Improving the Modelling of Human-centric Aspects of Software Systems:

A Case Study of Modelling End User Age in Wirefame Designs

Aria YukFan Jim

, Hyun Shim

, Jue Wang

, Lionel Richie Wijaya

, Rongbin Xu

Hourieh Khalajzadeh

, John Grundy

and Tanjila Kanij

Faculty of Information Technology, Monash University, Melbourne, Australia

HumaniSE Lab, Monash University, Melbourne, Australia

{hourieh.khalajzadeh, tanjila.kanij, john.grundy}@monash.edu

Keywords:

Human-centric Aspects, Modeling Frameworks, Wireframes, End-users’ Age, Human-centred Design.

Abstract:

Taking into account the diverse human aspects – gender, age, emotions, personality, language, culture, physical

and mental challenges, etc – is critical towards achieving more human-centric design of software systems.

Human-centric aspects affecting software have long been underestimated or even ignored as a result of the

lack of in-depth capture and understanding during development. The use of technology has become the norm

and the range of users has increased from just adults to children as well as seniors. Modelling frameworks are

methods to represent the way a software system should be deﬁned, and to date, little research has been done

on age-related issues within modelling frameworks. In this paper, we investigate how human-centric aspects

regarding age can be better modelled by extending these modelling frameworks. We introduce an extension to

wireframe-based designs so that they can cater for decisions regarding age within the modelling framework.

We have evaluated this modelling extension using multiple questionnaires as well as usability testing by using

the extended age-modelling wireframe approach to design a news app. Questionnaires were used to evaluate

the requirements of the users and developers for the extended wireframes. Our analysis shows that when using

our extended wireframes, developers can cater for different user types and their accessibility needs easily and

therefore users can use the prototypes with more ease.

1 INTRODUCTION

According to the internet usage report from Pew Re-

search centre in 2019, 73% of United States adults

over the age of 64 access the internet (Pew Research

Center: Internet, Science & Tec, 2019). The average

life expectancy of many high-income countries in the

world such as Australia, Japan, United States is over

75 years (M. Roser and Ritchie, 2019). There is an in-

crease in the average age of internet users which pro-

vides evidence that there is a need for better catering

for elderly users of software systems. Furthermore, a

study by Joyce and Nielsen in 2019 shows that teens,

growing up with technologies, are also complaining

about poor visual designs, such as font size, back-

ground colour and layout of certain websites (Hussain

et al., 2017). The study also compares web-browsing

behaviour between different age groups which indi-

cates that users’ age should be taken into account

when designing software or websites.

When handling human-centric aspects in soft-

ware design, age being just one example, developers

need to carefully consider the limitations and abili-

ties of these end users as well as their working en-

vironment (Kulyk et al., 2007). ”Elderly users can

face issues such as screen readability due to visual

impairments” – which is not usually an issue for

the mostly young developers who design and de-

velop most software (Parker Software, 2019). Con-

ventional modelling languages such as the Uniﬁed

Modelling Language mostly just model technical re-

quirements and (non-human-centric) non-functional

requirements. Human-centric aspects of end users of

software are one of the most signiﬁcant factors to the

success of a software system. However, these have

been largely ignored and not modelled properly dur-

ing the system development process to date (Grundy

et al., 2020).

In this paper, we seek to improve the modelling

process to better meet the human-centric issues re-

garding age. Our objective is to critically analyse the

existing models, extend these models, and eventually

Jim, A., Shim, H., Wang, J., Wijaya, L., Xu, R., Khalajzadeh, H., Grundy, J. and Kanij, T.

Improving the Modelling of Human-centric Aspects of Software Systems: A Case Study of Modelling End User Age in Wirefame Designs.

DOI: 10.5220/0010403000680079

In Proceedings of the 16th Inter national Conference on Evaluation of Novel Approaches to Software Engineering (ENASE 2021), pages 68-79

ISBN: 978-989-758-508-1

build a prototype. For the evaluation, we will iden-

tify whether developers are easily able to create soft-

ware systems with better usability for all age groups

as well as whether the different user types are able

to easily use the software systems. Our research aim

is to investigate how human-centric aspects regard-

ing age can be expressed in current modelling frame-

works, and to this aim, we focus on three key research

questions:

RQ1. Which Modelling Frameworks are Best Fitted

for Adding Human-centric Aspects Regarding Age?–

For example, use case diagrams, class diagrams, wire-

frames, etc. To determine whether creating a new

modelling framework or extending an existing one

will be better, and to ﬁnd out which one is the most

extendable if we choose the later. To do this, we have

gathered results from a developer survey and relevant

research papers.

