraphical user interfaces. This prototype was then used in the user test.

3.5 Test

The last phase in the chosen design process was to test the design on users. Thiswas made with the purpose of getting a better understanding of how the users areable to understand the elements of the design, and also investigate the usabilityof the prototype as described by Nielsen [11]. By testing prototypes in the designprocess, flaws and usability issues can be found at an early stage [10]. The testingwas conducted by doing a qualitative testing session with five users. By doing

4https://www.figma.com

12

Page 16: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

qualitative user tests with five users, Nielsen claims that most usability issues canbe found [12].

3.5.1 Test Case

The test was conducted by giving the tester an instruction on what to do. The testerwas supposed to book a time with a doctor for help with stomach problems, andwanting to meet the doctor in person instead of choosing to have a voice or videocall. Apart from the instruction, the user was given free hands with the prototype,and asked to think aloud to better be able to find usability issues in the prototype.By observing what the user do, and not direct their attention to a specific part ofthe screen, usability issues can be found without having influenced the user [11].After the test, the testers were asked questions about how they interpreted eachstep of the application in order to understand which parts of the application thatperforms well, and which that do not. The test was conducted remotely by usingGoogle Meet5. The prototype was shown by sharing the screen in the video call andletting the tester talk about where he or she would like to press if they had theirprototype in their hands.

Due to the covid-196 situation during the spring of 2020 making it harder to findusers for testing that has a limited knowledge of the Swedish language, the prototypewas tested with German copy instead of Swedish. This was done with the hope toachieve the same type of situation where the user does not understand the languagewell. German was chosen since, as Swedish, it is a Germanic language [13].

3.5.2 Test Participants

The user testing was conducted on people with different backgrounds. In orderto discuss how the result could be applied to the target group the testers were askeda number of questions about their background. A summary of their background ispresented below.

• All five test participants were women.

• The age span of the test participants were between 27-41, with a mean of 35,2,and a median of 38.

• All women were born outside of Sweden. Time in Sweden spanned between 5months to 27 years. The mean was 12,5 and the median was 8.

• Two of the test participants were at the time of the test studying, one worked asa UX designer, one as an accounting consultant and one with administration.

• All test participants used a computer and smartphone on a daily basis intheir work or studies. They also used their smartphone or other digital toolsextensively outside of their work, meaning that all participants had a solidtechnical experience.

• Examples of services used on their smartphone in their everyday lives werebank applications, communication services like Messenger, WhatsApp, and

5https://meet.google.com6https://covid19.who.int

13

Page 17: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

Facetime, shopping applications, and entertainment services such as Netflixand Youtube.

14

Page 18: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

4 Theoretical Framework

This section presents a theoretical framework that is based on an extensive literaturestudy in the areas of icon design, cross-cultural design, design for illiterate users,and design thinking. The framework will present information and areas that needto be considered when designing in a cross-cultural setting.

4.1 Icon design

Since this project aims to develop a design for users with a limited understandingof a specific language, icon design is an important part of the user interface. Whendesigning icons for graphical user interfaces (GUI), there are some criteria to meet.According to Huang et al. the most important of these are styling, message quality,meaningfulness, locatability, and metaphor [14]. Easterby described five characteris-tics of a good icon by using the concept of ”figural goodness” in Gestalt psychology.These are closure, continuity, symmetry, simplicity, and unity [15].

The new liberal art of design thinking points toward the impossibility of rigidboundaries between industrial design, engineering, and marketing. Furthermore, itpoints toward the fact that what one might call ”impossible”, actually just be alimitation of imagination that can be overruled by better design thinking [16].

Designing icons can be done in may ways, one can use icons only, or the combi-nation of label and icon. Both in terms of ease of use, and performance it is best toprovide the user with a combination of icon and label [17].

McDougall et al. [18] have explored the effects of icon characteristics on userperformance, in terms of concreteness, complexity, and distinctiveness. When itcomes to concrete icons, they are likely to be most useful in situations when iconlearning needs to occur quickly or instantly (e.g., in public transport). They arealso likely to be more effective when icons are not seen frequently (e.g., emergencywarnings). The effect icons complexity concerns the time a user takes to searcha display and act accordingly. Because of this, simple icons are most likely moreeffective in a time-critical setting. It should be prioritised in situations where thespeed of response is critical. By creating distinctive icons, the user can respondmore quickly.

4.2 Cross-cultural design

When designing something that is to be used by someone from a different countryor culture, the aspects of cross-cultural design must be considered. To understandcross-cultural design, we must first understand what culture is. Hofstede [19] con-ceptualised culture as ”programming of the mind” in 1991. He means that certain

15

Page 19: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

reactions are more likely in certain cultures than in other ones, based on differencesbetween the basic values of the members of different cultures. Heimgartner [20]compares three groups from different cultures and can conclude that their interac-tion pattern differs significantly between each other. Pfeil et al. [21] draws the sameconclusions, that we act based on our cultures online, from a similar study.

4.2.1 Challenges in cross-cultural design

As a designer, one must be aware of the differences mentioned above and act ac-cordingly to the challenges they bring. According to Kamppuri [22, p 174-175], themain challenges of cross-cultural design are:

Ideas of culture When working with cross-cultural design, the theories and meth-ods used are based on particular ideas about culture. These ideas must be madeexplicit to be able to evaluate their benefits and weaknesses for the intended purpose.

Values and worth Meaning that the design process should be based on an un-derstanding of the positive and negative value of technology as seen by the users.Users and designers might not always agree on what is worthwhile not only due toindividual differences but also because of cultural aspects. This gives the risk ofethnocentric thinking, which is evaluating others against the standards of one’s ownculture.

