Usability Testing on Android-based Mobile Application ”Smart Assistant

Diabetes”

Firman Sadewo Priatmadji

, Ike Pertiwi Windasari

and Kurniawan Teguh Martono

Computer Engineering, Faculty of Engineering, Diponegoro University, Semarang

Keywords:

Usability, Mobile Application, Diabetes, SUS, SEQ.

Abstract:

Diabetes is a chronic disease characterized by disorders of blood sugar regulation in the body due to lack of

insulin production by the pancreas, lack of the body’s response to insulin, and the inﬂuence of other hormones

that inhibit insulin performance. Mobile application could help people with diabetes to assist their needs

such as medicine reminder, exercise guidance, or blood sugar tracking. To maximize their usage, the apps

needs to be usable and easy to use. This paper aims to test the usability of Smart Assistant Diabetes Mobile

Application based on SUS and SEQ method. The result from SEQ method has median value of 6, so the apps

can be classiﬁed as Easy. Using SUS method, the average SUS values were 71.08 and can be classiﬁed as

Good.

1 INTRODUCTION

Diabetes mellitus, commonly known as diabetes, is

one of the diseases with the highest number of suffer-

ers in the world with the number of sufferers reach-

ing 422 million (Organisation, 2017). Indonesia, has

the six-highest diabetes prevalence rate after PRC, In-

dia, and USA with 10.3 million adults living with dia-

betes(Federation, 2019). The diabetes prevalence has

also rising from 6,9% in 2013 to 8,5% in 2018.

Diabetes is a metabolic disorder characterized by

chronic hyperglycaemia and metabolic disorders of

carbohydrates, fats and proteins caused by abnor-

malities in insulin secretion, insulin action or both

(Kardika et al., 2015).

Many people assume that effective diabetes treat-

ment is only enough with insulin injections or drugs

only. In fact, this kind of treatments require a quite

large cost. Another important effort that is quite easy

to do and does not require a large cost but less noticed

by people is to regulate a healthy diet and exercises,

especially diabetes gymnastic and other simple exer-

cises regularly.

The ”Smart Assistant Diabetes” was built with the

aim of designing a personal data assistant application

interface for Diabetes patient by combining previous

research and modifying it by adding various features

that are different from previous studies. This apps is

built on android platform. Android is one of the mo-

bile device platforms with the most users (Atmodjo

and Krisjanti, 2016). In addition, this study also tested

the usability and ease of its interface by using the Sin-

gle Ease Question (SEQ) and System Usability Scale

(SUS) methods that proposed to users.

2 LITERATURE REVIEW

There are several applications that helps people with

diabetes (Perwira, 2012)(Widiastuti and Syahbani,

2015). (Perwira, 2012) made an android apps that

helps users to manage their diet with the consulta-

tion feature of the daily diet. This apps helps users

in maintaining their daily diet. (Widiastuti and Syah-

bani, 2015) made an Android-based apps that was

similar to the previous application, but equipped with

a daily calorie recording feature. This apps also have

a medicine and meal time reminder. Kalpajar (2018)

made a similar application with a diet menu consulta-

tion feature and a daily calorie calculator feature that

was made more detailed.

Diabetes management is not only about medicine

but also healthy daily diet and exercise, especially di-

abetes and light exercise regularly (Yendi and Ad-

wiyana, 2014). Healthy diet and doing diabetes

and light exercises regularly will help control patient

blood sugar level and also avoid various kinds of other

disease complications (Novita et al., 2018). In ad-

dition, diabetes exercise itself helps insulin work in

diabetics because through movements in diabetes ex-

Priatmadji, F., Windasari, I. and Martono, K.

Usability Testing on Android-based Mobile Application "Smart Assistant Diabetes".

DOI: 10.5220/0009431300640071

In Proceedings of the International Conferences on Information System and Technology (CONRIST 2019), pages 64-71

ISBN: 978-989-758-453-4

ercise makes sugar in the blood ﬂow through mus-

cle cells which is then converted into energy so that

blood sugar levels in the body of diabetics also de-

crease (Sharoh and Salmiyati, 2017)(Nurahmatya and

Asnindari, 2014) (Rehmaita et al., 2017)(AFRIZA,

2015).

An apps with low usability can frustate user

and stop user from using the apps (Widodo et al.,

2017). Some usability research has been applied

in healthcare (Davis and Jiang, 2016)(Alturki et al.,

2017), also in academic (Ardiansyah and Ghazali,

2016)(Santoso, 2018). Some usability testing method

used in this research are Nielse usability testing

(Widodo et al., 2017), Single Ease Question (SEQ)

(Alturki et al., 2017) and a mix of System Usabil-

ity Scale (SUS) and SEQ (Ardiansyah and Ghazali,

2016)(Santoso, 2018).