RQ2. What Progress Can Be Made on Human-centric

Aspects Regarding Age for Requirements Collection

and Modelling? – How can we best model user

characteristics regarding age in software requirements

and/or design models? For example, by enlarging

fonts based on age, simpler user interfaces for el-

derly people, explanation on certain keywords, etc.

We need to deﬁne the age range for each of the age

groups as well as identify human-centric aspects that

can be improved for each of the age groups.

RQ3. What Application Domains Would Beneﬁt from

Human-centric Aspects Regarding Age? – For exam-

ple, common applications such as news apps, discus-

sion forums, social media, etc. We wanted to ﬁnd out

whether there are certain domains that can best use or

extend our new modelling framework, and to ﬁnd any

patterns or reasons that make this so.

Section 2 brieﬂy explains the methodologies used

throughout the research, and provides an overview of

the background for choosing age as the main objective

among all the human-centric aspects for this study.

We present our approach in Section 3 and our eval-

uation approach in Section 4. Section 5 presents our

extended modelling framework including examples of

the artefacts resulting from the research. We also dis-

cuss evaluation responses from our developer survey

questionnaires. Section 6 provides a summary of the

research papers that are related to our study. We will

ﬁnally conclude the paper in Section 7.

2 METHODOLOGY

We reviewed a number of research papers related to

extending modelling languages to support the mod-

elling of different human-centric aspects, includ-

ing modelling emotions, age, culture, language etc

(Curumsing et al., 2019; Kamalrudin et al., 2012;

Spichkova et al., 2015; Alves et al., 2020). We inves-

tigated how the languages were extended and which

human-centric aspects were modiﬁed in these exist-

ing research papers. We also reviewed various works

that extended existing modelling languages to capture

additional non-functional characteristics (Goncalves

et al., 2018; El-Attar et al., 2015; Grundy and Pa-

tel, 2001). We investigated what aspects could be

changed in the existing modelling languages, in or-

der to capture information about different age groups

of end users. After our analysis, we developed a set of

extensions to the widely used wireframe based design

notation (de Lange et al., 2020) in order to model dif-

ferent end user ages, age-related implications on the

design, and different design decisions based on user

age characteristics.

To evaluate our extended design modelling nota-

tion we used a set of different questionnaires, aimed

at different classes, i.e., users and developers. We pro-

vided multiple solutions to modelling software design

decisions regarding human-centric aspects related to

age. We wanted to ask target end-users to identify the

ones that are good for the age group, in their opinion.

We used a range of target end users from different

age groups to evaluate a prototype based on our aug-

mented wireframe design models. Using their feed-

back, we checked to see which solutions are the best

ﬁt for our prototype use cases.

Our extended wireframe modelling frameworks

are to be used by Software Engineers to create an end-

product that is more age-aware and supports different

interfaces and interface components for differently-

aged end users. We prepared a different set of ques-

tions to ask a range of developers their opinion on the

new modelling framework. We wanted to identify key

current issues that can be found with the pre-existing

model frameworks not supporting modelling of end

user age. We also wanted their opinion on the new

framework and whether it helped them in address-

ing these issues. Using the questionnaire results, we

have reﬁned our extended wireframe-based design ap-

proach to better match the needs of the developers.

3 OUR APPROACH

There were multiple modelling frameworks that could

be extended to support end user age difference mod-

elling in our research. We decided that extending

wireframe-based design models would be the best op-

tion as most of the changes required for the different

age groups were based on the application user inter-

Improving the Modelling of Human-centric Aspects of Software Systems: A Case Study of Modelling End User Age in Wirefame Designs

face. Wireframes provide a simple way for develop-

ers to create the basic design of an application and our

extension point was to create a workﬂow that a devel-

oper can follow for each of the different age groups.

In recent years, various other works have looked to

extend the use of wireframe-based UI design and to

suggest wireframe-based designs or partial designs

(de Lange et al., 2020; Chen et al., 2020). None

have focused on supporting different design decisions

based on end user ages.

Based on the research review, our age groups are

deﬁned as: 1) Kids; Under 15 2) Adults; Aged be-

tween 15 and 64; and 3) Seniors: Aged above 64.