Wider context of use When designing for intracultural contexts, designers usuallyshare knowledge about the wider context of use at the level of society. In cross-cultural settings, a designer might face challenges since this shared knowledge mightbe missing.

Wider context of design The lack of knowledge, experience, and informationabout the wider context also affects the design process. A designer also needs topossess an understanding of cross-cultural communication to be able to communicatewith users.

Design theories and methods as cultural products The theories and methods thatare used today have been developed in a setting or context that is different fromthe cross-cultural design setting of today. Therefore, a designer must be critical andopen-minded about improvising to come up with new design techniques that bettercould suit the local setting.

Furthermore, Kamppuri [22, p 64-67] mentions the problem with generalisationas a challenge in cross-cultural design. This means that one must be aware ofcultural assumptions, such as treating national culture as a single variable thatdetermines whether or not a specific technique will work or not. This does not takeinto account cultural variety inside a country and the situational factors. Kamppurifurther discusses how usability attributes depend strongly on the context of use.Some usability attributes can be highly relevant in one culture, while not matteringin another culture, instead, other aspects of use then become more relevant.

Bourges et al. [23] have found several challenges in cross-cultural design. Thefirst one being that understanding representations is a common usability problem.Differences in values, language, and taste are seen as representational differencesbetween cultures. The understandability of a representation affects its learnability.If the meaning of a representation is specific to a culture, it is not suitable for aheterogeneous user group. Understandability and learnability also might influence

16

Page 20: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

the users’ preferences. Bourgees et al. [23] also finds that users favour systems thatthey understand better, regardless of their native language, over systems that havebeen translated into their native language.

4.2.2 Designing in a cross-cultural setting

If we consider computers as of today not being able to possess empathy, we mustas designers be able to immerse ourselves in the cultural domains, user groups,and context to be able to extract the relevant requirements for the design of theinteraction. However, one must be conscious of the challenges of empathy. Whenworking with cross-cultural design, it is still clear that for the communication tobe successful, both parties must be open and through that use certain empathicskills [24].

When doing usability think-aloud sessions for testing interaction in a cross-cultural setting, there is a rich non-verbal component of communication in the formof gesture-clusters and body movements [25]. As a designer, it is a valuable inputin a user testing session to include gestures from the user.

4.3 Universal design

The Center for Universal Design at North Carolina State University has developeda set of design principles for universal design, presented by Story [26]. Some ofthese principles can be useful in this project. Their third principle is simple andintuitive use with guidelines to eliminate unnecessary complexity, accommodate awide range of literacy and language skills, and provide effective prompting andfeedback during and after task completion. Principle four is perceptible informationwhich comes with the guidelines to use different modes to for redundant presentationof essential information (e.g. pictorial, verbal, and tactile), maximise legibility ofessential information, differentiate elements in ways that can be described (i.e., makeit easy to give instructions or directions. The fifth principle is tolerance for error.This means to arrange elements to minimise hazards and errors by making the mostused elements most accessible and vice versa, and lastly, provide fail-safe features.The principles of universal design have successfully been implemented in differenteducational settings, proving their validity [27][28]. When using the principles ofuniversal design, it is still important to think about the main goal which is tosupport all users, and not only focus on the users who might have some form ofdisability or difficulty to cope with the technology [28].

4.4 Communication

Communication is used in many different situations in our everyday lives. Whendesigning a digital product that is to be used by a person that does not speak thelanguage in the region one must have an understanding of what communicationactually is, and how it can be divided.Liddicoat [29] describes communication as an act of sociality. It is simply not thecase that information is transferred between participants, but rather that language

17

Page 21: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

is used to create and maintain social relationships. Furthermore, he talks aboutlanguage as a marker of one’s identity, and the use of it is an act of social identity inthat it encodes how the speaker is presenting him/herself in a particular interaction.

4.4.1 Handling illiteracy in communication

When designing for people that are new to the Swedish language, comparisons can bemade with illiterate people. Harrison-Walker [30] implies four practical implicationsfor direct mail marketers to be able to reach all potential customers. These are:

• Focus on how the receiving audience will use the product, rather than detailingproduct/service benefits.

• Use someone who is not connected to the organisation or product for proof-reading the copy in terms of clarity, accuracy, and comprehension.

• The organisation should be pretested on a sample of the target audience tobe able to match the reading level of the piece that can be matched to thecapabilities of the potential customers.

• Various options for channels used for marketing should be evaluated, to meetas many potential customers as possible.

By using these implications when designing for the target audience, we raise ourchances of the design being understood.According to Chipcase, we can use audio as a way to help people that are illiterateto use a mobile phone, by letting them speak in their textual input or having theapp or service reading out loud what the text means on a given screen [31].

4.4.2 Second Language Communication Strategies

Tarone phrased a number of communication strategies for second language in 1980(as cited in [32] by Tarone herself, as well as in [33], [34] ). Some of these are:

• Approximation - the use of a single vocabulary item or structure, which thelanguage learner knows is not correct, but which shares enough semantic fea-tures in common with the desired item to make sure to be understood (e.g.pipe for waterpipe).

• Word coinage - means that the language learner makes up a new word in orderto communicate the desired concept (e.g. airball for balloon).

• Circumlocution - describing the characteristics or elements of the thought ofobject or action instead of using the appropriate term or structure in the targetlanguage.

By being aware of how people are communicating in a second language, we can usesimilar approaches when designing user interfaces for the same type of situations.