Smart Assistant Diabetes for people with diabetes

was created by combining features in previous stud-

ies and existing applications and adding new feature

them such as the meal and medicine reminder feature.

This application also features a food catalogue and its

calories that are suitable for diabetes and BMI (Body

Mass Index) calculator. Furthermore, this application

also has a blood sugar tracking feature that functions

to monitor the user’s blood sugar level. Finally, this

application also features a simple guide to do some

exercises suitable for users with diabetes. In general,

this study discusses how to test a Smart Assistant Dia-

betes mobile application interface with its various fea-

tures and the level of usability and ease of use using

the SUS and SEQ methods.

3 METHOD

Data is collected by giving questionnaires to thirty re-

spondents with diabetes. Then the test is carried out

by lending an Android device containing the Smart

Assistant Diabetes apps to the respondent. Respon-

dents test the application by carrying out tasks in ac-

cordance with the tasks on the questionnaire given

which are then ﬁlled out after trying the application.

In this test, there are two methods used, namely

Single Ease Question (SEQ) and System Usabil-

ity Scale (SUS). The Single Ease Question (SEQ)

method is used to measure the ease felt by users af-

ter completing a given task. This method is used be-

cause the implementation is fast and does not require

a long time because the questions are asked immedi-

ately. System Usability Scale (SUS) method is used

to measure how high the usability and the acceptabil-

ity levels of application design are developed (Ardian-

syah and Ghazali, 2016). This method is also used be-

cause the test is very practical and easy, but the results

remain valid and can be justiﬁed.

The Single Ease Question (SEQ) method used in

this study contains 9 tasks that must be performed by

the user. SEQ are administered at the end of every

task in a test session. The task is coded F1 to F9.

User rate the difﬁculty of the completed task, from

Very Easy to Very Difﬁcult on a 7-point rating scale.

We then calculate the average of each task from the

respondents with Equation 1. We then calculate the

median as described in Equation 2 for even data and

Equation 3 for odd data (Wetzlinger et al., 2014).

Scorepertasks =

taskvaluesamount

numbero f respondents

(1)

Median = (Xn2 + Xn2 + 1)/2 (2)

Median =

(Xn + 1)

(3)

Information :

X = data sequence

n = amount of data

As for the scale of values in the SEQ method, it

can be seen inFigure 1.

Figure 1: Scale Value of SEQ test.

List of tasks and questions of the SEQ method can

be seen in Figure 2.

Usability Testing on Android-based Mobile Application "Smart Assistant Diabetes"

space

Figure 2: Single Ease Question (SEQ) task list.

In System Usability Scale (SUS) method, there

are 10 questions with ﬁve response options for re-

spondents that use a Likert scale with a value of

1 (strongly disagree) to 5 (strongly agree). In this

method, each question with an odd number (1,3,5,7,9)

has a positive tone and an even numbered question

(2,4,6,8,10) has a negative tone. For odd numbered

questions, the value of the respondent is calculated us-

ing Equation 4. Whereas in the even numbered ques-

tions, the value of the respondents was calculated us-

ing Equation 5. To calculate the ﬁnal value of SUS,

can be seen in Equation 6 (Bangor et al., 2009).

OddQuestionValue = X − 1 (4)

EvenQuestionValue = 5 − X (5)

SUSValue = (OddQuestionValueTotal+

EvenQuestionValueTotal)x2,5 (6)

Information:

X = data sequence

The assessment of testing using this SUS method

starts from 0 to 100. For the scale of the value of this

method can be seen in the Figure 3.

space

Figure 3: SUS Score Scale.

As for the list of SUS questions, it is shown in

Figure 4 below.

Figure 4: List of System Usability Scale (SUS) Questions.

4 RESULTS

The interface of this application is shown in the fol-

lowing ﬁgures.

4.1 Login Display

This page is displayed after the splash screen and

serves as a page for account authentication. On this

page, the user input the registered email and pass-

word. If it fails or the user enters the wrong email

or password, there will be a notiﬁcation that the email

or password entered is incorrect. Under the button,

there is also the text ”Daftar” (Register) which can be

selected by the user as navigation to go to the Register

page. The login display is shown in Figure 5 below.

CONRIST 2019 - International Conferences on Information System and Technology

space

Figure 5: Login Display.