To model different ages, age-related design consid-

erations, and age-related design differences to wire-

frame models, we added three features to the original

wireframe modelling framework:

1. Branching (Branch a screen into two or three path-

ways)

2. Merging (Merging two or three pathways into a

single pathway)

3. Splitting (Splitting a screen into multiple screens)

Figure 1: Example of branching.

Figure 1 shows an example of branching. A single

pathway has been branched into two pathways. This

is used when there is a need for different user interface

or functionality for the same screen for different aged

users. It thus gives developers a tool to express the

need for a change in a single screen for different ages.

Figure 2 shows an example of merging. Two path-

ways have been merged into a single pathway. This

is useful when multiple pathways arrive at a speciﬁc

Figure 2: Example of merging.

screen that will be the same for multiple age groups

that have been previously branched out. It saves the

developers from creating the same screen design for

different pathways. It also helps show that multiple

screens will be the same across different age groups.

Figure 3: Example of Splitting.

Figure 3 shows an example of splitting. A sin-

gle screen is split into two screens. This is useful for

ENASE 2021 - 16th International Conference on Evaluation of Novel Approaches to Software Engineering

age groups that require additional assistance in getting

a task done. For example, adults may prefer a sin-

gle register page as it allows them to quickly register.

However, kids or seniors may need additional help

when registering, e.g. by using a step by step multi-

ple screen approach. Using multiple screens will give

developers more room to explain key features within

each screen, make items larger, or choose different

interaction components. This will be a trade-off be-

tween speed and ease-of-use.

4 EVALUATION APPROACH

4.1 Prototype News App

We created a prototype news app based on our ex-

tended wireframe approach to evaluate how easy it

was for a developer to use this extended wireframe

model. The prototype was also used to evaluate the

ﬁnal prototype system by different age group target

end-users. We chose a news app as the prototype as

(i) most age groups use them or are interested to use

them; (ii) despite their seaming simplicity, they often

have quite complex interfaces and design decisions

are not always ﬁtting differently aged users; (iii) the

interface design is reasonably detailed but not over-

whelmingly so; and (iv) many lessons from news app

development can be applied to social media, commu-

nications, education and other widely used apps (Con-

stantinides et al., 2015). Hence, the news reading app

domain contains enough variation points between the

three age groups to try out the use of our extended

wireframe modelling language features.

We used our extended wireframes to design and

develop a Figma-based prototype. The extended wire-

frame designed to create the news app prototype is

shown in Appendix. By using the extended modelling

framework we captured key human-centric aspects re-

garding the age of the users and designed solutions to

accommodate these. This news application has differ-

ent functionality and user interface for different age

groups of its users. In this prototype, we decided

to keep the colours for each of the branches speciﬁc

to the age groups even if a screen was for all three

user types. For example, a screen that has not been

branched out which is used for all three user types

will still have different colour themes to allow for a

more consistent experience for the end user.

In the extended wireframe design model, we have

many common pages for all three age groups. We

needed to ﬁnd a balance point for these age groups

such as use of font size, user interface complexity,

and any related points to ensure that the design for the

common pages is clear enough for use from all three

age groups. An age selector is presented to users at

the start, as shown in Figure 4.

Figure 4: Age category selection page.

4.1.1 Children

For the kids application, the key aim was to make

the application enjoyable and attractive enough for the

children. Some user interface design decisions that

needed to be considered are:

• Colour: Research showed that children prefer

colours like red and yellow and thus we used a

rainbow gradient colour schema to catch the eyes

of the children.

• Font Style: Use a fun looking font size to help

children stay focused.

• Icon Selection: We chose a more cartoonish icon

for children to cater for their shorter attention

span.

Figure 5 shows a user interface designed for kids,

based on that speciﬁed in our extended wireframe.

4.1.2 Adults

For the adult application, the key aim was to make

the app clear and simple so that they will be able

to quickly navigate between pages without too much

overhead information. Thus key design decisions for

the adult app user interface design included:

• Font Size: Normal size.

Improving the Modelling of Human-centric Aspects of Software Systems: A Case Study of Modelling End User Age in Wirefame Designs

Figure 5: User interface for kids adapted from the extended

wireframes.

• Colour: Simple and conventional colours. More

professional compared to the kids version.

