2.1 Mobile Payment Service
Mobile technology has become an essential part of everyday life [14, 18, 19]. The main thing a customer needs for an MP is a cell phone that can connect to the Internet [20]. MPs use cell phones as an essential part of the process [7, 21]. The most significant difference between MPs and other payment methods. MP is the most important term that needs to be explained. There is no one definition of MP. A different way to electronically handle payments (Schierz, Schilke, & Wirtz, 2010) or "payments over a cell phone" [14] is the most commonly used definition for an MP. This cell phone is the primary identifying standard for MPs compared to other payment methods. Some studies [21, 22] look at all mobile communication devices. Other studies, however, only look at cell phones. MPs are considered the next step in evolving electronic payment transactions. They could use them to pay for trains, plane tickets, hotel rooms, and meals [23]. Dahlberg [2] include wireless and ten other communication ways as MPs definitions. An MP is "any payment in which a cell phone is used as part of a request to start, carry out, or possibly confirm a payment" [24, 25]. MPs are a type of payment transaction that takes place electronically. The buyer uses portable communication devices like cell phones to start, approve, or confirm a payment [26]. The second type is when a customer pays for services and goods bought over the Internet using a cell phone. The last kind is when a customer pays for something at the POS with a cell phone.
According to Boston's Federal Reserve Bank, USA, two prominent types of MPs are: far away and close by [18, 19]. The first two types of MPs listed above are remote MPs, while the third type is the example of a proximity MP. In this paper, we will focus on the best way to use "proximity MPs using NFC-enabled cell phones and the contactless monetary payment infrastructure" for obtaining purposes [11, 27]. Remote MPs are very useful for payments to dealers and person-to-person payments in places that don't have a standard POS system. Remote MPs also include paying for items bought online with a cell phone. Remote MPs might be completed by utilizing the current budgetary payment infrastructure [27, 28].
Most proximity MPs use RFID (Radio Frequency Identification) or NFC (Near Field Communication) technology, which is essential for POS and vending machines. Because of contactless payments, the customer never has to let go of the payment device [29]. This payment type is beneficial because it takes almost no time to finish and eliminates the need to use a physical card [30]. The phone's built-in NFC technology is the vendor's contactless payment-enabled POS system, just like the contactless devices and payment gadgets used today [27]. A standard barcode payment could be the MP option available at places like Starbucks and McDonald's. The cell phone has an RFID chip that works without touching it [24]. RFID technology usually has a stationary point of sale (POS). Also, RFID has a more extended transmission range than NFC. NFC is compatible with "many introduced contactless payment readers," which are used in many POS terminals [31, 32].
The method of Proximity payment has a few parts: (1) Payment Gateway, (2) Portable Device, (3) Contactless Reader, (4) E-Wallet App, and (5) Wireless Network [31]. Proximity payments could be beneficial if using credit cards is complicated or dangerous or if you only have a short time to purchase [30].
Prior researches in the field of customer adoption of MPs have concentrated on individual mobility [1, 14] compatibility [14, 16, 34–36]; convenience [2, 11, 23, 26]; subjective/social norms [2, 14]; perceived risk [7, 35, 37] ; trust [7, 34] perceived benefits/relative advantage [7, 35] Security [2, 11, 34] perceived ease of use and usefulness [2, 23] and cost [11, 16]. Security, technology anxiety, transaction speed, expressiveness, context, and observability are among the least researched aspects [1, 34]. Regardless of corporate projections about the massive capability of MPs, it is essential to comprehend what prevents clients from adopting this innovation for regular use in the purchasing process [33].
2.3 Technology Acceptance Model (TAM)
There are a variety of ways to study innovation utilization behavior. We employ Davis' model for consumer affirmation of diverse information architectures as a basis for this investigation, which is widely accepted (Fig. 1) [38]. "Technology Acceptance Model" (TAM) decodes customer perspectives on different mechanical progressions and will assist in investigating MP allocation from clients' points of view [17, 38, 39]. According to this concept, a customer's decision on whether or not to acknowledge a given improvement can be determined and quantified. Perceived utility and perceived ease of use are two of the model's primary factors of new invention recognition. The TAM is widely accepted as a reliable and effective tool for predicting customer behavior [40–43].
