Analysis of User Acceptance Using UTAUT2 Model in KAI Access Application

The convenience provided by mobile applications provided substantial opportunities for companies, especially those that engaged in the service sector to improved performance and service to customers and get the best feedback as the value of the company's competitive advantage. This research’s aims were to find out what factors that influence user acceptance (behavioral intention and use behavior) in KAI Access application using the UTAUT2 model for users of the KAI Access application. Fsctors/variables used in this research were effort expectancy, performance expectancy, social influence, facilitating conditions, price value, habit, hedonic motivation, behavioral intention, and use behavior. The number of respondents who could be used for data processing and analysis in this research were 257 respondents, with details of 27 offline respondents and 230 online respondents. The data was obtained from the distribution of google forms through various social media and the distribution of questionnaires directly at the Kertapati Palembang train station. The research method used quantitative methods and analytical techniques using validity and reliability tests with Rasch Modeling, Path Analysis, Classical Assumption Test, Anova Test and Partial Test. The results of the the research concluded that performance expectacy, facilitating condition, hedonic motivation, price value, and habit affect the behavioral intention and use behavior of the KAI Access application. The most influential/significant variables are hedonic motivation, facilitating conditions, price value, and habit.


INTRODUCTION
The growth of IT application increases the use intensity of mobile device such as smart phones, tablets, smart watches, etc. with the support of 3G or 4G internet connection Jurnal Teknologi Informasi dan Pendidikan Volume 15 networks. The total population of Indonesia is ±271 millions people [1] which is one of the countries with the largest1mobile device user in the1world [2]. According to data from Emarketer (2014), approximately in 2018 total of active mobile device user in Indonesia will operate more than 100 million user. Indonesia has been placed fourth for largest mobile device user in the world after India, the United States, and China [3]. One of the information technology products on mobile devices that are popularly used by the public is a mobile application. Mobile phones offer increasingly widespread content and services and provide various conveniences in accessing information [4]. The convenience provided by the mobile application provides a great opportunity for companies, especially those engaged in the service sector to improve performance and service to customers and get the best feedback as the value of the company's competitive advantage. One example is the improvement of services in the transportation sector, especially rail transportation in the current digital era carried out by PT KAI.
To improve services to the public in the current digital era, PT KAI is collaborating with creating a mobile-based e-ticket application. E-Ticketing is a paperless electronic system used for ticket reservations, usually used by ticket travelers [5]. E-ticketing made by PT KAI is called KAI Access.
KAI Access is the public application from PT Kereta Api Indonesia (Persero) that was released in 2014 and could be downloaded by anyone in app store and google play store. The KAI Access application provides features for booking train tickets, rescheduling tickets, cancelling tickets, e-boarding pass, etc. Socialization regarding the KAI Access application is intensively carried out. PT KAI recommends passengers to use the KAI Access application as a form of ease of service offered by PT KAI, especially during the current pandemic.
However, in its implementation, the use of the KAI Access application is still low. This is evidenced by the number of application downloads on the Play Store which still amount to 5 million downloads. Even though the number of train passengers during the January-November 2020 period cumulatively reached 172.6 million passengers [6]. The use of the KAI Access application which is still low indicates that the behavioral intention (BI) and use behavior (UB) by users to use the KAI Access application is still low.
In fact, behavioral intention (BI) and use behavior (UB) by users determine the success of application system adoption because behavioral intention (BI) and use behavior (UB) were form from user attitudes and perceptions of an application service. On the other side, PT KAI has never made any observations or knows about any constructs that affect the behavioral intention (BI) and use behavior (UB) of a user of the KAI Access application. Therefore, it is necessary to conduct research on the factors/constructs that affect the behavioral intention (BI) and use behavior (UB) of the KAI Access application.
There are many research models that already developed to determine factors that affect behavioral intention & use behavior of an information technology [7]. One model that is widely used in various kinds of research today is the Unified Theory of Acceptance and Use of Technology 2 (UTAUT2) [8].
The UTAUT2 model was developed by Venkatesh et al in 2012 [9]. The UTAUT1 model has four main construct, namely performance expectancy, effort expectancy, social influence, and facilitating conditions on behavioral intentions to use technology. Then, UTAUT2 add three more construct from the previous UTAUT model, that are hedonic motivation, price value, and habit [10].
The constructs/factors in the UTAUT2 model are appropriate to use because the constructs in UTAUT2 were capable to describe the influence on behavioral intention (BI) and use behavior (UB) of applications in a context of consumer use [11], UTAUT2 is more eminent than UTAUT1 construct whose context is more directed to applications in an organizational context [12]. The KAI Access application itself is a product of technological innovation in the context of consumer use, so it is very appropriate to do the reseach with the UTAUT2 model.
The UTAUT2 model was use to measure extent of behavioral intention (BI) and application use behavior (UB) on several factors that influence users to use the application and analyze the factors that influence it based on the constructs that exist in UTAUT2.
This research's purpose is to know the factors/consturcts in UTAUT2 model that affect/influencing behavioral intention (BI) and use behavior (UB) in KAI Access application. So that the research does not deviate from the discussion, the researcher only focuses on the official ticket booking application from PT KAI, namely KAI Access. Respondents in this research were user of the KAI Access Application. Respondents in this research were purely application users, not employees of KAI Access application operators.