18

Page 22: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

4.5 Designing for Illiterate Users

Since a lot of graphical user interfaces contain a lot of text, there are some specialconsiderations to have in mind when designing for a user group not proficient in thecurrent language. We can learn from the case study made by Lalji and Good [35],and use some of their key lessons learned in this paper. Some of these are:

• Step-by-step instructions are preferred and proved to be more effective than ageneric help button. If language is used, the choice of words or phrasing mustbe given plenty of thought.

• Lists were more effective and understandable than hierarchical structures inmenus.

• The use of symbols must be done carefully since they might have differentmeanings in different cultures. The proximity of similar colours may alsoaffect the user’s ability to navigate.

• It is of high importance to take into consideration the wider context of usewhen designing.

By using a user-centered design process we increase our chances that the productwill be enjoyable and that the thought of the user will actually use it. This processis all the more important when designing for non-traditional users [35].

Medhi et al. [36] has made a study and provides us with some guidelines whendesigning for illiterate users. The first guideline is to avoid text as much as possible.Numbers are usually understood. Secondly, they suggest to use semi-abstractedgraphics and increase photorealism with deeper interaction. This means that photo-realistic graphics worked better than complex abstract graphics. The third guidelineis to pay attention to subtle graphical cues, which means that a user’s response maydepend on biases such as psychological, cultural, or religious ones. One specific detailis the need to display motion in an illustration. Next is to provide voice feedback forall functional units. The last guideline given by Medhi et al. [36] is to provide theuser with a ”help” button on all screens, which will give the user instructions for theapp to be more autonomously used. When designing for illiterate users, interfacesthat use speech and dialogue interaction, display non-linguistic graphics, adapt tothe literacy level of the user and accept input by microphone are well-suited to thedomain [37].

4.5.1 Presenting information to illiterate users

For illiterate users, one can use many tools to present information. Using voiceannotation generally helps in the speed of comprehension, while bimodal audio-visualinformation can be confusing. Furthermore, richer information is not necessarilyeasier to understand. Lastly, the relative value of dynamic imagery versus staticdepends on various factors [38]. Non-literate and semi-literate users can completetasks better with a rich multimedia user interface compared to a spoken dialogsystem, however, the speed of completion is faster and less assistance is needed in aspoken dialog system [39]. Medhi invites the reader to not generalise the conclusionstoo much from the study, but that the findings should be considered and kept inmind when designing.

19

Page 23: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

4.5.2 Design guidelines

When designing for a heterogeneous user group, we can take a look at differentstandards for accessibility to secure a design that increases the chance that all canuse it. The World Wide Web Consortium has created web accessibility guidelines,where guidelines regarding language will be applied to this project. This includes thesuccess criterion 3.1.5 Reading Level which states that a text which requires readingability more advanced than lower secondary education level, should be available ina simplified version [40]. Through designing a questionnaire for low-literate users,Cremers et al. [41] have created a number of guidelines together with users whichcan be applied to this project. The most important ones are:

• Information overload Means to avoid users feeling overwhelmed with infor-mation.

• Simple language Use simple language and short sentences.

• Pictures and icons Use pictures and recognizable icons to enhance compre-hensibility.

• Concrete Illustrations should be as concrete as possible.

• Consistent All illustrations should have the same look.

• Photographs better than illustrations If possible, use photographs in-stead of illustrations.

• Consistent screen layout Each screen should be recognizable, looking assimilar to each other as possible.

• Clearly recognisable buttons Buttons should be recognisable and big enoughto be clearly seen on a screen. If possible, the form of a button should sym-bolise its function.

• Feedback Once a button is activated, it should be accentuated.

• Relaxed atmosphere Make sure people are comfortable, by creating an at-mosphere where there is no pressure.

After developing and evaluating an intercultural design concept, Schreick et al. [42]derived design principles for a user group with a diverse background. As previouslymentioned in this report [17], it is preferred when using icons that they are displayedin combination with text. The icon itself should also be validated across cultures toreduce the risk of it having diverse associations. Furthermore, an intuitive structureof information is indispensable. Derived from their results, they recommend that theoverall information should be clustered hierarchically to ease navigation. In combi-nation with the structure, the 3-Click-Rule should be fulfilled, i.e. users should beable to find the sought-after information within three clicks starting from launchingthe application. In terms of usability, common interaction patterns such as scrollingor swiping should be used. By involving the users in the design process, Cremers etal. [43] have formulated some design guidelines for ATM user interfaces suitable forilliterate persons. Some of these can be applied to this project. Regarding text, the

20

Page 24: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

recommendation is to use foreground-background contrasts as well as using a simplelanguage and placing text close to the associated button. When using icons, theyshould be used in addition to text, as previously mentioned in this paper, and theyshould be concrete. Icons should also show how to do things, preferably by usinganimation. When it comes to navigation, the number of menu options should besmall, buttons should have a label or color, effects of choices should be made clear,and the user should be able to control the pace of the interaction by inserting pauseswhen wanted.

21

Page 25: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

5 Related Works

Schreick et al. have developed some design principles for mobile applications thattransmit local information to refugees with different cultural backgrounds [42]. Thesedesign principles can be used in this study to understand the cultural difference ofusers. Bodner and MacKenzie find that by using animated icons to represent com-plex tasks, the user can more often identify which task is represented than if theicon is static [44]. Smith et al. have developed a process model for developing usablecross-cultural websites. Their conclusions most relevant to this paper are that semi-otic we cannot assume that western methodologies and techniques for user-centereddesign and participation can be used in other cultures without adaptation. Further-more, they find that semiotic attractors can successfully be used both to audit sitesfor indigenous user requirements and to identify potential design solutions [45].