2.4 Model Development
Kim, Mirusmonov [23] say that, although more and more people are using technology to buy things, few studies have been done to look at how people accept technology [40, 44–46]. But, as far as we know, the previously mentioned parts of the extended TAM (social influence, system usefulness, facilitating conditions, hedonic motivation, trust, risk, and attitude toward intention to use MPs) have not yet been looked at as a whole [43]. So, in this study, we want to find out how people see the system of usefulness, ease of use, trust, risk, hedonic motivation, and attitude toward their plans to use MPs for additional purchases. Using Venkatesh, Thong [47] research, we build on our ideas about the main TAM by adding an extra social construct that is a mix of the Theory of Reasoned Action (TRA) and the Theory of Planned Behavior (TPB). In addition to technology-oriented factors, personal use qualities must also be considered. This study's proposed conceptual model is based on a detailed review of pertinent writing about how MPs use. This research expands the TAM to the level of behavioral intention and adds the compatibility construct as follows:
2.4.1. Social Influence
Since social influence is the driving force that emerges after the initial transmission of something new, it is required to disseminate new products [48]. It is an example of social influence when other people impact an individual's thoughts, feelings, or behaviors. Viswanath [49] reinserted and approved social influences as a vital indicator of expectations in the first UTAUT (Unified Theory of Acceptance and Use of Technology), demonstrated and revalidated it in the updated UTAUT2 shows reinserted and approved social influences as a significant indicator of expectations in the first UTAUT Unified Theory of Accept [50, 51]. First and foremost, such assessments are a component of a broader arrangement of observations that reflect the extensive trip involvement. It is possible that particular aspects of the accessible travel experience distinctions, such as acknowledgements of the referents' NFC-MP perspectives, turned out to be of utmost significance to clients. Based on the discussion above, the following hypotheses were developed:
H1
There is a relationship between consumers’ social influences regarding MP
2.4.2. System Usefulness
The perceived ease of use afforded by innovation is directly proportional to a customer's motivation for purchasing and utilizing a product bearing that innovation [38]. The phrase "the degree to which a person believes that employing a certain system will boost his or her job performance" encapsulates the concept of "perceived system usefulness" [16, 36]. There is empirical evidence in the literature on mobile technology that supports a similar conclusion. Perceived system usefulness has a significant influence on customers' propensity to utilize MP technologies [23, 24]. The utility of a system will illustrate that the application of a specific technology may be advantageous for accomplishing a particular goal by a specific individual [52]. According to Moslehpour, Thanh [53], it is "the degree to which the client believes that the online purchase will allow them access to valuable data, make offer examination less difficult, and speed up the purchasing process." We propose the following hypotheses:
H2
There is a relationship between system usefulness regarding MP
2.4.3. Facilitating Conditions
A person's opinion that a specific framework is simple or plain is called "facilitating conditions" [38]. As a result of the TAM model's trustworthiness, Karnouskos [24] have developed a behavioral model that joins the ranks of their predecessors. Faith and usefulness are linked in this concept, where the enabling conditions influence belief.
Numerous experts have shown facilitating conditions to significantly impact whether consumers want to utilize new technology [38, 54]. Various researchers [2] emphasize the significance of creating favorable conditions for MP acceptability [26]. In the case of fundamental regular administration exchanges, MP channels are well-suited because they are self-service orientated. MPs, according to Dahlberg, Guo [55], cannot exist without the presence of enabling conditions. These observations lead to the following hypotheses:
H3
There is a relationship between facilitating conditions regarding MP
2.4.4. Hedonic Motivation
The degree to which customers believe using an IS (information systems) framework is entertaining is known as hedonic motivation [47, 56]. The focus of reception at first, when most consumer IS were supposed to be mostly errand situated, was on internal convictions and functional characteristics [57]. After learning that customers will use IS to complete tasks and interact, IS architects revised their plan's justification. According to Slade, Williams [58] the development of IS was marked by vivacity, excitement, esteem, and contentment, ineluctably signifying non-utilitarian capacities and energizing researchers' enthusiasm.
