1.1 Mobile Commerce - Shopping Evolution

According to [62], the volume of m-commerce is growing at 39 percent each year and was estimated to have reached US$31 billion by 2016. The increased mobile phone usage during daily activities, and new areas of mobile integration such as m-Payments, reveals that the connected-consumer is here to stay [28].

Mobile devices have reached critical mass, with a majority of consumers, 80% of global consumers in both developed and high-growth economies, now owning a mobile phone. Still a relatively new application, m-Payment usage is gaining traction globally. 47% of the consumers in emerging markets reported using their phones to make in-store payments. In developed markets, 20% of the consumers reported using their phones to make in-store payments and 30% indicated interest. In developed markets, more northern Europeans use m-Payments for transport whereas central and southern Europe uses them more on food / shopping (lifestyle). One-third of e-commerce's business worldwide is transacted via mobile devices, and the number of smartphone and/or tablet users making purchases on their mobile devices will increase significantly in the years ahead [99].

With the consumer shift to mobile-first now fully underway, brands and marketers are constantly seeking to better understand consumer mobile preferences, transforming their mobile operating models, strategies and processes, while answering consumer demand for more personalized and contextually-relevant experiences, as well as instant access to products and services. When in stores, mobile shoppers tend to scan product barcodes and read reviews more often than computer-based shoppers and therefore "require different sales, marketing and service strategies" [32]. Mobile shoppers want the benefits of personalization without the downsides. Mobile consumers are 46% more willing than computer-based buyers to provide information, if the result is a more personalized experience. Therefore, they are willing to share their personal data and shopping preferences only within loyalty and rewarding programs [14].

So, it can be surely inferred that the m-shopping channel has emerged as an influential medium for connecting customers with retailers and ultimately in generating sales [112]. Given the continuing growth and the dynamics of m-commerce, its comprehensive understanding has become a pursuit of various researchers and practitioners. Thus, several studies have been carried out in the m-shopping field from different perspectives, including m-commerce theory, wireless network infrastructure, mobile middleware, wireless user infrastructure, and m-commerce applications and cases [74], [116]; trying to gain an improved insight into users’ mobile behavior. However, the vast majority of them have focused on mobile services in general, whereas the examination of mobile shopping in the field of conducting a monetary transaction has been slightly studied [116]. Identifying key influential factors that impact mobile users on m-commerce can definitely help involved enterprises to formulate a target-oriented business strategy in order to improve customers’ satisfaction. Therefore, the purpose of this paper is to find out users’ reaction towards different parameters that would influence their intention to utilize their mobile devices in order to purchase products and services online; focusing on the current Greek reality. More specifically, the study aims to construct an instrument in order to provide an explanation of the determinants of m-commerce acceptance. The study suggests a conceptual model that shows how its differ ent constructs influence users to adopt m-commerce. Based on Greece’s geographical and cultural characteristics, it is expected that the hypotheses used can explain the intention phenomenon. Such a study in the mobile domain has not yet been investigated for this culture. The paper is expected to both contribute research and practice by providing enlightenment and thoughts for the academia and industry to build strategies on m-shopping development.

1.2 Behavioral Intention Theoretical Background

A literature review on information systems, innovative technology, e-commerce and m-shopping highlights a number of theories or conceptual models, which are commonly used or extended with other variables for investigating various contexts of technology adoption and especially the users’ intention to adopt m-services, as well as consumers’ reaction towards factors affecting their intention to use their mobile devices in order to purchase products and services. Therefore, there are several behavioral intention theories, and the most popular and widely used are following.