Figure 6 shows a user interface designed for adults,

using our extended wireframe.

4.1.3 Seniors

The senior application is similar to the adult appli-

cation, but we needed to make sure that everything is

large enough, easily accessible, and that the user in-

terface is simple enough. We added some additional

functionality to help aid seniors so that they can com-

fortably use aspects of the technology that they may

not be familiar with. Some key design decisions for

the senior age group include:

• Colour: Similar to adults, use simple and conven-

tional colours.

• Font Size: use a larger font size to aid visual im-

pairments.

• Voice Control: Senior users may have trouble typ-

ing on a mobile device or reading smaller font

sizes and speech-to-text and text-to-speech func-

tionality would be good for them.

• Larger Components: Make items such as buttons,

and links large enough for them to easily click.

Figure 7 shows a user interface designed for seniors,

using our extended wireframe.

We also provided a toggle setting, shown in ﬁg-

ure 9, so that users can customise the app styles such

Figure 6: User interface for adults adapted from the ex-

tended wireframes.

Figure 7: User interface for seniors adapted from the ex-

tended wireframes.

as the font size, font style, etc. This was necessary

because there may be some users who do not require

the customizations that were categorised within their

user age groups. Other customizations that were de-

ENASE 2021 - 16th International Conference on Evaluation of Novel Approaches to Software Engineering

termined necessary during our research are speciﬁed

in the extended wireframes. This allows developers

to choose whether to include a customization within

their applications.

Figure 8: Toggle setting.

4.2 Usability Evaluation

To evaluate the news app prototype’s usability, a us-

ability test was conducted to gain insights into what

went well and what needs to be improved in our ex-

tended wireframe modelling approach. Users were

asked to complete two tasks:

• Change the user mode along with some settings to

match their preferences; and

• Find an article they are interested in, save the ar-

ticle, and open the article again using the saved

articles page.

The details of the tasks given to the users were pur-

posely given without too much detail on how to nav-

igate within the application. This was so that we can

see how a new user will act when using this applica-

tion. These tasks helped us to gather user data on key

age-based design decisions in the news app:

• Whether each user type is able to use the inter-

action ﬂow designed for them within the applica-

tion, as well as change the settings to match their

accessibility needs; and

• Whether the navigation supported within the ap-

plication is easily understandable to new users.

The testers were asked to record whether the tasks

were successfully completed as well as to describe

any troubles that they encountered during performing

the tasks. We conducted a preliminary test, and us-

ing the results, we enhanced our extended wireframe

models to increase the usability of the prototype. We

plan to conduct a more comprehensive usability eval-

uation with developers to evaluate whether the ex-

tended wireframe is applicable to different domains

and can be easily used to create prototypes. We also

plan to conduct usability evaluations with end-users

from various age-groups to evaluate whether the cre-

ated prototype actually increased the usability of the

news app prototype for each age group.

5 RESULTS

In this section, we present the results of our survey

of developers and users. We collected a total of 27

responses – 6 from developers, 21 from app end users.

5.1 Evaluation Results: Developer

Questionnaires

Six developers responded to our survey – 3 with 5 to

10 years development experience, 2 with 1 to 4 years,

and 1 with more than 20 years of experience. Among

all of these developers, 4 said they think there is a

need for human-centric aspects regarding age in mod-

elling frameworks e.g UML diagrams, Wireframes,

User stories, etc. One had never though of this issue,

and one though it is not important. We collected a list

of approaches and frameworks they currently use to

deal with age related issues of the users in the soft-

ware that they develop. This included use of wire-

frames (3 of the developers), UML (3), user stories

(2), use cases (2), and BPMN diagrams (1).

We asked the 6 developers whether they have en-

countered any issues regarding age during software

development. Based on their responses, as shown in

ﬁgure 9, one of the key issues is that the majority of

developers do not address differing age of their end

users at all in their software development. To ad-

dress this issue in this study, we captured a lot of end

user requirements from our end user survey question-

naires. The second issue is that some developers be-

lieve it is hard to deﬁne an “age group”. To address

this, we decided to research more about the target au-

dience of the particular software and then deﬁned the

age groups accordingly.

