Professional Documents
Culture Documents
User-Centered Design in Developing Countries
User-Centered Design in Developing Countries
Abstract—User-centered design (UCD) is a well-established de- by big differences in education and livelihood. Lack of training,
velopment methodology that focuses during the whole design pro- lack of experience and awareness, and lack of project manage-
cess on users’ goals and context. This is especially important in ment skills characterize the software development situation in
software development projects in developing countries, where the Ethiopia [4][8]. Heterogeneous cultures among different ethnic
development and use situations typically greatly diverge. We are groups and between rural and urban contexts are additional chal-
currently conducting a joint research project pursuing technol- lenges for Ethiopian software development [3].
ogy-enabled maternal and child healthcare in rural areas of Ethi-
opia. In this paper, we report our findings and lessons learned TEMACC, or Technology-Enabled Maternal and Child
from employing a UCD approach to address challenges stemming Healthcare in Ethiopia, is a research project that aims to explore
from lack of education, lack of training, and mostly illiterate users. the potential of information and communication technologies
Keywords—User-centered design, developing countries, Ethio-
(ICT) to improve healthcare access and quality for mothers and
pia, case study, healthcare children in rural communities. It is an attempt to contribute to
aspects of two of the UN Sustainable Development Goals [13],
I. INTRODUCTION namely SDG 3: Ensure healthy lives and promote well-being for
all at all ages and SDG 5: Achieve gender equality and empower
User-centered design (UCD) is a development methodology,
all women and girls. TEMACC is an interdisciplinary effort by
first described around 30 years ago, that emphasizes usability
professionals from the ICT fields together with specialists from
goals, user characteristics, and context of use during all stages
public health and medicine.
of the design process [6]. In [9], UCD is defined as the active
involvement of users for a clear understanding of user and task In this paper, we report on a case study we conducted while
requirements, iterative design and evaluation, and a multi-dis- eliciting user requirements and developing software prototypes
ciplinary approach. User requirements are thus central to UCD in the TEMACC project. The software was developed at Addis
processes; the design is typically iteratively developed, evalu- Ababa University (AAU) in Ethiopia to support mothers and dif-
ated, and refined through investigative methods such as usability ferent stakeholders of the local health system in rural areas. We
testing, prototype testing, or field studies [9]. An analysis of ex- report lessons learned and findings from several field studies and
isting work in this area shows a positive correlation between user workshops held to collect requirements and validate the soft-
participation or involvement and system success [1]. ware prototypes with end-users.
UCD is a well-established software development methodol- In Section 2, we present the goals and background of the
ogy in industrialized countries, but it is even more important in TEMACC project and describe the health system in Ethiopia.
developing countries, where the gap between software develop- Section 3 presents the research questions and methods. Section
ment and the local use situation is larger than in western coun- 4 illustrates the case study we conducted, and Section 5 presents
tries [2]. This gap is larger still in rural areas, where even the our findings and lessons learned. We conclude the paper in Sec-
basic conditions expected in industrialized countries cannot be tion 6 with an outlook for the community-based intervention
met. Developing countries, such as Ethiopia, are characterized study.
12
The underlying research question is whether the maternal During this round of field visits, we also gave HEWs diaries to
and child healthcare practices of rural communities can be sig- record any peculiar activities, challenges, or best practices they
nificantly improved using information and communication tech- experienced in their day-to-day activities over a period of three
nologies. weeks.
