Improving the Usability User Experience of Peduli Lindungi Application based on Heuristic Evaluation

-Peduli Lindungi is an application developed to help relevant government agencies in tracking to stop the spread of COVID-19. Peduli Lindungi is expected to help the government in the tracking process to facilitate the detection of the reach of COVID19. However, over time, many complaints about the user experience written by users in the App Store and Play Store applications. This study uses the Heuristic Evaluation method to analyze usability. This method is used to get an evaluation of the shortcomings of usability. In this study, it was found that the usability of the Peduli Lindungi Application was good enough. However, there are still aspects that require improvement. This aspect is the Flexibility and Efficiency of Use with a severity rating of 3.8 and Aesthetic and Minimalist Design with a severity rating of 3.4. So that design improvements are made based on these aspects. The test was carried out and found that the Flexibility and Efficiency of Use had a severity rating of 1.5. While Aesthetic and Minimalist Design with a severity rating of 2.6. It is concluded that the design is better than before.


INTRODUCTION
Coronavirus, or severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), is a virus that attacks the respiratory system under the name COVID-19 [1]. The virus causes several disorders ranging from mild respiratory system disorders to severe lung infections and even death. According to data released by the Task Force for the Acceleration of Handling COVID-19, the number of confirmed positive cases as of November 4, 2021, is 4,246,802 people, with a death toll of 143,500. This figure shows that the death rate from COVID-19 is around 3.4% [1].
With the increasing number of positive confirmed cases due to the coronavirus, the government has to take serious steps. One of the government's efforts to help reduce the spread of the coronavirus is to develop the Peduli Lindungi application to track the spread of Coronavirus Disease . This application relies on community participation to share location data while traveling so that contact history tracing with COVID-19 sufferers can be carried out. Users of this application will also get a notification if they are in a crowd or are in a red zone, which is an area where it has been recorded that there are people infected with COVID-19 positive or there are patients under surveillance [2].
Indirectly, the government requires the public to be able to use the application as an effort to stop the spread of Coronavirus Disease . There are rules in some crowded locations that you must scan the barcode to be able to enter the place using the Peduli Lindungi application. Therefore, the Peduli Lindungi application plays an essential role in the daily life of the Indonesian people. However, based on the author's initial observations of user reviews on the App Store and Play store, it was found that many users gave low ratings and negative comments. From the user review, the authors conclude that the Peduli Lindungi application has a display that is not user-friendly, some of the main features required are too many steps, and the lack of information about the features available in the application. Based on the user review, it causes inconvenience & delays for users to use the application. User dissatisfaction will have an impact on accessing caring and protecting mobile applications, which will experience a decrease in users [3]. To see the quality of a product, especially a mobile application, by hearing the opinions of users directly with a user experience approach or user experience and user interface or appearance [4]. The improper interface can cause users to leave the application [5]. This is what happens to Peduli Lindungi apps. Therefore, it is necessary to measure the usability of the Peduli Lindungi application.
Usability is a component that makes up the quality to state whether the user interface is easy to use. Heuristic evaluation refers to expert-based methods to find usability problems in user interface design. The heuristic evaluation aims to find usability problems in existing designs so they can be improved [6]. By using the heuristic evaluation method by Nielsen, it is hoped that researchers will be able to find quality and evaluate usability errors & deficiencies in care and protect applications. From these tests, it can be concluded that the lack of application as an evaluation material for the Peduli Lindungi application is to be used as the basis for the redesign process so that users can get a better experience and pleasurable when using the application. So, it can be used as input for a Peduli Lindungi developer.
Several studies have been done before. First, it was conducted in 2021 by Muhammad Subhan and Aries Dwi Indriyanti, who analyzed the user interface and user experience on the BCA Mobile application using the heuristic evaluation method. From this research, the evaluation of the user interface design and user experience of the BCA Mobile application was obtained using the heuristic evaluation method. Thus, it can be part of the redesign process [4]. The second research was conducted in 2019 by Ika Richa Savitri, who evaluated the user experience of the BRI Internet Banking application against the perspective of the Madurese community using the heuristic evaluation method. From the results of this study, it was concluded that almost all parts of the BRI internet banking application had problems but did not interfere with user comfort and did not need to be repaired [7]. Third, research was conducted in 2018 by Annisa Hardianty Nasution, who evaluated the usability of the HKBP Nommensen University library website using the heuristic method. Based on the calculation of 10 heuristic aspects, the average value for each point is 83%, indicating that the library website's usability has been categorized as good and successful [8]. Furthermore, research was conducted in 2019 by Mumtaz Haya Waralalo, who analyzed the user interface and user experience at AIS UIN Jakarta using heuristic evaluation and Webuse methods with ISO 13407 standards. This study found that AIS UIN Jakarta still has an ordinary appearance, lacking attractiveness to unattractive, with a total interface and usability value of 46.33%. So, these results make recommendations for AIS UIN Jakarta using the same methods and standards. Furthermore, these recommendations were measured, and the results showed that the respondents considered the AIS academic information system to have a good scale, namely the interface and usability values of 96% [9]. The previous research was conducted in 2018 by Yemima Monica Geasela, French Ranting, and Johanes Fernandes Andry with the title user interface analysis of E-Learning-based websites with heuristic evaluation methods. This study shows that, in general, an assessment of the interface design and ease of use of an e-learning-based website found that the website is quite good. However, some things still need to be improved, especially the comfort of use characteristics [10].