Several studies have been made comparing how users from different cultures in-teract with computers. Choong and Salvendy have made a study which investigatesthe differences between American and Chinese users and finds differences regardingperformance time by letting the users try an application both with pictorial andalphanumerical mode [46].

Walsh et al. [47] have made a UX evaluation of a smartphone among differentcultures, to be able to see any differences. What they mention is that by usingHofstede’s [19] cultural dimensions they could understand differences in how usersevaluated the smartphone. They also mention that to truly understand the user andhow they experience the evaluation, interviews face to face is preferred. By usingthese insights, this thesis can benefit from doing interviews with users, rather thanproviding them with an online survey.

22

Page 26: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

6 Results

The following chapter will present the results of this thesis project. It includes themain parts of the design process which are understanding, exploring, and material-ising.

6.1 Understanding

Below the results from the understanding phase of the design process will be pre-sented.

6.1.1 User Research

Through the case study explained in section 2.1.2, a broader understanding ofthe targeted user group was achieved. From the case study, it was learned that thetargeted users need digital products that provide help in everyday life and that theyshould be able to be used regardless of language abilities. The users have a needand are motivated to use digital services, but sometimes lack the language abilitiesor cultural understanding to do so. Some key points from the research are:

• There is a need for a cultural understanding when designing a product (e.g.not everyone is used to navigate with a map in their phone).

• The language must be simple to understand and read.

• The use of icons is preferred to help the user understand written text.

• The technology must be simple, and not feel complicated to use.

The insights gained from the user research were used to better understand whattype of literature research was needed to fulfill the needs of the users.

6.1.2 Takes From Literature Study

The literature study gained insight into what areas and aspects of a design thatis relevant and needed when designing both in a cross-cultural setting but also todesign for users that are not proficient in a certain language. The literature studyresulted in a number of design guidelines used in the design (see section 6.2.2 below).

6.1.3 Persona

Based on the information gained from the empathise phase, a persona was generatedto gain information as to how the user’s needs and wants play a role in the design.The persona is Farah, see figure 4, a 27-year-old single mother living with her two

23

Page 27: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

children in a two-room apartment. She has studied SFI for two months but iscurrently on maternity leave. She has a relatively good technical experience, but witha lack of knowledge in the Swedish language. This is causing her to experience issuesusing some of the digital services in everyday life in Sweden such as applications forpublic transport, banking, job search, etc.

Figure 4: Visualisation and information about the persona created

The key aspects of the persona are:

• She is used to technical devices such as a smartphone but can have a hardtime understanding the user interface in a new language.

• The lack of knowledge of the language is the primary barrier to be able to usedifferent digital services.

• She has acquired a basic knowledge of the Swedish language through SFI.

• The majority of her friends have the same mother tongue as her, meaning thatshe is struggling with learning Swedish since there is a lack of situations to usethe language in.

The persona was used as a foundation in the design to design a product that isculturally suitable for the target group.

24

Page 28: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

6.2 Exploring

Below the results from the exploring phase will be presented, ideate and prototypingthat is.

6.2.1 Low-Fidelity

By using the association cards described in the method, several quick and roughsketches were produced. In figure 5, several different versions for the doctor bookingapp is displayed. The sketches display the different steps that the user needs to taketo book a time. The session resulted in several different structures of the flow andhow to order information on each screen. Each of the sketches resulted in a flowwhere the user only focuses on one thing on every screen. Furthermore, differenttypes of customisation features to the booking emerged, such as choosing a day forthe meeting, a preferred way of contact, visualisation of symptoms, and differentways of available inputs for the user. These sketches were analysed and combinedinto the high-fidelity prototype presented below.

Figure 5: Sketches on paper of the booking prototype

6.2.2 High-Fidelity

The high-fidelity prototype resulted in a prototype that displayed the flow of a userthat is supposed to book a time with a doctor, either remotely or in person. The usermust first enter the app by choosing ”book an appointment”. Then the user is askedto choose from a number of categories of symptoms, which one that suits him/herbest for this occasion. To give the doctor as much information as possible, the useris asked to describe its symptoms either in text or by speech. Before finalising thebooking, the user must choose in what way he/she would like to get in contact. The

25

Page 29: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

options are voice call, video call, or a physical meeting at the clinic. Once the userhas made his/her choice, a summary of the booked time is given. The prototype isdisplayed in Figure 6 - 11. The prototype shown is with German copy, as explainedin the method.

Figure 6: Start page of proto-type

Figure 7: A list of categorisedsymptoms

26

Page 30: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

Figure 8: Input from the user,either by text or voice

Figure 9: The user has typedin information abouttheir symptoms

Figure 10: Choosing how theuser wants help

Figure 11: Summary of thebooked appoint-ment

27

Page 31: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

6.3 Materialising

Below the results from the user testing are presented.

6.3.1 User testing

As shown in 6.2.2, the user was shown several screens throughout the prototype. Allusers were able to understand the context of the page and able to complete the giventask, find stomach i.e., on the screen with categories of symptoms (see Figure 7). Allusers claimed that they started with reading the text, and then confirming their ideaof what it meant with the icons. All users understood that the toilet symbolised thestomach issues. The icons were overall understandable and the users were able tounderstand their representation. The icon for headache and physical meeting werea bit unclear for some users.