We asked developers what tool(s) they used to

help them address differing age issues of their app

and web site end users. They mostly stated they never

used any tools. Some suggested that although they

Improving the Modelling of Human-centric Aspects of Software Systems: A Case Study of Modelling End User Age in Wirefame Designs

Figure 9: Developer responses on designing in issues regarding different end user age during development.

did not use any tools in particular, they tried to ad-

dress age related problems such as security restric-

tions in the identity server (e.g. some websites are for

adults only); and informally captured non functional

requirements in text, annotate the odd use case/user

story. For a news application, age selection is impor-

tant to restrict inappropriate news from the children.

Icons, text, and background colours might also need

to be changed to accommodate different types of users

sight limitations and preferences based on their age.

Regarding what other human-centric aspects they

think could be a good addition to better support in

software modelling frameworks, developers included

gender, culture, end user language, physical and men-

tal challenges of users, accessibility needs of end

users, Convenience of usage, usability test (user test),

cognitive load (mental effort) test (including perfor-

mance measures in terms of user’s response time

taken to complete a task and its accuracy, or physi-

ological measures, e.g. pupil dilation and blink rate).

5.2 Evaluation Results: End-user

Survey

We received 21 responses, 13 from 18-25 (Young

Adults) respondents, 6 from 26-49 (Adults), and 2

from those 50-64 years old. The age demographics

for the questionnaire participants are mostly young

adults with ages ranging from 18 to 25. We did not

have any younger participants (below 18) and senior

participants (above 65). Most of the participants used

technology for daily activity for a signiﬁcant amount

of time (8 hours+). A majority of the participants

found that using technology is easy for them (14 peo-

ple said they are proﬁcient with it, ﬁve said they are

good with it, and one said they are doing okay with it).

We also found out that the most popular device is a

smartphone used for social media, entertainment and

news apps. We asked if any participant had any acces-

sibility requirements. The majority of people saying

they had visual impairments, such as needing glasses

to read, having short sight, and one person answered

that they needed to use voice output to “read” app text.

5.3 Research Question Answers

RQ1. Which Modelling Frameworks are Best Fitted

for Adding Human-centric Aspects Regarding Age?

– From the result in ﬁgure 10 we can see that most of

the problems are User Interface related issues. This

is one of the main reasons we decided to extend wire-

frames since wireframes provide early visuals that can

help with these problems, they are also easy to adapt

compared to conceptual designs.

RQ2. What Progress Can Be Made on Human-

centric Aspects Regarding Age for Requirement

Collection? – We used questionnaires for collect-

ing data. We created both end user and developer

questionnaires to capture data from multiple angles.

We got a range of feedback from different age groups

that indicated usability enhancement in the prototype

news app for different age groups of target end users.

However, we need to recruit more participants to gain

a larger sample size, especially younger (under 18)

and older (over 65) end users.

RQ3. What Application Domains Would Beneﬁt

from Human-centric Aspects Regarding Age? – Ac-

cording to our survey results in ﬁgure 11, we can see

that our participants are commonly using a range of

different application domains such as social, enter-

tainment, study, etc. These results helped us in de-

ciding on which domain our example extended wire-

frame and prototype would focus on. Generalis-

ing our experiences to those other common domains

would help to show if the approach can enhance dif-

ferent age group usability for them as well.

5.4 Discussion

We chose to extend wireframe design models with in-

formation about the age of target end users and alter-

native user interface design decisions based on these

ages. However, other modelling frameworks such as

interaction and sequence diagrams may help devel-

opers understand the human-centric aspects regard-

ing age further and use similar approaches based on

the wireframe extensions we have created. For ex-

ENASE 2021 - 16th International Conference on Evaluation of Novel Approaches to Software Engineering

Figure 10: Key pain points of using software systems.

Figure 11: Devices used by participants and for what purposes.

ample, ﬁgure 12 contains an interaction diagram that

is adapted from the extended wireframes presented

in this paper. Adapting other human-centric aspects,

such as gender, emotion, and physical support, into

the extended modelling framework presented in this

paper may also be beneﬁcial. Additional research can

be done to increase the range of our user types and

changes to the designs to accommodate their needs.

Using approaches developed in the accessibility re-

search ﬁeld could also help us to better understand

the needs of users with a range of disabilities. The

speciﬁc requirements of these user types need to be

identiﬁed so that the developers can create an appro-

priate pathway and design decisions for those users in

our extended wireframe models.

6 RELATED WORK

We reviewed the key works done in three different

categories related to designing for children, seniors

and extending models to capturing emotions of end

users. The age classiﬁcation we used in this paper is

called Life Cycle Groupings (Statistic Canada, 2017).