Research Stages
In this study, the method that will be used to analyze the usability of user experience is heuristic evaluation. Heuristic evaluation is one of the evaluation methods to find usability problems of computer software by referring to 10 heuristic aspects to obtain improvements [11]. While the method used to do the redesign is User-Centered Design (UCD). UCD is a method that focuses on potential users in creating a product [12]. The flow of the research method can be seen in Figure 1.

Problem Identification
At this stage, the author determines the studied object and looks for references related to the problem. Furthermore, the authors look for problems that often appear on these objects through user reviews on the app store and play store applications. The author chose to use the heuristic evaluation method. However, the author has previously determined the procedure to be used. Among the methods of Think Aloud Evaluation (TA), Cognitive Walkthrough (CW), and Heuristic Evaluation (HE). Heuristic Evaluation is a method that can produce the most usability problem findings compared to other methods [13].

Data analysis based on heuristic evaluation
Data collection is done by distributing questionnaires to evaluators. By Nielsen's principles, several things that will be assessed are aesthetics, control, and efficiency of product use. The appraiser consists of 5 evaluators because this number is proven to find about 75% of all product usability problems. Each evaluator must deeply understand digital product development and come from a related background, for example, UI/UX design [14].

No
List of Questions 1. The application displays information about a specific condition. 2. The application displays feedback after the command is completed.

14.
The application displays a choice of solutions for the error that occurs (for example, when you forget the password, there is an option to update the password). 15. The "Help" menu is easy to find.
After getting the results on a likert scale, the writer tested the validity and reliability. A validity test is used to determine whether the data generated from a questionnaire is feasible or not [15]. Measurement of item validity by correlating the item score with the total item score [16]. Furthermore, a correlation coefficient (r) will be obtained to measure an item's validity level and determine whether it is valid or invalid. The following are the requirements for decision-making in determining the validity [17]: -If r value > r table, it can be stated Valid question item -If r value < r table, it can be stated that the question item is Invalid Reliability testing is carried out to determine the consistency of an instrument for testing that is carried out more than once to determine the extent to which the instrument can be trusted [18]. Decision-making to determine reliability [19].
-If the Cronbach Alpha value > 0.600 is declared Reliable -If the Cronbach Alpha value < 0.600 is declared Unreliable Suppose the questions for the questionnaire are valid and reliable. Then the severity rating is calculated from each aspect to determine the assessment for each element of the Peduli Lindungi application. The evaluation is carried out using a scale of 1-5, with a value of 1 being the lowest usability problem and a value of 5 being the most significant usability problem. The severity rating value in one usability aspect is obtained from the average severity rating score for that aspect [20].

Recommendations for improvement based on User Centered Design
In the redesign process, the UCD method is used. By going through several stages such as understanding the context of the user, specifying user requirements, designing a solution, evaluating against requirements.

Trial of Design Recommendations
After the recommendation is successfully made, the evaluators will test the design. The trial will be tested by making a prototype feature simulation using Figma.