When shown the next screen, Figure 8, four out of five users chose to write theirsymptoms instead of speaking. Several users mentioned the word ”symptome” inGerman as a key to understand what type of input they were supposed to give.Users also mentioned that they were helped by the large icon with a pen and lineswritten on a piece of paper. Several users immediately recognised the text input fieldand said that this screen was somewhere they were supposed to type something, andthrough that they realised what type of input they were supposed to give assistedby the other cues on the screen.

Next, the user was given a prompt on how they would like to get in contact, seeFigure 10. The majority of the test participants mentioned that they immediatelywere sure on which option to choose, based on the instruction given, thanks to thesimilarity in language between German and Swedish on the first two options. Thesecould be ruled out and the third was the one to choose. Several testers interpretedthe screen as to how the clinic could contact themselves, the patient i.e., rather thanhow they would like to get help. One participant thought that the icon for a physicalmeeting meant to get in contact with some form of a chatbot.

On the last screen, see Figure 11, all users were able to understand the infor-mation presented, that it was a date, time, and place. Several users were insecureabout whether or not the information could be changed to suit their needs, or ifthey were simply given a time slot to come to the clinic.

Usability Issues

Throughout the testing, a few usability issues were found with the prototype:

• When asked how the patient would like to be assisted, it could be interpretedas how the clinic can contact them.

• The summary to the user when the booking is complete was hard to understandwhether or not it was a suggestion or a confirmation of a completed booking.

• Some icons were hard to interpret properly, especially the one for headacheand the physical meeting as an option.

28

Page 32: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

7 Design choices

In the following sections, the motivation behind design choices will be presented.

7.1 Icons

As Wiedenbeck [17] mentions, that by using a combination of text and icons, per-formance and ease of use is improved. This has been used throughout the prototypewith good results. The choice is also supported by Schreick et al [42]. The icons havebeen designed as concrete and consistent as possible, as recommended by Cremerset al. [41] and McDougall [18] to help the user quickly understand the icon. Theicons have been designed to be semi-abstract when it comes to the graphical level,as suggested by Medhi et al [36]. In this prototype, this means that the icons havebeen designed to be as realistic as possible. Lastly, the icons have been designedwith Easterby’s [15] characteristics of a good icon in mind, especially closure andunity. The icons are designed without any linguistic elements, helping users thatmight not be able to read the language [37].

7.2 Layout

As recommended by Cremers et al. [41], the layout of each page has been designed ina minimal manner, to minimise the risk of information overload. Also, each screenhas been designed as consistent as possible. Furthermore, their recommendationto have clearly recognisable buttons has been used, designing large buttons with alabel. This is also supported by another study made by Cremers et al [43].

Lists have been used instead of hierarchical structures when presenting severaloptions to the user. This is recommended by the work done from Lalji and Good [35].

7.3 Interaction

The interaction flow has been designed to provide step-by-step instructions, givingthe user only one task at each screen in the process. This is supported by the casestudy made by Lalji and Good [35]. As Cremers et al. [43] suggests, the user shouldbe able to control the pace of the interaction by inserting pauses when needed. Thishas been achieved by dividing the interaction in a number of steps, each with a clearpurpose.

By dividing the interaction flow in a step-by-step manner, the use of principlethree from The Center for Universal Design which is simple and intuitive use [26] hasbeen used. More specifically, it means to provide effective prompting and feedback

29

Page 33: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

during and after task completion. By labeling each screen with a text and/or icontelling the user what to do the users are given feedback during the task completion.This is also the reason for implementing the progress bar in the header, to providethe user with feedback on how far he or she has come in the task.

In the prototype, the user is given the possibility to speak in their textual inputwhen describing their symptoms, as supported by Chipcase [31] and [37].

Each clickable area in the prototype has been given a clear indication of whatcomes next. Each list item has been given an arrow symbolising that by pressingthe option, the user is taken to a new screen. The same goes for the large buttonswhich generally are used as a confirmation or call to action. This has been donebased on Cremers et al. [43] suggestion to make the effects of choices clear to theuser.

7.4 Text

Even though the Swedish text in the prototype was not tested, it has still beendesigned with some aspects in mind. One of them is the recommendation by Cremerset al. [41] to use simple language and short sentences. No excessive text has beenused. The text has also been designed with a clear foreground-background contrastto simplify the reading, as supported by Cremers et al [43].

30

Page 34: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

8 Discussion

In the following chapter the results of the study, and methodology behind it will bediscussed.

8.1 Result Discussion

Thanks to the similarity between the German and Swedish language, most testparticipants were able to understand some of the text in the app, helping them tocomplete a task. By noticing the similarity of words, the test participants used thecommunication strategy of approximation, as cited in [32] by Tarone, as well asin [33], [34], to understand the meaning of the term.

Some usability issues were found during the testing phase. These are to beconsidered solvable if the time would have been sufficient to conduct more iterationsof the design and testing it.

Since all users, without many issues, were able to complete the task given in theinstruction the result from the study is to be considered successful. The hypothesisset up in the initial phase was then correctly assumed. The literature study, com-bined with the user research, proved to be a good foundation to be able to achieve agood result. The guidelines and methods provided by that gave the project a solidground to stand on, and the design choices(see 7) made based on that was a key tothe successful project. A special thing which was an important part to achieve agood result was to early on get a proper understanding of the users’ needs, issuesand opportunities. This understanding helped the design to be adapted to theirspecific situation. Without that understanding, the result would most likely not beas successful.

8.2 Methodology Discussion

In this section the methodology in the project will be discussed.