The research methods applied are participatory [5] and eth- We conducted the second set of interviews with mothers,
nographic action research [7] with the direct participation and aiming to understand their existing level of awareness with re-
active collaboration of the user community. Ethnographic action spect to antenatal care (e.g., follow up, danger signs, birth pre-
research is an approach particularly suited for understanding dif- paredness, nutrition, hygiene, safe delivery), neonatal care (e.g.,
ferent cultures. A key method in the approach is participant ob- breast feeding, cleanliness, baby bathing, burping, balanced diet,
servation, where the ethnographer participates in the culture or care for sick babies), postnatal care (e.g., family planning aware-
society being studied [7]. The purpose of action research is to ness, preventing unwanted pregnancies, abortion, counseling,
impact social change with specific actions [5]. Participants ac- consultation facilities), nutrition (e.g., optimal feeding practices,
tively contribute to the research in all phases of the process [5]. growth monitoring, body building foods, disease-protecting
foods), sick child services (e.g., diarrhea, malaria, pneumonia,
For understanding usage context, the main data collection vaccination), and hygiene (e.g., clean water usage, personal hy-
methods were interviews and focus group discussions. Inter-
giene, food hygiene, toilet usage). The interviews with mothers
views and discussions were in a free, conversational style with also gave us the opportunity to observe the sociocultural and
the goal of revealing more detailed information than structured economic situation of the community. During face-to-face dis-
interviews would have. Also, interview questions and discussion cussions, we made extra effort to create a friendly atmosphere
guides were piloted before actual use. Based on the results of the prior to starting the interviews through, for example, warm cul-
pilot, we made a decision to make interviews conversationally tural greetings, asking whether coffee is ready, and playing with
flexible and to use audio recordings of the sessions for later anal- children. We only started interviews once the team felt that
ysis. Various brainstorming sessions were held with respective mothers were ready for a friendly conversation, and we at-
stakeholders regarding the communication methods employed tempted to build emotional trust and openness.
amongst community members, usual sources of healthcare in-
formation, extent of use of technology (such as mobile phones) All interviews and focus group discussions were conducted
for communication and access to information, challenges en- in Amharic with a semi-structured interview guide. Since the in-
countered in sharing information, and existing knowledge-shar- terviews were more conversational in style, a digital audio re-
ing mechanisms. corder was used to record the discussions. In each interview ses-
sion, the team worked in pairs (one member from public health
IV. CASE STUDY and one from software development, paying particular attention
Ethnographic action research and participatory design meth- to gender: at least one member of each team was female). One
odologies were applied to gather user requirements and evaluate team member moderated the interview, and the other took notes.
software prototypes. First, we worked actively to create a com- Audio recordings were later transcribed and translated into Eng-
mon and clear understanding of ethnographic research through- lish. We also created meta-data and summaries of the interviews
out the team. We followed this with field visits to conduct inter- and focus group discussions in order to support the specification
views and focus group discussions with mothers, HEWs, and of user requirements.
health officers. The international standards for the ergonomics The identified usage context also helped us to learn about the
of human system interaction and human-centered design for in- availability and access to information on maternal and child
teractive systems have standardized UCD, defining it in terms healthcare. The issues identified were prioritized and used as the
of six principles and four activities [10]. The four activities are: basis for design solutions, including health content, based on pa-
(1) understanding and specifying the context of use, (2) specify- per mockups and mobile application prototypes. The paper
ing user requirements, (3) producing design solutions to meet mockups, prototypes, and initial content were used as additional
user requirements, and (4) evaluating the design against the re- instruments to gather more user requirements. In preparing to
quirements. develop the prototypes, the content development team was
In order to understand and specify the context of use, we first tasked to compile relevant content based on healthcare materials
conducted interviews and focus group discussions with HEWs made available by the Ethiopian Ministry of Health (MoH). The
with the objective to discuss issues related to their sociocultural initial version of the health content was developed with refer-
conditions, activities at their health posts, and the community ence to national maternal- and child health–related guidelines.
outreach, infrastructure, and challenges they encounter in their The major topics covered were antenatal care, postnatal care,
daily work, as well as the level of help provided by various vaccination, nutrition, hygiene, and baby sickness. We also at-
stakeholders. Also discussed were selected topics in maternal tempted to align content with user stories that we had already
and child health education, major knowledge gaps on health is- collected and identified during the preliminary investigation. In
sues in the community, and level of ICT utilization (such as total, over 100 digitized content slides were prepared for use in
smart phone usage, computer resources, and barriers). Prior to the mobile prototype application. The content was prepared in
the start of the interviews, we introduced the project objectives both text and audio formats.
and focus and participatory design approaches. In the process, To evaluate the design against the requirements, we gathered
we informed the health workers that the idea was to jointly de- feedback from the use of both the paper prototypes and the mo-
sign the envisaged technology platform with them and explained bile prototype applications in several sessions with prospective
that nothing without their consent would be developed for use.