These elements have been shown to influence shopper IS choices and be crucial in managing customers' behavioral results [58, 59]. Therefore, the current MP relies on the outline that fits the hedonic character of the initial steps of consumers' buying successions while tending to one of the final stages of the utilization procedure—payment. We anticipate that the impacts of social influence, system usefulness, and conducive conditions on trust will be moderated by hedonic incentives. Thus, the following hypothesis was developed.
H4
There is a relationship between hedonic motivation and MP
2.4.5. Trust
Another expansion of the TAM is trust, which is observed to be the center advancement that considers driving consumer acceptance [60]. Trust, with regards to MPs, is characterized as how much mobile payments are good with the qualities, encounters, and behavioral examples that purchasers have[14, 61]. For instance, if they now utilize MPs for additional purchases, they will probably also use them in restaurants, hotels, etc. Versatile administrations' similarity with purchaser needs positively affects the aim to utilize these services [16, 36]. Trust, together with perceived facilitating conditions and system usefulness in a roundabout way, influences a purchaser's aim to utilize MPs [23]. Individuals' ways of life will incredibly influence their choice to use MP services [35]. Schierz, Schilke [14] have found that perceived trust is a helpful augmentation of the TAM. Along these lines, it could expand the proactive power in the basic leadership procedure of utilizing technology. In this study, we characterize trust as the shoppers' conviction that mobile payment exchanges will be handled as per their desires. Mirroring the expanding significance of trust in mobile commerce, in this study, we propose trust as an antecedent variable to the attitude and intention to utilize a mobile payment. Customers' attitude toward its use reflects the swelling significance of trust in m-commerce [62]. Therefore, the intention to use it will significantly be improved by greater trust in the payment system [35]. Therefore, we propose the following:
H5
A relationship is relationship between trust and MP
2.4.6. Risk.
Risk examination is based on two aspects: vulnerability (consumers' lack of knowledge regarding possible outcomes when they make a purchase) and the unavoidable adverse effects of the purchase. Later, this same scientist stated that every shopping behavior is risky because it cannot predict the outcomes with certainty [53, 58].
It as the outcome of a choice that "mirrors the variety of its inevitable outcomes," Al-Jabri and Sohail [63] define it as the likelihood that a development's use cannot be protected. Kim, Mirusmonov [23] define it as "a buyer's recognition of the instability and the unfriendly results of an exchange performed by a merchant". Additionally, risk will negatively impact the client's confidence in the flexible installment framework [64], making using the new installment framework less of a goal [65].
Organizations are in danger from security concerns since they join many small vendors [16, 55]. Security has been the most crucial problem given the current level of prosperity for electronic exchanges and corporate information interchange [66]. One of the major obstacles preventing an MP promotion from moving forward is the security [23, 67]. According to this theory, subjective security refers to how much a person "trusts that employing a specific mobile payment approach will be secure" [53]. Therefore, we hypothesized the following:
H6
There is a negative relationship between risk and MP
2.4.7. Attitude
Fishbein [68] argues that attitudes are formed over time as individuals have more and more experience with a specific activity. To build a particular behavior, the distinct hypothetical models (TAM, TRA, and TPB) have shown that attitude is a fundamental precondition [69]. According to Fishbein [68] attitude is a multi-faceted construct comprised of a psychological, an enthusiastic, and a conative or behavioral evaluation. The psychological segment refers to a person's knowledge about a product or service (experience, beliefs, and opinions). In contrast, the active component refers to a person's proclivities toward a specific question (feelings, emotions, and evaluations) [67]. We conceptualized the following:
H7
There is a positive relationship between attitude and intention to use MPs.
The conceptual research model implied by our six hypotheses and empirically tested in the following section appears in Fig. 2.
1. Research Methodology
The research model was tested with an online questionnaire survey. Quantitative research methods were used, which can't be backed up by outside sources and could lead to one-sided answers. Convenience sampling technique has used to select the study samples. SPSS 26 and IBM AMOS was used to analyze the collected data.