8.2.1 Research methodology

The methodology chosen in this thesis has been able to provide a good result. Eventhough, there is room for improvement. At the time of the thesis, the covid-19pandemic struck Sweden, which affected how this thesis project was carried out. Atthe beginning of the project, the plan was set out to work with the partner TidigtForaldrastod (TFS) when it comes to user testing of prototypes and sketches. Thishas not been possible to do due to the situation in society. The target group atTFS could have been of great help for this project if early sketches and prototypes

31

Page 35: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

could be tested to minimise the risk of using interaction patterns, icons, colours, orother graphical elements that might have a different meaning or interpretation inanother culture. Due to the lack of this contact, the result has not been validatedwith the target group. Instead, users as close as possible to the target group havebeen chosen in order to get a result that could be seen as relatively legit (see 3.5.2for information about the participants).

By talking to members of the target group in the initial phase of the project,through the HIVE project (see 2.1.2), an understanding of the culture could beestablished. Specifically, Kamppuri’s [22, p 174-175] Ideas of culture and valuesand worth could be understood. This means that more than understanding theculture, the design process could then be based on an understanding of the positiveand negative values of technology as seen by the users. This was important to avoidethnocentric thinking, which means to evaluate others against the standards of one’sown culture. By continuing this research further on in the project, the result couldbe improved even more with a deeper understanding of the users. This could alsominimise the risk of generalisation, as described by Kamppuri [22, p 64-67], whichmeans to be aware of cultural assumptions, and that some usability attributes can behighly relevant in one culture, but not mattering in another culture. Furthermore, animprovement that could be made is that the icons made should be validated acrosscultures to reduce the risk of them having diverse associations. This is supportedby Sxhreick et al [42].

Furthermore, the methodology could be improved by earlier deciding on whichcontext the theory is to be used in. For this project, the aim at first was to investigatehow different aspects of user experience design can help users complete a numberof tasks in a simple application. If the context of the application would have beendecided at first, the project would probably have benefited from it. This since itcould have simplified both the literature study and the user research. Users couldhave been asked questions that were directed more specifically towards the contextof use. This is also supported by what Kamppuri [22, p 174-175] and Lalji andGood [35] says about the wider context of use, that in an intracultural setting itmight be hard to understand the wider context of use of an application. This is ofcourse even harder to understand if the context is not decided early enough in theprocess.

What could have helped this project be more successful would be to, in an earlystage, interview professional designers from different cultures and backgrounds abouttheir experiences with designing products that are to be used by a culturally diverseuser group. With this, a deeper understanding of the work progress could have beenestablished.

8.2.2 Testing methodology

When testing the designed prototype, all users were given the same instruction tosearch medical care for some symptoms regarding their stomach. All users were alsoasked to book a physical meeting with a doctor. If the users would have been givendifferent instructions on how to complete the task, e.g. switching the symptoms orchoosing another way of getting in contact, more usability issues could possibly havebeen found. The chosen test method only tested a fixed case, and not the design asa whole, which is a limitation in this study.

32

Page 36: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

Three out of five test participants were more or less fluent in the Swedish lan-guage, therefore all tests but one was done with German copy in the prototype. Thisto simulate that the tester understands some parts of the written text, as the targetgroup would do. Several testers mentioned that the text helped them understandthe context and task they were going to complete, thanks to the similarity with theSwedish language. If the language chosen would not be closely related to Swedishhowever, the tester would have to rely solely on design elements such as icons andbuttons. This could have been beneficial to be able to test the iconography andother graphical elements more thoroughly. Despite this, the choice of using Germanas the language in the prototype is seen as the right one due to it was able to simu-late a situation that was relatively close to how it would have played out in real lifefor a user from the target group. The single test with Swedish copy experienced thesame type of minor understanding of words when they were similar to the word inEnglish.

As mentioned in the result, all test participants were able to complete the giveninstruction without so many issues. This raises the question if the test was designedto be too easy? Were the participants able to do wrong at all? They could choosethe wrong option on the category of symptoms page, but were the options simplytoo distinct? As mentioned by [18], distinct icons help the user respond quickly andfind what they are searching for. Since all icons in the list were fairly distinct, thissupports why the participants were able to find the correct choice so quickly.

The test session, done remotely, was not recorded. Notes were taken while thetest was taking place. However, if the test would have been recorded, a richer non-verbal communication in the form of gesture-clusters and body movements couldhave been found. These could tell us more about how the user reacts to differentparts of the prototype. Some movements and facial expressions were noticed duringthe testing, but far from all. This is something that could have been improved,which is supported by Yammiyavar et al. [25].

The test participants were not in the target group specified in the project.Through the collaboration with Tidigt Foraldrastod, the initial plan was to conductuser testing with their participants. Due to the covid-19 pandemic, their premiseswere closed during this thesis project, which made it impossible to test on their par-ticipants, even remotely. Therefore another test group was used. These were chosento be as close to the thought of target group as possible to be able to have a rep-resentative result. All testers were women, whom all were born outside of Sweden.All of them can be categorised as high performers, with jobs or occupations wherethey use computers every day, diverging them quite a far way from the target group.This could have an influence on the result, in terms of the testers being used to alot of similar apps or services, helping them use the prototype. All users were eitherfluent in Swedish or English, which most likely helped them understand the Germancopy. If the test would have been performed on the target group however, the userswould probably have more difficulties with the language due to their native tongueand Swedish not being closely related.