According to Statistics Canada, age groups can be de-

ﬁned as: Children aged from 0 to 14 years; youth aged

from 15 to 24; adult aged from 25 to 64 years; and

senior aged from 65 and above. The Australian Bu-

reau Statistics (ABS) uses a similar convention except

they group both youth and adult as working-age pop-

ulation (Australian Bureau of Statistics, 2019), which

we named it as adult throughout the research.

6.1 Children

Mobile Educational Applications are used by (Ma-

sood and Thigambaram, 2015) for usability testing.

By recording with eye tracking glasses, they found

some children had problems working with mobile ap-

plications. The system status is not apparent for them

and they have a hard time ﬁnding out what to do

next. They also had problems where they could not

remember which page or button was accessed earlier.

This children-oriented software needs to more clearly

show the current state of the page, and sometimes

the child users may need some guidance to do the

next step. Help toolboxes and documentation were

identiﬁed as necessary. Another important factor is

to consider whether the buttons and menus are sim-

ple enough for children. Buttons and menu links

should be easily identiﬁed as being clickable while

items such as menu headings and titles should be eas-

ily identiﬁed as being not clickable.

A Fingerprint app (Pan, 2010) is used to describe

Improving the Modelling of Human-centric Aspects of Software Systems: A Case Study of Modelling End User Age in Wirefame Designs

Figure 12: Interaction diagram adapted from the extended wireframes.

how to design the software user interface for the chil-

dren. This work discusses four key points for the vi-

sion element regarding kids – integer vision effect,

functional area design, icon and button design, and

font design. For example, children may not under-

stand the text inside a button but icons can be de-

signed as buttons to show they are clickable.

The work of (Michaels, 1924) and (Boyatzis and

Varghese, 1994) discusses colour preferences of chil-

dren users as well as the effect of colour on children’s

emotions. They did colour tests with children around

6 years old, and found that their top three favorite

colours are yellow, red, and green. Another type of

testing they did was called the story test. If a child

heard a happy story, most of them would select yel-

low. Using this research, they used yellow and red as

the gradient colour for the theme colour, and made the

whole app look funny and attractive.

6.2 Seniors

(Boll and Brune, 2015) provide a set of user interface

design guidelines for people between 55 to 75 years

old. They found 41 participants between 55 and 75

years to ﬁll out questionnaires to research the actual

requirement and the problems they have. They used

the results to make the user interface design guide-

lines. According to the guidelines, for font design,

the sans serif fonts provide good readability. For

colour selection in the main content page, they rec-

ommended a light grey background with black fonts.

Regarding the size of the icons and the buttons, most

of the users reported that the icons are too small, and

that double-clicking a small button is a problem for el-

derly people. The last important thing is the structure

of the page. For example, in menu page, a good menu

structure helps users to navigate through the user in-

terface more easily. The menu needs to be put in con-

ventional positions to make sure the position of the

menus are consistent in the entire software.

ENASE 2021 - 16th International Conference on Evaluation of Novel Approaches to Software Engineering

(Curumsing et al., 2019) focuses on designing

emotion-oriented software, based on the smart home

device for elderly. They used extended Goal models,

Interaction models, Scenario models, Role models,

and Behaviour models to keep track of the “cared for”

to a list of emotions. Analysis of these emotions helps

developers to understand the expectations of an older

adult using the smart home. Using this approach, a

goal model for the smart home device was created.

The model includes different emotions for the elderly

people to help get the elderly people to accept the de-

vice and feel like this is what they need. The soft-

ware can catch the emotion of the user, and analyse

their expectations. Thus, the software can make cor-

responding responses to meet the needs of users. The

software can understand users’ emotions, this could

become one of the determinants of software success.

(Curumsing et al., 2015) demonstrates a case

study of an emergency alarm system for elderly peo-

ple, presenting the entire suite of models for this case

study. They suggested a few important factors on de-

signing the framework and also keeping the interest

of the elderly people. Firstly, in order to encourage

elderly people to adopt technological solutions, they

have to be designed in such a way that they suit the

needs of its users, are easy to use, and cost effective.

