Exploring Uber Taxi Application Using the Technology Acceptance Model

: The goal of this study was to investigate the accountability of the technology acceptance model (TAM) for Uber taxi applications. As antecedents of usefulness, this study selected three elements: time information, price information, and driver information. Data was collected using a survey on Amazon Mechanical Turk. The survey participants were users of the Uber taxi application. The number of valid observations for the data analysis was 412. For data analysis, frequency analysis, conﬁrmatory factor analysis, correlation matrix, and structural equation modeling were utilized. The results show that the usefulness is signiﬁcantly inﬂuenced by time and driver information. They further indicate that ease of use positively affects both usefulness and attitude. In addition, usefulness signiﬁcantly determines attitude, and attitude toward the Uber taxi application exerts a substantial effect on intention to use. This study contributes to the literature by demonstrating the explanatory power of TAM using the Uber taxi application as a research subject


Introduction
The Uber taxi app is a representative business model in the sharing economy.The system is attractive in the market because it allows users to move more conveniently.As such, both tourists and local residents have widely used the transportation system.According to Insider [1], the market share of Uber taxis was approximately 80 percent in the second quarter in 2018.Statista [2] also reported that the market share of Uber taxis in July 2021 was approximately 75 percent in the US.Both statistics indicate that Uber could become a representative ride-hailing company; the follower Lyft showed rapid growth in the market because its market share grew by about 25 percent within 3 years [2].This implies that competition for the taxi mobile application system has intensified.Under market conditions, it is worthwhile to first determine user characteristics to outperform in the market.
The technology acceptance model (TAM) is the main theoretical framework of this research.TAM notes the association between usefulness, ease of use, attitude, and intention to use [3][4][5].Numerous scholars have chosen TAM as a theoretical underpinning to test the utility of various technologies and scrutinize the link between ease of use, usefulness, attitude, and intention to use a certain technology [3][4][5][6][7][8][9][10]. Nevertheless, few studies have been conducted to ensure the explanatory power of TAM by selecting Uber taxi applications as research subjects.Although Min et al. [11] demonstrated the accountability of TAM in the area of the Uber taxi business, a single empirical study might be insufficient to understand user characteristics.Moreover, TAM could be varied depending on antecedents, and it helps to understand Uber taxi service users more deeply.However, it has been insufficiently explored in the area of Uber taxi application research.To fill these voids, this study adopts TAM as a theoretical foundation to inspect the psychological mechanism of Uber taxi application users.A primary function of a taxi application is to inform users about their service because of its intangible features [12,13], which is likely to motivate potential clients to attain more information to minimize uncertainty.In fact, scholars claimed that Uber taxi provides various pieces of information, which include the time, price, and driver [14][15][16][17][18]; the information could become a clue for users to evaluate whether the application system is useful or not.Although Min et al. [11] investigated the explanatory attributes for TAM's main attributes: perceived ease of use and perceived usefulness, the focus for information provision of an application system has not been well explored.To fill such a research gap, this research aims to scrutinize information-related elements to understand Uber taxi application users.Given these aspects, this research posits three attributes as determinants of the usefulness of Uber taxi applications.The contribution of this research is its scrutiny of the user behaviors of Uber taxi applications.Ultimately, this research aims to inform taxi application service providers to upgrade their service level, which might make service providers more competitive and provide customers with better services.

Taxi Application and Information
The information asymmetric condition is a theoretical underpinning of this work.Scholars contends that information asymmetry increases the likelihood of adverse selection [19,20].Therefore, offering information to consumers is an essential element to minimize the risk in consumption [21,22].The main function of the taxi application system is to provide information to potential customers.Service products are intangible; the customers' main concern is ambiguity because they cannot assess the product without experience [12,13,23].To minimize such a concern, information could become a solution because it assists customers in imagining and anticipating the service features [13,24,25].Indeed, prior works argued that the main function of taxi service applications is to offer information: time, price, and service provider [15,18,26].Previous studies have documented the details of information from the Uber application system.First, scholars have found that taxi application systems offer time information, including duration and waiting [14,27].Time information allows customers to save resources because time is a resource [14,28].Second, the extant literature states that price information is the second piece in the taxi application system because price information makes taxi service customers decrease the likelihood of over-payment [16,17,29].Finally, prior research has noted that driver information contributes to the utility of customers using a review system that provides information about drivers [15,30].Specifically, customers can minimize service failure risk using review information, and drivers can become kinder by the evaluation rating system [16,27,30].In sum, the information reduces the risk to prospective service customers and saves their resources (e.g., time and money), which results in the creation of more utility.Based on the literature review, this study proposes the following research hypotheses: H1.Time information positively affects perceived usefulness.