33

Page 37: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

9 Conclusion

The aim of this thesis was to investigate what aspects of user experience design thatcould be used to develop digital services in order to help users complete tasks andunderstand the context, even when the language used might not be fully understood.The underlying purpose was to understand what could be done to help people thatare new to a country to understand the essential digital services that are beingused in public transport, health care, etc. By conducting a literature study andinterviewing people that are new to the Swedish society a deep understanding ofuser’s issues and potential solutions and methods to use was gathered. Once thetheoretical framework was completed, the next phase of the thesis begun, being theprototype phase. Prototypes were created in rising fidelity order, designing a servicethat was to be tested on real users. The testing phase of the thesis proved to besuccessful, meaning that by using aspects of cross-cultural design, design for illiterateusers, and icon design it is possible to create services that are understood by userseven though the language might not be fully understood. Therefore confirming thehypothesis constructed in the initial project phase.

9.1 Future Work

Due to the limited time frame in this thesis, a limited service was designed. Inorder to draw further conclusions, more services must be created using a similarmethodology and theories, preferably in a number of diverse situations and withdifferent complexities.

34

Page 38: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

10 Acknowledgements

The author would like to thank everyone at Humblebee for their support throughoutthis project. A special thank you goes to Mattias Hagbard for being a mentor duringthe project and helping out with feedback along the way. The team at HIVE alsodeserves a big thank you for their work with the research and help with finding users.Thank you to all the test participants who participated in the user tests. Last butnot least, the author would like to thank everyone that helped review this paper.The peer-reviewers Filip Bark, Oscar Thorwid, Petter Poucette, and the mentor atUmea University, Ole Norberg.

35

Page 39: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

References

[1] H. Esser, Migration, language and integration. WZB Berlin, 2006.

[2] K. Huotari and J. Hamari, “Defining gamification: a service marketing perspec-tive,” in Proceeding of the 16th international academic MindTrek conference,pp. 17–22, 2012.

[3] D. Norman and J. Nielsen, “The definition of user experience (ux).” https:

//www.nngroup.com/articles/definition-user-experience/. Accessed:2020-02-06.

[4] M. Hassenzahl, “User experience and experience design,” The encyclopedia ofhuman-computer interaction, vol. 2, 2013.

[5] D. Sless, “Defining information design.” https://communication.org.au/

defining-information-design/, 2007. Accessed: 2020-03-04.

[6] R. Pettersson, “Information design–principles and guidelines,” Journal of Vi-sual Literacy, vol. 29, no. 2, pp. 167–182, 2010.

[7] T. A. KELLEY, The art of innovation: Lessons in creativity from IDEO, Amer-ica’s leading design firm, vol. 10. Broadway Business, 2001.

[8] S. Gibbons, “Design thinking 101.” https://www.nngroup.com/articles/

design-thinking/, 2016. Accessed: 2020-02-18.

[9] “Personas.” https://www.interaction-design.org/literature/topics/

personas/. Accessed: 2020-03-10.

[10] M. Walker, L. Takayama, and J. A. Landay, “High-fidelity or low-fidelity, paperor computer? choosing attributes when testing web prototypes,” in Proceedingsof the human factors and ergonomics society annual meeting, vol. 46, pp. 661–665, SAGE Publications Sage CA: Los Angeles, CA, 2002.

[11] J. Nielsen, “Usability 101: Introduction to usability.” https://www.nngroup.

com/articles/usability-101-introduction-to-usability/, 2012. Ac-cessed: 2020-04-17.

[12] J. Nielsen, “Why you only need to test with 5 users.” https://www.nngroup.

com/articles/why-you-only-need-to-test-with-5-users/, 2000. Ac-cessed: 2020-04-17.

[13] Isof, “Svenska.” https://www.isof.se/om-oss/for-dig-i-skolan/

sprak-for-dig-i-skolan/spraken-i-sverige/svenska.html, 2014. Ac-cessed: 2020-05-05.

36

Page 40: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

[14] S.-M. Huang, K.-K. Shieh, and C.-F. Chi, “Factors affecting the design of com-puter icons,” International Journal of Industrial Ergonomics, vol. 29, no. 4,pp. 211–218, 2002.

[15] R. S. Easterby, “The perception of symbols for machine displays,” Ergonomics,vol. 13, no. 1, pp. 149–158, 1970.

[16] R. Buchanan, “Wicked problems in design thinking,” Design issues, vol. 8,no. 2, pp. 5–21, 1992.

[17] S. Wiedenbeck, “The use of icons and labels in an end user application pro-gram: an empirical study of learning and retention,” Behaviour & InformationTechnology, vol. 18, no. 2, pp. 68–82, 1999.

[18] S. J. McDougall, O. de Bruijn, and M. B. Curry, “Exploring the effects of iconcharacteristics on user performance: The role of icon concreteness, complex-ity, and distinctiveness.,” Journal of Experimental Psychology: Applied, vol. 6,no. 4, p. 291, 2000.

[19] G. Hofstede et al., “Organizations and cultures: Software of the mind,” Mc-GrawHill, New York, 1991.

[20] R. Heimgartner, “Cultural differences in human computer interaction: Resultsfrom two online surveys.,” in Isi, pp. 145–157, 2007.

[21] U. Pfeil, P. Zaphiris, and C. S. Ang, “Cultural differences in collaborative au-thoring of wikipedia,” Journal of Computer-Mediated Communication, vol. 12,no. 1, pp. 88–113, 2006.

[22] M. Kamppuri, Theoretical and methodological challenges of cross-cultural in-teraction design. University of Eastern Finland, 2011.