13
users: mothers, HEWs, and health experts. In these sessions,
groups of five to 10 users used the system and provided feedback
to members of our user study group. We used this feedback in
early stages to revise requirements. In later stages, we gave this
feedback to the software development group in order to enhance
or change specific aspects. For example, we use icons for the
user interface on the mobile phones for mothers. Mothers mis-
understood or simply found unacceptable some of the icons. For
example, an icon showing a sick baby in a bed was not accepted,
because in rural areas of Ethiopia, children do not sleep in beds.
Another example is an icon of a child washing hands under a
faucet. Again, this situation is not common in rural areas. Moth-
ers misunderstood the icon as a child playing with bubbles.
Evaluation has shown that proper design increased the
chance that mothers would be interested in using these applica-
tions. According to the observations we made, most were eager
to actively try the prototypes and provide feedback, despite their
indifference and relatively limited participation during the ear-
lier preliminary investigation. We assume that the fact that they
were able to have a hands-on experience contributed to their ac- Fig. 1. Mothers’ App
tive participation. It was also surprising to learn that even those
mothers who claimed to be illiterate had no difficulty identifying
numbers from 1 up to 10.
An example requirement is that mothers should have an ap-
plication on their mobile phones that contains basic maternal and
child health care information. They should be able to listen to
the available information in their own language, supported by
video. Fig. 1 shows the user interface of the mothers’ app, which
contains icons that mothers can use to get information on spe-
cific content areas such as antenatal care, nutrition, hygiene, and
sickness. Navigation information is shown using images in order
to allow illiterate users to easily navigate. Content itself is shown
in the local language and read out loud, again to allow illiterate
users to understand the information. Fig. 3 shows members of
the user study group instructing mothers in the use of the Fig. 2. HEW System
TEMACC app on their phones.
Another requirement from a different group of users was that
information about pregnant women should be recordable. Fig. 2
shows a dialog for HEWs to enter that information. The HEW
system is web-based and can be used on personal computers or
on tablets. The use of tablets is important, because HEWs visit
mothers and pregnant women in their villages and need to be
able to access relevant information while on the move.
V. LESSONS LEARNED
We now present the lessons learned from the described field
studies with respect to the software development context, the us-
age context, and the content of the software systems.
A. Software Development Context
From our experiences in the TEMACC project so far, the
main challenges with respect to software development in devel- Fig. 3. Members of the User Study Group Instructing Mothers
oping countries are:
• Scarcity of human resources in rural areas, particularly • Poor infrastructure in terms of electric power reliability
regarding ICT experts; and network availability; and
• Poorly educated software developers and service or • Lack of established organizational structures,
maintenance personnel in rural areas; especially in rural areas.
14
B. Usage Context be made for health professionals working in the health centers
The main challenges with respect to users and the usage con- of these districts to use the web-based systems and the
text of software systems in developing rural areas are: knowledge-exchange platforms to support virtual consultation
and referral services at the expert and professional levels. The
• Low education level of user groups (e.g., mothers poorly third district will serve as a control group with no technology
understand educational materials, healthcare personnel intervention. After nine months of intervention, an end-line sur-
cannot speak English); vey will be conducted to examine the effect of the intervention
and to make comparisons. The adoption of the technology solu-
• Many illiterate users (e.g., mothers cannot read or write tion will also be reflected upon in terms of increasing awareness
or put their signatures on documents); and evaluating the extent to which the community demonstrated
• Users have little or no experience with ICT and need an interest in the innovation as compared to other available op-
training on the basic operation of computers (e.g., users tions.
have difficulty operating tablets and lack awareness that
information published on the Internet may not be relia- ACKNOWLEDGMENTS
ble); and The TEMACC project is funded by APPEAR, the Austrian
Partnership Program in Higher Education and Research for De-
• Cultural and gender issues (e.g., mothers feel uncomfort- velopment. APPEAR is a program of the Austrian Development
able taking advice from a younger person or from a Cooperation (ADC) and is implemented by OeAD, the Austrian
woman who does not have children, mothers in general Agency for International Cooperation in Education and Re-
do not trust healthcare personnel, mothers are not al- search.
lowed to use their husband’s mobile phone).