H2. Price information positively affects perceived usefulness.
H3. Driver information positively affects perceived usefulness.

Technology Acceptance Model (TAM)
The theoretical underpinning of this study is the technology acceptance model (TAM), which evaluates a technology's utility using various attributes: perceived usefulness, perceived ease of use, attitude, and intention to use [3,31].The central components of TAM are perceived ease of use and perceived usefulness.Scholars define usefulness as how technology users perceive improved efficiency in their task, and ease of use as how users control a technology without struggling [32][33][34][35].Numerous studies have documented the explanatory power of TAM in diverse domains.For instance, Chen et al. [36] inspected sustainable energy technology users using the TAM; Kamal et al. [37] researched telemedicine service users using TAM; Manis and Choi [38] explored the behavioral characteristics of virtual reality technology users using TAM.Furthermore, Zhong et al. [10] tested TAM's accountability in the context of facial recognition payment technology.Wong et al. [5] investigated personal protective equipment technology by employing TAM.Min et al. [11] demonstrated the accountability of TAM in the domain of Uber mobile applications.
Primarily, TAM rationalizes the association among variables, which are ease of use, usefulness, attitude, and intention to use.Specifically, ease of use is the antecedent of usefulness and attitude because it is positively related to both working efficiency and the impression of technology [4,[7][8][9].Additionally, the extant literature claims that usefulness positively determines both attitude and intention to use because technology users assess the utility of technology depending on enhanced task efficiency, which in turn impacts attitudes and intentions [33,39,40].TAM posits that attitude influences intention to use because positive impressions lead users to choose the system [9,34,35,40].Considering TAM, this research proposes the following research hypotheses: H4.Perceived ease of use positively affects perceived usefulness.

H7.
Perceived ease of use positively affects attitude.

H8.
Attitude positively affects intention to use.

Research Model and Data Collection
Figure 1 illustrates our research model.Seven attributes (time, price, driver, ease of use, usefulness, attitude, and intention to use) appeared in the research model.The determinants of usefulness are time, price, driver, and ease of use.The antecedents of attitudes include usefulness and ease of use, and the intention to use is influenced by usefulness and attitude.All the directions were positive.Amazon Mechanical Turk (https://www.mturk.com(accessed on 17 January 2022)) is the main instrument for data collection.Amazon Mechanical Turk is an Internet crowdsourcing system that allows task creators to collect data from paid task completers.Numerous works have used Amazon Mechanical Turk for data collection, and reported Amazon Mechanical Turk (https://www.mturk.com(accessed on 17 January 2022)) is the main instrument for data collection.Amazon Mechanical Turk is an Internet crowdsourcing system that allows task creators to collect data from paid task completers.Numerous works have used Amazon Mechanical Turk for data collection, and reported substantial statistical inference, which expresses the quality of data from Amazon Mechanical Turk [41][42][43].This study collected data between 20 January 2022 and 21 January 2022.This research selected participants who are experienced with the Uber taxi application by asking, "Have you ever used the Uber taxi app?"This study used the samples that responded "yes" to this question.At first, 450 observations were selected, after which 38 were eliminated because they responded "no" to the screening question.The effect size of our sample appeared as 118.37 by following the instruction of Boomsma [44].It is approximately 119; the sample size of this work is 412.It indicates that the number of observations for SEM is sufficient.

Survey Description
This study employed a five-point Likert scale (1 = strongly disagree to 5 = strongly agree) to measure all constructs other than attitude.Attitude was measured on a five-point scale (1 = negative, 5 = positive).For the derivation of information-related measurements, three tourism and service area researchers were consulted.The measurements of variables within the TAM adopted prior research items with modifications for the purpose of the current research.Perceived usefulness is defined as how the working efficiency of application users is enhanced by adopting the system [32][33][34][35].The specification for perceived ease of use is how users feel they can control the system without difficulty [4,8,11,34].This research defined attitude as users' predisposition to the application system based on their experience [8,33,35] and intention to use as how users select the application for commercial transactions [7,39,40].Table 1 lists the measurement items used in this study.Time information and perceived usefulness were measured with three items; other constructs (e.g., price information, driver information, perceived ease of use, attitude, and intention to use) contained four items.PI3 I could reach price information using Uber taxi app.

PI4
Uber taxi app presented price details.
Driver information DI1 I could subscribe driver information using Uber taxi app.