RESEARCH METHOD
This reseach use UTAUT2 model as theoretical framework. According to space the scope of research, UTAUT2 model used in this research does not use a moderator.
The researcher uses independent variables, namely effort expectancy (EE), performance expectancy (PE), social influence (SI), facilitating conditions (FC), hedonic motivation (HM), price value (PV), and habit (HB). While the dependent variable is behavioral intention and use behavior. Furthermore, path analysis is carried out as shown in the following figure and equation:  Model 1, Z = p1X1 + p2X2 + p3X3 + p4X4 + + + p5X5 + p7X6 + p8X7 + e1 Model 2, Y = p1X1 + p2X2 + p3X3 + p4X4 + p5X5 + p7X6 + p8X7 + p10Z + e2 Model 3, Y = p6X4 + p9X7 + e3 After obtaining the results of the path equation, the ANOVA test and partial test were use to determine how independent variable affect depentdent variable. The research hypotesis can be assumed below: H1: Performance expectancy (PE) affect behavioral intention (BI) to use the KAI Access application. H2: effort expectancy (EE) affect behavioral intention (BI) to use the KAI Access application. H3: social influence (SI) affect behavioral intention (BI) to use the KAI Access application. H4: facilitating conditions (FC) affect behavioral intention (BI) to use the KAI Access application. H5: hedonic motivation (HM) affect behavioral intention (BI) to use the KAI Access application. H6: facilitating conditions (FC) affect use behavior (UB) for using the KAI Access application. H7: price value (PV) affect behavioral intention (BI) to use KAI Access application. H8: habit (HB) affects behavioral intention (BI) to use the KAI Access application. H9: habit (HB) affects use behavior (UB) to use the KAI Access application.

Population & Sample
The population in this reseach are every users of the KAI Access application. The population in this research is classified as an infinite population, because the size of the population is not known with certainty or is infinite.
While the sample in this research are users of the KAI Access application which were taken according to research needs. In this research, sampling techniques that are used by researchers is non-probability sampling. Non-probability sampling is one of the sampling technique that did not allow equal chance or opportunities for each member of population to be sampled [13][14] [15]. As for the approach, the researcher uses a purposive sampling approach. Purposive sampling is an approach where sampling is done with a certain consideration.
The researcher uses the Wibisino equation to determine the sample size of the unknown population, namely:

RESULT AND DISCUSSION
Data were collected using online and offline questionnaires. A total of 257 respondents were obtained, with details of 230 online respondents and 27 offline respondents. The questionnaires were distributed from April 2021 to May 2021. The collected questionnaires were then distributed by frequency.
Based on the results of the frequency distribution, the majority of respondents in this research were women, namely 136 users (52.9%) and men as many as 121 users (47.1%). Then the majority of respondents in this research were between 17-29 years, as many as 204 users (79.4%) and at least 2 users (0.8%). The majority of respondents in this research had a high school education of 158 people (61.5%) and at least 1 person with the last education of D4 (0.4%). Then most of the respondents in this research have used the KAI Access application for 1-3 years as many as 129 people (50.2%) and a small proportion have used the KAI Access application for more than 5 years as many as 5 people (1.9%). Most of the respondents in this research used the KAI Access application 1-2 times a year as many as 109 people (42.4%) and a small portion used the KAI Access application 8-10 times a year as many as 11 people (4.3%).

Validity and Reliability Test
Validity is sorted by measure in descending order. Question items with answers that tend to vary is Q9 (question number 9), with a measure value of 1.44. While the questions with answers that tend to be uniform are Q1 (question number 1), with a measure value of -0.96.  Table 2 presents two table results, namely individual abilities and question abilities. The value is 0.83 for Cronbach's Alpha value of individual ability that describes if respondent's ability to fill out the questionnaire is good. While the question ability value is 0.94 which states that the questions that have been prepared are categorized as special.
It can be concluded that the respondents represent the entire population and the questions are reliable to measure the research variables. Gleaned from table 7, it is discernible a value of sig. of 0.000 < 0.05 for variables X5, X6 and X7. Then obtained the value of sig. of 0.021 < 0.05 for the X4 variable. Then obtained the value of sig. of 0.012 < 0.05 for the X1 variable. This means that the variables X1 (performance expectancy) and X4 (facilitating conditions) affect the variable Z (behavioral intention). Variables X5 (hedonic motivation), X6 (price value), and X7 (habit) have a significant effect on variable Z (behavioral intention). Gleaned from table 8, it is discernible a value of sig. of 0.000 < 0.05 for variables X4 and X7. This means that X4 (facilitating conditions) and X7 (habit) variables have a significant effect on Y (use behavior).