4.2 Exercises Info Display

This page is the main page when the users access the

application. On this page, there are ﬁve main sub

menus regarding exercises information suitable for

people with Diabetes Mellitus, namely Muscle Train-

ing, Swimming, Cycling , Jogging, Yoga and Gym-

nastic . On the Muscle Training and Yoga and Gym-

nastic sub menus, there are several more sub menus

containing info and exercise programs that can be

selected by the user. The Exercises info display is

shown in Figure 6 below.

space

Figure 6: Exercise Info Display.

4.3 Body Mass Index (BMI) Calculator

Display

This page contains the Body Mass Index (BMI) cal-

culator. On this page, there are forms ”Height” and

”Weight” that can be ﬁlled by the user. Then there is

the ”Calculate” button to display the calculation re-

sults of the value of the Body Mass Index (BMI) en-

tered by the user. The Body Mass Index (BMI) calcu-

lator display is shown in Figure 7 below.

Usability Testing on Android-based Mobile Application "Smart Assistant Diabetes"

space

Figure 7: Body Mass Index (BMI) Calculator Display.

4.4 Blood Sugar Graph Display

This page contains a graph of the random and fasting

blood sugar development based on the value and date

entered by the user. On this page there are two sub

menus namely ”Sewaktu” (random blood sugar) and

”Puasa” (fasting blood sugar). On both sub menus,

there is a blood sugar form and a date that the user can

ﬁll in. Then there is the ”Input Data” button which

functions to display the graph value and date along

with the indicator based on input from the user. Blood

Sugar Graph display is shown in Figure 8 below.

space

Figure 8: Blood Sugar Graph Display.

4.5 Food Database and Calories Display

This page contains a list of foods and their calories

that can be seen by users. On this page there is a list

of various foods along with an explanation of calories

and weight that serves to help people with Diabetes

Mellitus measure the food they want to consume. The

food database and its calories display is shown in Fig-

ure 9 below.

CONRIST 2019 - International Conferences on Information System and Technology

space

Figure 9: Food Database and Calories Display.

4.6 Reminder for Mealtime, Take

Medication, and do Exercises

Display

This page contains a list of reminder alarms for meal-

time, take medication and do exercise that can be set

by the user. On this page there are three sub menus

namely ”Makan” (Eat), ”Minum Obat” (Take Med-

ication), and ”Olah Raga” (Exercise). In each sub

menu, there is a list of each alarm that has been set and

a button to add a new alarm. If the button is selected,

a new page will open where the user can change the

name and set the time of the alarm as desired by the

user. The reminder for mealtime, take medication,

and do exercises display is shown in Figure 10 below.

space

Figure 10: Reminder for Mealtime, Take Medication, and

do Exercises Display.

We conduct usability test using the SEQ method

questionnaire that is given to thirty respondents. User

rate the difﬁculty of the completed task, from Very

Easy to Very Difﬁcult on a 7-point rating scale. The

obtained data and results is described in the Figure 11.

To get the value of each task in Table 4, we calcu-

late the average values of each task using Equation 1.

We then calculate the median as described in Equa-

tion 2 for even data and Equation 3 for odd data. The

results are described in Figure 12.

Usability Testing on Android-based Mobile Application "Smart Assistant Diabetes"

space

Figure 11: SEQ test results.

Figure 12: Average values of each tasks using the SEQ

method.

Figure 13 summarize the results of each task from

Single Ease Question (SEQ) method. We then calcu-

late the scale value by ﬁnding the median using Equa-

tion 2 and Equation 3 so that the median results ob-

tained from all tasks from F1 to F9 are 6. That average

value of each task indicates the usability as ”Easy”.

Figure 13: Graph of SEQ test results

The obtained data and results from System Usabil-

ity Scale (SUS) is shown in Figure 14.

Figure 14: The Sample Measurement of Three Tourist Des-

tination.

Based on the SUS rating scale in Figure 3, the av-

erage SUS values were 71.08 and can be classiﬁed as

Good.

CONRIST 2019 - International Conferences on Information System and Technology

5 CONCLUSIONS

Overall, testing F1 to F9 tasks using the SEQ method

has a median value of 6 which when adjusted in Table

1 is included in the ”Easy” classiﬁcation according

to respondents. We also using the System Usability

Scale (SUS) method. The average SUS values ob-

tained based on Table 6 were 71.08. Based on the

SUS rating scale in Figure 2, it was found that testing

using the SUS method with a value of 71.08 was in-

cluded in the ”Good” category with a ”C” rating. In

other words, the Smart Assistant Diabetes apps has

good usability according to respondents.

The advice that can be given from the results of

testing and analysis of this research is that it is ex-

pected that in further application development “task

success rate” testing can be done which serves to mea-

sure the level of user success in completing tasks in

the application and can be done “time-ontask” testing

which functions to measure how much time is needed

for users to complete the task.