DI2
Uber taxi app provided driver information.

DI3
Uber taxi app offered details of driver.

DI4
Using Uber taxi app enabled me to acquire driver information.

Perceived usefulness UF1
Using Uber taxi app enabled me to arrive destination more quickly.

UF2
Using Uber taxi app improved my service experience.

UF3
Using Uber taxi app enhanced the effectiveness of transportation service.

Construct Code Item
Perceived ease of use EU1 Uber taxi app was easy to use.
EU2 It was simple to use Uber taxi app.
EU3 Uber taxi app provided easy system to use.
EU4 It was straightforward to use Uber taxi app.

Attitude AT1
Using Uber taxi app is bad-good.

AT2
Using Uber taxi app is negative-positive.

AT3
Using Uber taxi app is unfavorable-favorable.

AT4
Using Uber taxi app is stupid-wise.
Intention to use IU1 I intend to use Uber taxi app.

IU2
I am going to use Uber taxi app.

IU3
Uber taxi app will be selected for me.

IU4
I will use Uber taxi app.

Data Analysis
First, frequency analysis was implemented to analyze demographic information, followed by confirmatory factor analysis, a correlation matrix, and structural equation modeling (SEM) analysis.This study used covariance-based (CB) SEM using the maximum likelihood because the extant literature found that CB-SEM is adequate for theory-based hypotheses testing among variables [45][46][47].Additionally, CB SEM of this work is appropriate for the data analysis because this research had sufficient observations for normality based on the central limit theorem [45,46].To ensure convergent validity, we examined both factor loadings using 0.5, and construct reliability using 0.7 as the criteria, following Hoyle [48] and Hair et al. [45].We then followed the rule that the square root of the average variance extracted should be greater than the correlation coefficient to ensure discriminant validity [45,49].Next, as Hoyle [48] and Hair et al. [45] stated, the goodness of fit of the structural equation model was evaluated using multiple indices: Q (CMIN/degree of freedom) < 3, goodness of fit index (GFI), normed fit index (NFI), relative fit index (RFI), incremental fit index (IFI), Tucker-Lewis index (TLI), comparative fit index (CFI) > 0.8, and root mean square error of approximation (RMSEA) < 0.05.

Convergent Validity and Discriminant Validity
Table 3 illustrates the results of the confirmatory factor analysis.Regarding the goodness of the indices (Q (χ 2 /df) < 3, GFI, NFI, RFI, IFI, TLI, CFI > 0.8, and RMSEA < 0.05), the results were statistically acceptable (Q (χ 2 /df) = 1.800,GFI = 0.909, NFI = 0.923, RFI = 0.910, IFI = 0.964, TLI = 0.958, CFI = 0.964, and RMSEA = 0.044) [45,49].All indices satisfy the criteria to appraise the goodness of fit.In addition, all factor loading, construct reliability, and average value values indicate convergent validity because they meet the criteria proposed in the existing literature.Table 3 presents the mean values of these variables.Next, the square root of the average value values in Table 4 is greater than the correlation coefficients, which suggests that discriminant validity was confirmed.All things considered, this research derived seven constructs: time information, price information, driver information, usefulness, ease of use, attitude, and intention to use.

Discussion and Conclusions
This study evaluated whether TAM is applicable for explaining the behavior of Uber mobile application users.Using usefulness as an endogenous variable, it could be inferred that time information, driver information, and ease of use allowed Uber taxi application users to attain more utility, which could be related to the individual's working efficiency.Regarding the browsing magnitude, Uber taxi application users viewed ease of use as the most important element for their usefulness, and driver information appeared as the second most important attribute.However, price information was identified as a non-significant element to account for its usefulness.It could be assumed that application users might take price information for granted in the taxi service application system.This means that offering price information may be insufficient to create merits from the perspective of users.The results also imply that usefulness and ease of use are significant in building a positive attitude among Uber taxi application users.The magnitude of ease of use is greater than usefulness, indicating that users value simple control of technology in the area of taxi service applications to build positive attitudes.In addition, the results imply that positive attitudes encourage users to build a higher level of intention to use.However, the results showed that usefulness was not a critical attribute for intention to use, even though it influenced positive attitudes.This suggests that the impact of usefulness could be limited to Uber taxi application services.Indeed, the range of mean values for usefulness (3.96-4.11) is lower than that of ease of use (4.24-4.33)and attitude (4.11-4.26).