The second important factor determining the success-

ful adoption of technology consists of willingness and

enthusiasm for acquiring new knowledge. However,

this is rarely the feelings expressed by elderly people

when it comes to using a system which is linked with

a stigma. For example, some refused to use the pen-

dant because of its visibility to others. It was viewed

as a sign of stigma and old age. These are the way

users perceived technology from an emotional aspect.

(Wagner et al., 2014) explains the impact of age

on usability. It states that there is an increase in the

average age of internet users which provides evidence

that there is a need for catering elderly users for ap-

plications. Usability helps organisations by improv-

ing the job performance, gaining higher productivity

and reduced costs by the users. There are ﬁve con-

ceptualisations of age; Chronological or calendar age,

Functional or performance-based age, Subjective or

psycho social age, Organisational age, and Life span

concept of age. Currently, mainly research chrono-

logical age is used to allow for consistency and com-

parison with the existing literature. We will also be

using chronological age as the other conceptualisa-

tions are mostly related to chronological age.

Finally, (Holzinger et al., 2008) aims to derive

metric-based benchmarks for measuring usability.

This study suggested two aspects of usability: pas-

sive and active interaction. Passive interaction means

users are not directly interacting with it, but it helps

out in the background. Active interaction means users

directly interact with the technology. Several ques-

tions were developed to ensure it meets both passive

and active interaction. They present an analogy be-

tween user anxiety and metrics. For our prototype, it

will be a good metric to evaluate our product’s trust-

worthiness and acceptance.

6.3 Emotions

An emotion orientated software development method

was developed by (Chen et al., 2015). Existing

emotion-aware applications lack accuracy in terms of

emotion recognition due to the small scale of data col-

lected. Through cloud-based computing and cloud

assisted resources, mobiles phones can collect much

more data. Combined with the architecture itself, ap-

plications can recognise user’s emotional changes by

big data analysis. Based on the user’s current emo-

tion, a common list of feedback is generated in the

remote cloud. The information is transferred back to

the local cloud, providing users with personalised ser-

vices. Various in-home devices were used to accom-

modate the user’s emotion. We could modify this ap-

proach and base the categorisation on age group. An

application would analyse the habits of that speciﬁc

age group in order to adapt the application to better

supporting the user, based on their age.

7 SUMMARY

There is a need to incorporate human-centric aspects

into modelling frameworks in order to improve the

suitability of software produced for diverse end users.

There are many existing modelling frameworks, and

we expect that modifying existing modelling frame-

works will be better than creating a completely new

one. Most do not currently support modelling the

age of end users and providing different design so-

lutions for different age groups and needs of end

users. We developed a set of extensions to the com-

monly used wireframe modelling approach to incor-

porate different designs for child, adult and senior

end users. We evaluated our modelling approach with

developers and a prototype news app developed us-

ing our approach with a range of differently aged

end users. Further work includes incorporating other

human-centric aspects into the extended wireframe

model e.g. gender, culture, language, and trying the

same model extension approach in other modelling

frameworks, such as user stories, use cases, and se-

quence diagrams.

Improving the Modelling of Human-centric Aspects of Software Systems: A Case Study of Modelling End User Age in Wirefame Designs

ACKNOWLEDGEMENTS

Support for this work from ARC Laureate Program

FL190100035 is gratefully acknowledged.

REFERENCES

Alves, T., Nat

alio, J., Henriques-Calado, J., and Gama, S.

(2020). Incorporating personality in user interface de-

sign: A review. Personality and Individual Differ-

ences, 155:109709.

Australian Bureau of Statistics (2019). Australian demo-

graphic statistics, jun 2019. https://www.abs.gov.au/.

Boll, F. and Brune, P. (2015). User interfaces with a touch

of grey?–towards a speciﬁc ui design for people in the

transition age. Procedia Computer Science, 63:511–

516.

Boyatzis, C. J. and Varghese, R. (1994). Children’s emo-

tional associations with colors. The Journal of genetic

psychology, 155(1):77–85.

Chen, J., Chen, C., Xing, Z., Xia, X., Zhu, L., Grundy,

J., and Wang, J. (2020). Wireframe-based ui design

search through image autoencoder. pages 1–31.

Chen, M., Zhang, Y., Li, Y., Mao, S., and Leung, V. C.

(2015). Emc: Emotion-aware mobile cloud comput-

ing in 5g. IEEE Network, 29(2):32–38.