Data collection
This study uses a Likert scale of 1 to 5, with a value of 1 as the lowest usability problem and five as the most significant usability problem. The following is a description of the Likert scale used [4]. No problems or flaws were found. 2 The problems found do not affect the application's users much. 3 There are potential problems that can make it difficult or annoying for users. 4 There is a problem that has the potential to make it difficult for application users. 5 There is a fatal problem that needs to be fixed on the application.

Validity test
Validity testing is done by comparing the r value and r table. The r value can be obtained by correlating the item score with the total item score [16]. The authors rely on the SPSS software application with a bivariate analysis menu to facilitate the calculation. Meanwhile, the r table is obtained by looking at the r table product. For five evaluators, so that DF = 3 with a significance of 0.05 was obtained, the r table value was 0.878 a and the results of the validity test can be seen in the Table 4.

Table 4. Validity Test Results
The Question r Value r Based on the calculation results, all question items have an r value that is greater than the r table, and it can be concluded that all question items in the questionnaire variable are declared valid.

Reliability Test
In this reliability test, Cronbach's Alpha method is used. Cronbach's Alpha value is obtained using SPSS software on the bivariate analysis menu. The results of the Reliability Test processed using SPSS according to the Table 5.

Cronbach's Alpha
N of the question 0,983 15 Based on the data processing results above, Cronbach's Alpha questionnaire value is 0.983. So it can be concluded that the questionnaire in the variable is declared reliable.

Design Improvement Analysis and Recommendations
In conducting the analysis, the author uses a questionnaire involving five people as evaluators with experience in the field of user experience and users of the Peduli Lindungi application. From the results of the questionnaire that has been obtained from the evaluators, then the severity rating is calculated for each aspect. Based on the calculation of the severity ratings above, the evaluators can tolerate several aspect points. The higher the severity rating value means, the more needed for improvement. The highest severity ratings were found in the aspect of Flexibility and Efficiency of Use, with a value of 3.8. The following are the results of the complete user experience analysis questionnaire data processing for the Peduli Lindungi application.
In the questionnaire, in addition to the statement above, the author will also provide a section for respondents to use to fill in subjective assessments and details regarding criticism when using the Peduli Lindungi application. From several questions regarding each aspect of heuristic evaluation, the following are some aspects with the highest severity rating value: a. Aspects of Flexibility and Efficiency of Use Flexibility and Efficiency of Use have the highest severity rating value of 3.8. In this aspect, evaluators think that several essential features lack efficiency. b. Aspects of Aesthetic and Minimalist Design Aesthetic and Minimalist Design has a severity rating of 3.4. In this aspect, evaluators think that the design is not simple, so the care in the Peduli Lindungi application design should look more aesthetic.
In addition, the author provides a field for filling in the questionnaire so that evaluators can offer suggestions to recommendations for improving the Peduli Lindungi application design. The following are some points of advice given by respondents: a. The COVID-19 certificate & test result in the display feature is placed in an easily accessible position. b. The use of more minimalist icons so that they are suitable for all groups. c. Consistency of language used on all pages.
Based on the analysis by the evaluator on the evaluation of user experience, several problems were concluded. With the regulations that require the use of the care in Peduli Lindungi application, evaluators need to rely on the care in Peduli Lindungi application features to be used efficiently and have an aesthetic appearance. Therefore, the author recommends designing the Peduli Lindungi application display according to the UCD method.

Figure 2. Splash Screen
The Peduli Lindungi application splash screen is the initial display of the application when the user opens the application.  The page shown in figure 7 has a button to display the documentation on "Using Scan QR Code." When going to check in, users can press the "Scan QR Code" button and point the camera at a QR Code at a location. In addition, users can choose to check in individually or in groups.

Figure 8. Check in Success Page
After the user has successfully checked in, the following screen, like figure 8, will appear that provides information if the user has successfully checked in. In addition, it also includes information about the user name, check-in location, time, and total crowd.

Improvement Testing
After getting the prototype for improvement recommendations, a simulation is carried out using the Figma software. The trial was carried out on evaluators who had previously provided an assessment of the data collection process. The author gives the evaluators a task to access the entire prototype design. Then, the evaluators were