3.1 Questionnaire Design
An online survey was used to compile a report on 31 requests for information. The research utilized qualifying questions to narrow the field of potential participants. If they had used a cell phone in the last five years, they would participate in the study. For this questionnaire, those clients who now use cell phones were excluded. Results were then presented to each respondent utilizing a rating scale. This review identifies social influence, system utility, facilitation conditions, hedonic incentive, trust, and risk aversion as the most critical factors in accepting proximity payments as a payment method of choice. The respondents used a Five-point Likert scale to indicate their degree of comprehension which was ranging from 1 (strongly disagree) to 7 (strongly agree). The questionnaire has two sections. The questionnaire begins with asking about the respondent’s basic demographic and socioeconomic backgrounds (gender, age, occupation, job title, years of experience, level of education, years of using cell phones, income level), then the next part was designed for measuring the related variables considered for the that study.
3.2 Measurements Scales
The questionnaire items were adopted and adapted from the scales that were used in the previous studies on similar researches (Table 1).
Table 1
Measurement Scales using in the research model
Constructs | Items | Sources |
Social Influence (Socinf) | People who are imperative to me, find the use of mobile payment services valuable. People who inspiration my behavior think that I am capable of using mobile payment services. People who are significant to me think I should use mobile payment services. | Venkatesh and Davis [70] |
System Usefulness (Sysuse) | Use of mobile payment enable me to make payment easily. Use of mobile payment enable me to conduct transactions. A consumer experience is increased with improved flexibility by using mobile payment services. I have found the system of mobile payment useful. | Srivastava, Chandra [64] |
Facilitating Conditions (Faccon) | Learning to use a mobile payment would be easy for me. It would be easy to get a mobile payment system to do what I want it to do. My interaction with a mobile payment system would be clear and understandable. Overall, I would find the mobile payment system to be easy to use. | Van der Heijden [71] |
Hedonic Motivation (Hedmot) | Using mobile payments for purchasing purposes is fun. Using mobile payments for purchasing purposes is enjoyable. Using mobile payments for purchasing purposes is entertaining. Using mobile payments for purchasing purposes is pleasant. | Chang, Eckman [72] |
Trust (Trust) | I trust mobile payment systems to be reliable. I find mobile payment services secure for conducting my payment transactions. I believe mobile payment systems are trustworthy. Even if the mobile payment systems are not monitored, I’d trust them to do the job correctly. I trust mobile payment systems to be secure. | Srivastava, Chandra [64] and Theng, Tan [73] |
Risk (Risk) | Information about my mobile payment transactions would be known to others. I believe mobile payment transactions may be modified or deleted by others. I would label adopting mobile payment systems as a potential loss. I believe that overall riskiness of mobile payment systems is high. | Srivastava, Chandra [64] |
Attitude (Att) | I have a positive attitude towards this portal. I intend to visit the portal frequently. Using mobile payment services is beneficial. Using mobile payment services is a good idea. | Kasavana [31] |
Intention to use (IU) | I intend to use MPs frequently to buy products and services. I intend to use MPs in the future I intend to use MPs as much as possible | Chen and Barnes [77]; Islam, S., Islam, M. F., & Zannat, N.-E [78] |
3.3 Data Collection
An organized, electronic questionnaire was circulated throughout one month via online using Google form in Uzbekistan between February-April, 2023. The number of participants in the study consisted of buyers in Uzbekistan more than 20 years old and who had utilized a cell phone inside the most recent five years. Toward the finish of the information accumulation period, 276 review results were considered as legitimate for factual examination. Keeping in mind the end goal to guarantee exactness and increment outer legitimacy of the review comes about, 24 of 300 gathered surveys were precluded because of missing information, invalid reactions, or inadequate reactions. In this manner, 276 polls were at last used for experimental investigation. The ratio of male and female respondents of the study is (male: 62.2% and female: 37.8%). A large number of respondents fall between the age of 20 to 30 years (63.5%) while 30.7% from 31 to 40 years old.
2. Analysis and Results
In order to determine the effects of eight factors (system usefulness, social influence, facilitating conditions, trust, risk, hedonic motivation, and attitude) the linear regression was conducted to examine the usage of MPs. There was a total of 276 responses utilized for the analysis. A prior study Venkatesh, Thong [47] shows that age and gender are the most significant demographic characteristics that explain MP adoption. Nevertheless, recent research [4] indicates that young and middle-aged MP users will be the most active MP adopters in the coming years.