[23] P. Bourges-Waldegg and S. A. Scrivener, “Meaning, the central issue in cross-cultural hci design,” Interacting with computers, vol. 9, no. 3, pp. 287–309,1998.

[24] R. Heimgartner, L.-W. Tiede, and H. Windl, “Empathy as key factor for suc-cessful intercultural hci design,” in Design, User Experience, and Usability.Theory, Methods, Tools and Practice (A. Marcus, ed.), (Berlin, Heidelberg),pp. 557–566, Springer Berlin Heidelberg, 2011.

[25] P. G. Yammiyavar, T. Clemmensen, and J. Kumar, “Influence of cultural back-ground on non-verbal communication in a usability testing situation,” Interna-tional Journal of Design, vol. 2, no. 2, 2008.

[26] M. F. Story, “Maximizing usability: the principles of universal design,” Assis-tive technology, vol. 10, no. 1, pp. 4–12, 1998.

[27] K. Rao, P. Edelen-Smith, and C.-U. Wailehua, “Universal design for onlinecourses: applying principles to pedagogy,” Open Learning: The Journal ofOpen, Distance and e-Learning, vol. 30, no. 1, pp. 35–52, 2015.

37

Page 41: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

[28] R. Dolan, T. E. Hall, M. Banerjee, E. Chun, and N. Strangman, “Applyingprinciples of universal design to test delivery: The effect of computer-basedread-aloud on test performance of high school students with learning disabili-ties,” The Journal of Technology, Learning and Assessment, vol. 3, no. 7, 2005.

[29] A. J. Liddicoat, Teaching languages for intercultural communication. PhD the-sis, Federation Internationale des Professeurs de Langues Vivantes, 2005.

[30] L. J. Harrison-Walker, “The import of illiteracy to marketing communication,”Journal of Consumer Marketing, 1995.

[31] J. Chipchase, “Reducing illiteracy as a barrier to mobile communication,”Handbook of mobile communication studies, pp. 79–89, 2008.

[32] E. Tarone, “Some thoughts on the notion of communication strategy,” TESOLquarterly, vol. 15, no. 3, pp. 285–295, 1981.

[33] Z. Dornyei, “On the teachability of communication strategies,” TESOL quar-terly, vol. 29, no. 1, pp. 55–85, 1995.

[34] A. L. Wenden, “Learner development in language learning,” Applied linguistics,vol. 23, no. 1, pp. 32–55, 2002.

[35] Z. Lalji and J. Good, “Designing new technologies for illiterate populations: Astudy in mobile phone interface design,” Interacting with computers, vol. 20,no. 6, pp. 574–586, 2008.

[36] I. Medhi, A. Sagar, and K. Toyama, “Text-free user interfaces for illiterateand semi-literate users,” in 2006 international conference on information andcommunication technologies and development, pp. 72–82, IEEE, 2006.

[37] M. P. Huenerfauth, “Design approaches for developing user-interfaces accessibleto illiterate users,” University College Dublin, Ireland, 2002.

[38] I. Medhi, A. Prasad, and K. Toyama, “Optimal audio-visual representations forilliterate users of computers,” in Proceedings of the 16th international conferenceon World Wide Web, pp. 873–882, 2007.

[39] I. Medhi, S. N. Gautama, and K. Toyama, “A comparison of mobile money-transfer uis for non-literate and semi-literate users,” in Proceedings of theSIGCHI Conference on Human Factors in Computing Systems, pp. 1741–1750,2009.

[40] W. W. W. Consortium, “Web content accessibility guidelines (wcag) 2.1.”https://www.w3.org/TR/2018/REC-WCAG21-20180605/, June 2018. ”Ac-cessed: 2020-03-16.

[41] A. H. Cremers, M. Welbie, K. Kranenborg, and H. Wittink, “Deriving guidelinesfor designing interactive questionnaires for low-literate persons: development ofa health assessment questionnaire,” Universal Access in the Information Soci-ety, vol. 16, no. 1, pp. 161–172, 2017.

38

Page 42: Designing a Digital Service for Users with Low ... › smash › get › diva2:... · 3.2.1 Persona 10 3.2.2 Design Challenge 10 3.3 Ideate 10 3.4 Prototype 11 3.4.1 Low-Fidelity

[42] M. Schreieck, J. Zitzelsberger, S. Siepe, M. Wiesche, and H. Krcmar, “Support-ing refugees in everyday life-intercultural design evaluation of an application forlocal information.,” in PACIS, p. 149, 2017.

[43] A. H. Cremers, J. G. de Jong, and J. S. van Balken, “User-centered design withilliterate persons: the case of the atm user interface,” in International Confer-ence on Computers for Handicapped Persons, pp. 713–720, Springer, 2008.

[44] R. C. Bodner and I. S. MacKenzie, “Using animated icons to present complextasks,” in Proceedings of the 1997 conference of the Centre for Advanced Studieson Collaborative research, p. 4, IBM Press, 1997.

[45] A. Smith, L. Dunckley, T. French, S. Minocha, and Y. Chang, “A processmodel for developing usable cross-cultural websites,” Interacting with comput-ers, vol. 16, no. 1, pp. 63–91, 2004.

[46] Y.-Y. Choong and G. Salvendy, “Design of icons for use by chinese in mainlandchina,” Interacting with computers, vol. 9, no. 4, pp. 417–430, 1998.

[47] T. Walsh, P. Nurkka, and R. Walsh, “Cultural differences in smartphone userexperience evaluation,” in Proceedings of the 9th International Conference onMobile and Ubiquitous Multimedia, pp. 1–9, 2010.

39