Constantinides, M., Dowell, J., Johnson, D., and Malacria,

S. (2015). Exploring mobile news reading interactions

for news app personalisation. In Proceedings of the

17th International Conference on Human-Computer

Interaction with Mobile Devices and Services, pages

457–462.

Curumsing, M. K., Fernando, N., Abdelrazek, M., Vasa,

R., Mouzakis, K., and Grundy, J. (2019). Emotion-

oriented requirements engineering: A case study in

developing a smart home system for the elderly. Jour-

nal of Systems and Software, 147:215–229.

Curumsing, M. K., Lopez-Lorca, A., Miller, T., Sterling, L.,

and Vasa, R. (2015). Viewpoint modelling with emo-

tions: a case study. International Journal of People-

Oriented Programming (IJPOP), 4(2):25–53.

de Lange, P., Nicolaescu, P., Rosenstengel, M., and

Klamma, R. (2020). Collaborative wireframing for

model-driven web engineering. pages 373–388.

El-Attar, M., Luqman, H., Karpati, P., Sindre, G., and Op-

dahl, A. L. (2015). Extending the uml statecharts no-

tation to model security aspects. IEEE Transactions

on Software Engineering, 41(7):661–690.

Goncalves, E., Castro, J., Araujo, J., and Heineck, T.

(2018). A systematic literature review of istar exten-

sions. Journal of Systems and Software, 137:1–33.

Grundy, J., Khalajzadeh, H., and Mcintosh, J. (2020). To-

wards human-centric model-driven software engineer-

ing. pages 229–238.

Grundy, J. and Patel, R. (2001). Developing software com-

ponents with the uml, enterprise java beans and as-

pects. Proceedings 2001 Australian Software Engi-

neering Conference, pages 127–136.

Holzinger, A., Searle, G., Kleinberger, T., Seffah, A., and

Javahery, H. (2008). Investigating usability metrics

for the design and development of applications for the

elderly. In International Conference on Computers for

Handicapped Persons, pages 98–105. Springer.

Hussain, A., Abd Razak, M. N. F., Mkpojiogu, E. O., and

Hamdi, M. M. F. (2017). Ux evaluation of video

streaming application with teenage users. Journal of

Telecommunication, Electronic and Computer Engi-

neering (JTEC), 9(2-11):129–131.

Kamalrudin, M., Grundy, J., and Hosking, J. (2012). Mara-

maai: tool support for capturing and managing consis-

tency of multi-lingual requirements. pages 326–329.

Kulyk, O., Kosara, R., Urquiza, J., and Wassink, I. (2007).

Human-centered aspects. In Human-centered visual-

ization environments, pages 13–75. Springer.

M. Roser, E. O. O. and Ritchie, H. (2019). Life expectancy.

https://ourworldindata.org/life-expectancy.

Masood, M. and Thigambaram, M. (2015). The usability

of mobile applications for pre-schoolers. Procedia-

Social and Behavioral Sciences, 197:1818–1826.

Michaels, G. M. (1924). Colour preference according to

age. The American Journal of Psychology, 35:79–87.

Pan, X. (2010). Research of iphone application ui design

based on children cognition feature. In 2010 IEEE

11th International Conference on Computer-Aided In-

dustrial Design & Conceptual Design 1, volume 1,

pages 293–296. IEEE.

Parker Software (2019). Key considera-

tions for making age-friendly software.

https://www.parkersoftware.com/blog/key-

considerations-for-making-age-friendly-software/.

Pew Research Center: Internet, Science & Tec

(2019). Internet/broadband fact sheet.

https://www.pewresearch.org/internet/fact-

sheet/internet-broadband/.

Spichkova, M., Zamansky, A., and Farchi, E. (2015). To-

wards a human-centred approach in modelling and

testing of cyber-physical systems. pages 847–851.

Statistic Canada (2017). Age categories, life cycle group-

ings. https://bit.ly/3dopioH.

Wagner, N., Hassanein, K., and Head, M. (2014). The im-

pact of age on website usability. Computers in Human

Behavior, 37:270–282.

APPENDIX

The full extended wireframe designed to create the

news app prototype is shown in ﬁgure 13.

ENASE 2021 - 16th International Conference on Evaluation of Novel Approaches to Software Engineering

Figure 13: Extended wireframe designed to create the news app prototype.

Improving the Modelling of Human-centric Aspects of Software Systems: A Case Study of Modelling End User Age in Wirefame Designs