4.1 Reliability and Validity
4.1.1 Reliability analysis of measurement tools
Internal consistency was measured by utilizing Cronbach's α coefficient as a technique to build the reliability of estimation devices by finding and dispensing with things that bother reliability when utilizing different things to quantify a similar idea individually, Cronbach alpha indicator (Table 2) is applied to examine the reliability of the constructs, since the Cronbach’s Alpha score of .70 was believed suitable by [74]. All the variables show relatively high reliability estimates, ranging from .788 (trust) to .917 (hedonic motivation).
Table 2
Reliability of each scale
Variables | Items | Cronbach's ⍺ |
Social influence | 3 | .863 |
System usefulness | 4 | .892 |
Facilitating conditions | 4 | .892 |
Trust | 5 | .788 |
Risk | 4 | .907 |
Hedonic motivation | 4 | .917 |
Attitude | 4 | .873 |
Intention to use | 3 | .817 |
4.1.2 Confirmatory factor analysis
To test the convergent and divergent validity of the scales, a confirmatory factor analysis was performed. In order to evaluate construct validity, it is necessary to examine intensive validity, discriminant validity, and legal validity.
Discrimination validity is also made by examining the relationship between potential variables as well as the law validity. Discrimination validity can be evaluated in the following way. (1) Examine whether the average variance extraction value (AVE) is greater than the square of the correlation coefficient between concepts. In other words, if the correlation coefficient is squared, it is judged that there is a validity of discrimination. (2) It is a method of judging whether or not to reject the hypothesis that the concepts are the same. In other words, if the correlation coefficient ± 2 × standard error is not 1 in the 95% confidence interval, the discriminant validity is considered. (3) After selecting a pair of theoretically similar concepts, we set a constraint model with fixed correlation coefficient between two concepts and a free model with free correlation between the two concepts, (P = .05 to 3.84 or more), the discriminant validity between the two concepts seems to be reasonable.
In this study, before analyzing the hypothetical relationship between constitutional concepts, a measurement model was set up to analyze the determinants of social influence, system usefulness, facilitating conditions, hedonic motivation, trust, risk and attitude are the most frequently used indexes for the evaluation of conformity with the model. The goodness-of-fit index (GFI), the adjusted goodness-of-fit index (Normed Fit Index), IFI, Root Mean Square Residual (RMR), and Root Mean Square Error of Approximation (RMSEA).
Table 3
Factor analysis of confirmatory factors and SE fitness
Interpretation criteria | Index | Interpretation criteria | Fit Index |
| χ2: p < .05 | χ2 = 1036.915 *** | χ2: p < .05 | χ2 = 3692.059** |
GFI | 0.9 | .774 | .9 More than | .907 |
AGFI | 0.9 | .725 | .8 More than | .888 |
CFI | 0.9 | .923 | .9 More than | .947 |
NFI | 0.9 | .900 | .9 More than | .923 |
IFI | 0.9 | .924 | .9 More than | .948 |
RMR | .0.8 Below | .057 | .08 More than | .082 |
RMSEA | .1 Below | .027 | .1 More than | .025 |
Note: * :P < 0.1, **:P < 0.05, ***:P < 0.01 |
The Table 3 contains various indices used for evaluating the fit of a statistical model, particularly in the context of structural equation modeling (SEM). χ2 (Chi-Square): This criterion assesses the discrepancy between the observed and expected covariance matrices. A significant p-value (p < .05) suggests that the model does not fit the data well. In your table, the calculated χ2 value is 1036.915, which is significant (*** denotes significance). GFI (Goodness of Fit Index): GFI measures the proportion of the observed covariance that can be explained by the model. A value closer to 1 indicates a better fit. In the table (4), the GFI value is .774, which is below the desired cutoff of .9 (denoted as .9 More than). AGFI (Adjusted Goodness of Fit Index): AGFI is similar to GFI but adjusts for the degrees of freedom. It also ranges from 0 to 1, with values closer to 1 indicating a better fit. In your table, the AGFI value is .725, which is below the desired