REFERENCES

AFRIZA, A. (2015). Pengaruh senam diabetes terhadap

kadar glukosa darah pada penderita diabetes mellitus

di puskesmas lapai kecamatan nanggalo kota padang.

Jurnal Ilmu Keolahragaan, 1(1).

Alturki, R., Gay, V., and Alturki, R. (2017). Usability test-

ing of ﬁtness mobile application: methodology and

quantitative results. Proceed. Int. Confer. Computer

Sci. Engineer. Applic, pages 97–114.

Ardiansyah, A. and Ghazali, M. I. (2016). Pengujian usabil-

ity user interface dan user experience aplikasi e-reader

skripsi berbasis hypertext. Jurnal Ilmiah Teknologi In-

fomasi Terapan, 2(3).

Atmodjo, G. E. and Krisjanti, M. N. (2016). Preferensi kon-

sumen terhadap merek smartphone berdasarkan sis-

tem operasi. Fakultas Ekonomi, Universitas Adma-

jaya.

Bangor, A., Kortum, P., and Miller, J. (2009). Determining

what individual sus scores mean: Adding an adjective

rating scale. Journal of usability studies, 4(3):114–

123.

Davis, D. and Jiang, S. (2016). Usability testing of exist-

ing type 2 diabetes mellitus websites. International

journal of medical informatics, 92:62–72.

Federation, I. D. (2019). Idf diabetes atlas - across the

globe.

Kardika, I. B. W., Herawati, S., and Yasa, I. (2015). Pre-

analitik dan interpretasi glukosa darah untuk diagno-

sis diabetes melitus. Bagian Patologi Klinik Fakultas

Kedokteran Universitas Udayana Rumah Sakit Umum

Pusat Sanglah, 1:1689–1699.

Novita, A. R., Kartini, A., and Pradigdo, S. F. (2018).

Pengaruh frekuensi senam diabetes melitus terhadap

kadar gula darah (studi pada kelompok umur ≥

45 tahun di kota semarang). Jurnal Kesehatan

Masyarakat (e-Journal), 6(4):168–175.

Nurahmatya, W. A. and Asnindari, L. N. (2014). Hubun-

gan Senam Diabetes dengan Kadar Kolesterol Darah

pada Penderita Diabetes Mellitus Tipe II di RS

PKU Muhammadiyah Yogyakarta. PhD thesis,

STIKES’Aisyiyah Yogyakarta.

Organisation, W. H. (2017). Diabetes who.

Perwira, R. I. (2012). Sistem untuk konsultasi menu diet

bagi penderita diabetes mellitus berbasis aturan. Jur-

nal Teknologi, 5(2):104–113.

Rehmaita, R., Tahlil, T., et al. (2017). Pengaruh senam dia-

betes dan jalan kaki terhadap penurunan kadar gula

darah pada pasien dm tipe ii di puskesmas krueng

barona jaya aceh besar. Jurnal Ilmu Keperawatan,

5(2):84–89.

Santoso, J. (2018). Usability user interface dan user ex-

perience media pembelajaran kamus kolok bengkala

berbasis android. Jurnal Sistem dan Informatika (JSI),

12(2):174–181.

Sharoh, S. M. and Salmiyati, S. (2017). PENGARUH

SENAM DIABETES TERHADAP KADAR GULA

DARAH PADA PENDERITA DIABETES MELLITUS

TIPE II DI WILAYAH KERJA PUSKESMAS GAMP-

ING 1 SLEMAN YOGYAKARTA. PhD thesis, Univer-

sitas’ Aisyiyah Yogyakarta.

Wetzlinger, W., Auinger, A., and D

orﬂinger, M. (2014).

Comparing effectiveness, efﬁciency, ease of use, us-

ability and user experience when using tablets and lap-

tops. In International Conference of Design, User Ex-

perience, and Usability, pages 402–412. Springer.

Widiastuti, K. C., A. D. and Syahbani, M. H. (2015).

Aplikasi pengambilan keputusan dan pengatur pola

makan bagi penderita diabetes berbasis android.

Widodo, I. D. et al. (2017). Usability testing for android

based application “jogja smart tourism”. In IOP Con-

ference Series: Materials Science and Engineering,

volume 215, page 012031. IOP Publishing.

Yendi and Adwiyana (2014). Pengaruh latihan jasmani

senam diabetes melitus terhadap penurunan kadar

gula darah penderita diabetes melitus di wilayah kerja

puskesmas rasimah ahmad kota bukittinggi tahun

2014.

Usability Testing on Android-based Mobile Application "Smart Assistant Diabetes"