C. Health-Related Content REFERENCES
[1] U. Abelein and B. Paech, “Understanding the Influence of User
The main challenges with respect to the content of the soft- Participation and Involvement on System Success: A Systematic Mapping
ware systems in developing rural areas are: Study,” Empir Software Eng (2015) 20: 28.
[2] Degif T., Y. Dittrich, and Mesfin K., “Usability Challenges in an
• Content selection, (e.g., health content can be selected Ethiopian Software Development Organization,” Proceedings of the
only from approved material by the Ethiopian MoH); Workshop on Cooperative and Human Aspects of Software Engineering
(CHASE 2016), Austin, TX, 2016, pp. 114-120.
• Content organization and structure (e.g., mothers were
[3] Degif T., Y. Dittrich, and Mesfin K., “Contextualizing User Centerd
often observed to simply click on numbers without first Design with Agile Methods in Ethiopia,” Proceedings of the 2017 IEEE
listening to the content, so content needs to be sepa- AFRICON, Cape Town, 2017, pp. 911-916.
rated into small, easy-to-understand chunks); and [4] Degif T., Y. Dittrich, Mesfin K., C. Ardito, and R. Lanzilotti, “User
Involvement and Usability Evaluation in Ethiopian Software Orga-
• Cultural issues related to content (e.g., mothers did not nizations,” The Electronic Journal of Information Systems in Developing
immediately identify an image of a baby getting vac- Countries (2017), 83: 1-19.
cinated as such). [5] C. Macdonald, Cathy, “Understanding participatory action research: A
qualitative research methodology option,” Canadian Journal of Action
VI. CONCLUSION AND FUTURE WORK Research (2012), 13: 34-50.
We are conducting a research project to explore the potential [6] D. A. Norman and S. W. Draper, User Centered Design: New Perspec-
tives on Human-Computer Interaction, L. Erlbaum Associates Inc., Hills-
of ICT to improve healthcare access and quality for mothers and dale, NJ, USA, 1986.
children in rural communities. So far, we have elicited user re- [7] J. Tacchi, D. Slater, and G. Hearn, Ethnographic Action Research,
quirements and developed a set of software systems for mothers, Unesco, 2003.
HEWs, and health professionals. UCD was, it turned out, an ab- [8] Tesfaye B., “Reflective Steps: A Collaborative Learning Oriented
solute necessity to identify users’ needs. Compared to a devel- Approach to Software Development and Process Improvement,” PhD
oped-world context, we had to deal with many unfamiliar chal- dissertation, University of Hamburg, 2008.
lenges related to software development, usage, and content, in- [9] K. Vredenburg, J.-Y. Mao, P. W. Smith, and T. Carey, “A Survey of User-
cluding illiterate users, poor connectivity, and challenges related Centered Design Practice,” Proceedings of the SIGCHI Conference on
Human Factors in Computing Systems (CHI '02), Minneapolis,
to content structure. Minnesota, USA, 2002, pp. 471-478.
We are currently conducting a community-based interven- [10] ISO 9241-210:2010, “Ergonomics of Human-System Interaction, Part
tion study until the end of 2019 to assess the impact of the re- 210: Human-Centred Design for Interactive Systems,” International
Organization for Standardization, 2010.
search project and to promote the educational aspect of the re-
[11] Central Intelligence Agency, The World Fact Book, https://www.cia.gov/
search. The study is being performed in a selected district within library/publications/the-world-factbook/geos/et.html, last accessed:
100 to 200 km of Addis Ababa. After conducting a baseline sur- March 13, 2019.
vey, the target groups will be provided with tablets and mobile [12] The World Stats, Usage and Population Statistics, https://www.internet
phones and given appropriate training to access vital infor- worldstats.com/stats1.htm, last accessed: March 13, 2019.
mation on various health-related topics in various formats in- [13] United Nations, Sustainable Development Goals, https://www.un.org/
cluding video, audio, text, and images. Arrangements will also sustainabledevelopment/, last accessed: March 20, 2019.
15