cutoff of .9 (denoted as .9 More than). CFI (Comparative Fit Index): CFI compares the hypothesized model with a baseline model where variables are uncorrelated. Higher values, closer to 1, suggest a better fit. In your table, the CFI value is .923, which is above the desired cutoff of .9 (denoted as .9 More than). NFI (Normed Fit Index): NFI is another measure of fit, ranging from 0 to 1. Values closer to 1 indicate a better fit. In your table, the NFI value is .900, which is above the desired cutoff of .9 (denoted as .9 More than). IFI (Incremental Fit Index): IFI compares the fit of the hypothesized model to a null model. Values closer to 1 indicate a better fit. In your table, the IFI value is .924, which is above the desired cutoff of .9 (denoted as .9 More than). RMR (Root Mean Square Residual): RMR measures the discrepancy between the observed and predicted covariances. Smaller values suggest a better fit. In your table, the RMR value is .057, which is below the desired cutoff of .08 (denoted as .08 Below). RMSEA (Root Mean Square Error of Approximation): RMSEA estimates the discrepancy between the predicted and observed covariances, adjusted for model complexity. Smaller values indicate a better fit. In your table, the RMSEA value is .027, which is below the desired cutoff of .1 (denoted as .1 Below). Based on the provided criteria, the model seems to have mixed results. The χ2 value is significant, indicating a poor fit. Additionally, the GFI and AGFI values are below the desired cutoff of .9, suggesting inadequate fit. However, the CFI, NFI, IFI, RMR, and RMSEA values meet the desired criteria, indicating a reasonable fit. It is important to consider all these indices collectively and in the context of your specific research question or hypothesis.
Table 4
Confirmatory factor analysis
Items | Estimate |
Socinf1 | Social Influence | .863 |
Socinf2 | .919 |
Socinf3 | .878 |
Sysuse1 | System Usefulness | .786 |
Sysuse2 | .796 |
Sysuse3 | .737 |
Sysuse4 | .739 |
Faccon1 | Facilitating Conditions | .759 |
Faccon2 | .715 |
Faccon3 | .783 |
Faccon4 | .757 |
Trust1 | Trust | .844 |
Trust2 | .859 |
Trust3 | .894 |
Trust4 | .402 |
Trust5 | .829 |
Hedmot1 | Hedonic Motivation | .851 |
Hedmot2 | .919 |
Hedmot3 | .909 |
Hedmot4 | .901 |
Risk1 | Risk | .874 |
Risk2 | .887 |
Risk3 | .890 |
Risk4 | .889 |
Att1 | Attitude | .829 |
Att2 | .852 |
Att3 | .877 |
Att4 | .850 |
Int1 | Intention to Use | .844 |
Int2 | .878 |
Int3 | .847 |
In the Table 4, Confirmatory factor analysis estimates the factor loadings for each item. The above values can be easily traced by the significance value. The factor loadings provide insights into how well each item measures its corresponding latent factor. Higher factor loadings indicate a stronger association between the item and the construct. Researchers use these estimates to evaluate the quality and validity of the measurement model in CFA and make decisions about item retention, construct refinement, or model improvement.
4.2 Analysis of Research Model
4.3 Structural equation model (SEM)
The structural equation model was used to assess the study hypotheses in the literature review after examining the initial measurement scales for validity and reliability (SEM).
The information includes the estimated values for several fit indices and the results of the chi-square test.
GFI (Goodness of Fit Index)
The estimated value for GFI is .774. To assess model fit, it is generally desired to have a GFI value greater than 0.9. In this case, the estimated value falls below the desired cutoff, indicating that the model does not fit the data particularly well according to the GFI criterion.
AGFI (Adjusted Goodness of Fit Index)
The estimated value for AGFI is .725. Similar to GFI, it is desirable for AGFI to be above 0.9 for a good model fit. However, in this case, the estimated value falls below the desired cutoff, suggesting that the model does not fit the data well according to the AGFI criterion.
NFI (Normed Fit Index)
The estimated value for NFI is .900. To have a good model fit, it is typically expected for NFI to exceed 0.9. In this case, the estimated value meets the desired cutoff, indicating a reasonable fit according to the NFI criterion.