Discussion and Conclusions
This study evaluated whether TAM is applicable for explaining the behavior of Uber mobile application users.Using usefulness as an endogenous variable, it could be inferred that time information, driver information, and ease of use allowed Uber taxi application users to attain more utility, which could be related to the individual's working efficiency.Regarding the browsing magnitude, Uber taxi application users viewed ease of use as the most important element for their usefulness, and driver information appeared as the second most important attribute.However, price information was identified as a non-significant element to account for its usefulness.It could be assumed that application users might take price information for granted in the taxi service application system.This means that offering price information may be insufficient to create merits from the perspective of users.The results also imply that usefulness and ease of use are significant in building a positive attitude among Uber taxi application users.The magnitude of ease of use is greater than usefulness, indicating that users value simple control of technology in the area of taxi service applications to build positive attitudes.In addition, the results imply that positive attitudes encourage users to build a higher level of intention to use.However, the results showed that usefulness was not a critical attribute for intention to use, even though it influenced positive attitudes.This suggests that the impact of usefulness could be limited to Uber taxi application services.Indeed, the range of mean values for usefulness (3.96-4.11) is lower than that of ease of use (4.24-4.33)and attitude (4.11-4.26).This means that users perceive relatively less usefulness in the Uber taxi application service as compared to other attributes, which might result in a constrained effect on intention to use.Moreover, the profile of the survey participants reported that male participants are more than female participants (41.0%).It might be possible that if there were more female participants, the mean values of the driver information might be higher because females might value the risk related to the profile of drivers more.Moreover, the mean value of ease of use appeared Systems 2022, 10, 103 9 of 11 to be relatively high, which might be explained by the age of the survey participants: approximately 80% were less than 40 years old.

Theoretical Implications
The foremost academic contribution of this research is to search for antecedents of usefulness in the context of Uber taxi applications, a representative sharing economy system in the area of ground transportation.This study anticipated three influential attributes: time, price, and driver information.Two attributes, including time information and driver information, were identified as substantial elements for usefulness.These results shed light on the literature by clarifying noteworthy elements for the usefulness of Uber taxi application users, indicating that information provision is an imperative element of the application system.Additionally, this study contributes to the literature by revealing the accountability of TAM for user behavior in Uber taxi applications.Therefore, this research expanded TAM into the Uber taxi application area.In addition, the outcome of this research could be aligned with Min et al. [11] in that TAM is the imperative theoretical underpinning for exploring the user behavior of Uber taxi application systems.Although previous empirical work has verified the accountability of TAM [11], additional research would be worthwhile to understand the user behavior of Uber taxi applications.Given this point, this research used TAM and demonstrated its explanatory power in the context of the Uber taxi application.This finding supports the results of the extant TAM literature [3][4][5]10,[35][36][37].

Practical Implications
The results of the current study might inform the management of Uber taxi application systems.The results implied that time and driver information were imperative to making customers perceive more utility in the use of Uber taxi applications.Hence, Uber taxi application managers might be able to invest more in developing time information and driver information; that is, managers could not only provide more real-time and precise time information for Uber taxi application users but also develop a more useful framework for driver review information.Next, Uber taxi application managers might need to contemplate how to control the system more easily from the perspective of users, because it would enhance both usefulness and attitude.Based on the magnitude, the priority for resource allocation can be suggested.Specifically, ease of use should become the most important domain, and driver and time information provision could become the second most crucial area for budgeting.In addition, Uber taxi application managers are likely to ponder how technology upgrades a user's working efficiency, which in turn gives rise to a better attitude among users.Finally, Uber taxi application management could dedicate their budget to their reputation management, such as corporate social responsibility, because attitude toward the Uber taxi application and its brand was influential for users to build intention to use, which ultimately results in a greater market share.

Limitations
This study has some limitations.First, it only inspected the antecedents of the usefulness of Uber mobile application systems.To overcome this limitation, future research could examine the influential attributes of ease of use.Moreover, the subjects of this study were limited to American consumers.Future research could choose users in another country considering that business conditions vary depending on the geographical situation.This might enable this study to confirm the generalizability of the results.Moreover, this research was limited to the consequences of TAM.Future research might be able to consider more diverse dependent variables (e.g., continuance intention and intention to reuse) in TAM to further understand consumer behavior.

Figure 2 .
Figure 2. Results of the hypotheses testing.

Figure 2 .
Figure 2. Results of the hypotheses testing.

Table 1 .
Description of measurement.
Note: * p < 0.05, Diagonal is the square root of the average variance extracted, SD stands for the standard deviation.