IFI (Incremental Fit Index)
The estimated value for IFI is .924. Similar to the other fit indices, an IFI value above 0.9 is generally desired for a good fit. In this case, the estimated value meets the desired cutoff, suggesting a reasonable fit according to the IFI criterion.
CFI (Comparative Fit Index)
The estimated value for CFI is .923. Again, a CFI value greater than 0.9 is typically desired for a good model fit. In this case, the estimated value meets the desired cutoff, indicating a reasonable fit according to the CFI criterion.
χ2 (Chi-Square) Test
The chi-square test evaluates the discrepancy between the observed and expected covariance matrices. The calculated χ2 value is 1036.915, with degrees of freedom (df) equal to 309. The p-value associated with this test is .000, which is less than the conventional significance level of .05. This suggests that the model significantly deviates from the observed data, indicating a poor fit according to the chi-square criterion.
Based on the information provided, the model appears to have mixed fit results. While some fit indices, such as NFI, IFI, and CFI, suggest a reasonable fit, others, such as GFI and AGFI, indicate that the model does not fit the data particularly well. Additionally, the significant chi-square test further suggests a poor fit. It is important to consider all of these indices collectively and in the context of your specific research question or hypothesis.
Table 5
Construct | Social influence | System usefulness | Facilitating conditions | Trust | Risk | Attitude | Intention | AVE |
Social influence | 1 | | | | | | | 0.636955739 |
System usefulness | 0.273529 | 1 | | | | | | 0.506137786 |
Facilitating conditions | 0.151321 | 0.8464 | 1 | | | | | 0.461908331 |
Trust | 0.030276 | 0.322624 | 0.378225 | 1 | | | | 0.413083853 |
Risk | 0.165649 | 0.007921 | 0.016384 | 0.058081 | 1 | | | 0.507002122 |
Attitude | 0.151321 | 0.481636 | 0.467856 | 0.412164 | 0.015129 | 1 | | 0.655670512 |
Intention | 0.1444 | 0.597529 | 0.525625 | 0.373321 | 0.084681 | 0.743044 | 1 | 0.542836604 |
In the Table 5, the underlined values represent the average variance extracted (AVE) values for each latent construct. The AVE value for the Social Influence construct is 0.636955739, indicating that approximately 63.7% of the variance in Social Influence is accounted for by its indicators. The correlation coefficient between System Usefulness and Facilitating Conditions is 0.8464, suggesting a strong positive correlation between these two constructs.
The correlation coefficient between Trust and Attitude is 0.412164, indicating a moderate positive correlation between these constructs. The AVE value for Intention is 0.542, suggesting that approximately 54.3% of the variance in Intention is explained by its indicators. The correlation matrix provides information about the relationships between pairs of constructs, indicating whether they are positively or negatively correlated. The AVE values offer insights into the amount of variance in each construct that is accounted for by its indicators. These values help assess the reliability and validity of the measurement model.
4.4 Hypotheses Testing
Based on the acquired data, the study undertook statistical analysis to analyze consumers' adoption and acceptance of MPs. The effect of the six indicated parameters (system usefulness, social influence, facilitating conditions, trust, risk, hedonic incentive, and attitude) on the likelihood of future usage of MPs was estimated using multiple regression. Multiple regression is "a method for selecting variables for inclusion in the regression model that begins with the selection of the best predictor of the dependent variable" [74].
Table 6
Results of the hypotheses testing
Hypothesis | Relationship | S.E | C.R | p | Result |
H1 | Social influence → Intention to use | .076 | 2.619 | .009 | Supported |
H2 | System usefulness →Intention to use | .241 | 1.903 | .057 | Supported |
H3 | Facilitating conditions →intention to use | .046 | 1.081 | .280 | Rejected |
H4 | Hedonic motivation →intention to use | .078 | 7.647 | .000 | Supported |
H5 | Trust→Intention to use | .163 | -2.432 | .015 | Supported |
H6 | Risk→Intention to use | .029 | -3.452 | .000 | Supported |
H7 | Attitude→Intention to use | .101 | .8668 | .000 | Supported |
Table 6 shows the proposed model’s predictors and hypothesized relationships for their statistical significance, a bootstrapping procedure with 2000 resamples was used. The results illustrate that except one hypothesis, rest of the hypotheses are accepted as expected.