The Diffusion of Innovations (DOI) perspective is introduced by Rogers [85] investigating a variety of factors which are considered to be determinants for the actual adoption and usage of Information Systems. According to DOI, potential adopters evaluate an innovation based on innovation attributes (relative advantage, compatibility, complexity, trialability and observability) [85]. The Theory of Reasoned Action (TRA) was first proposed by Fishbein and Ajzen [33], suggesting that users’ intention to adopt a technology is determined by two factors: personal in nature (attitude) and social influence (social or subjective norm). TRA was later evolved to the Theory of Planned Behavior (TPB) by adding perceived behavioral control to the initial determinants [4]. [4] has proved that perceived behavior control has a direct effect on behavior and desire through an empirical study, and proposed the Theory of Planned Behavior (TPB) based on the Theory of Reasoned Action (TRA). TPB emphasized the adopters’ attitudes towards technology, expressed social influence factors via the subjective regulated variables, and expressed the adopters’ personal characteristics via perceived behavior control. The TPB was also enriched with stable, decomposed beliefs structures for the TPB model and proposed the Decomposed Theory of Planned Behavior [88]. The Expectancy Confirmation Model (ECM) or the Expectancy Disconfirmation Model (EDM) has been used to explain the continuance behavior in electronic commerce. It is important for the m-shopping providers to understand the determinants of m-shopping continuance [16], [49].

Furthermore, the Technology Acceptance Model (TAM) indicates that perceived ease of use and perceived usefulness are the two main beliefs that determine one’s intention to use technology [26]. TAM has been the mostly constantly used or extended model regarding m-service adoption and m-shopping, comparing favorably to TRA and TPB [65], [86], [88], [108].

When deeper explanation of user adoption intention is desired, it allows other factors to be incorporated easily into its basic model [46]. Since the original proposal, TAM has been used and adapted extensively, and is seen as a leading model in explaining and predicting system use. In the context of mobile commerce, [107] used a revised TAM to explore the factors that determine user mobile commerce (including banking, shopping, investing, and online services) acceptance in Taiwan. They identified the following as factors affecting behavioral intention to use: perceived risk, cost, compatibility, and perceived usefulness. [104] used TAM, with the objective to predict consumer intention to use mobile services. [64] also used a revised TAM model to explore the factors affecting purchase intention on mobile shopping websites. Furthermore, [80] identified the availability of a mobile interface as a contributing factor to satisfaction with the online shopping experience. Additionally, [108] explored an integrative perspective, related to the adoption of mobile healthcare by hospital’s professionals.

Finally, the Unified Theory of Acceptance and Use of Technology (UTAUT) combined eight models, based on the above theories plus a few of their extensions, in a unified technology acceptance concept. The proposed unified model contains five determinants: performance expectancy, effort expectancy, social influence, facilitating conditions, and behavioral intention [98]. It is used so far for a number of technology types such as e-government [6], wireless Local Area Network (LAN) [8] and m-commerce [110].

When deeper explanation of mobile consumer adoption intention is desired, TAM allows other factors to be incorporated easily into its basic model [46]. Hence, in the current study, we keep the basic variables of TAM-perceived ease of use, perceived usefulness and behavioral intention-and through literature research we contribute with new variables-mobile skillfulness, enjoyment, anxiety, trust, innovativeness and relationship drivers-which are expected to have influence on the mobile shopping intention. All these constructs in this paper, are defined in a strict mobile context.

The rest of the paper is organized as follows: In the first section, the literature review regarding m-commerce adoption is presented. This is followed by an explanation of the hypotheses formed, based on literature review, and the description of the constructs that are included in the proposed model. In the third section a description of the applied methodology and the results of the empirical research are presented. The final section concludes with a discussion commenting on the data gathered and recommends some ideas for future research directions.

2.1 Behavioral Intention

In most of the well-established models of behavioral intention theories, such as the Technology Adoption Model (TAM) and the Theory of Reasoned Action (TRA), there has been an attempt to examine the factors that affect the consumers’ decision on using a technology studied [107]. [31] first defined the term Behavioral Intention to depict “a person’s subjective probability that he will perform some behavior”. [26], also, followed up with this idea to give shape to TAM, which finally concludes to the Actual System Use. Based on these concepts, in the paper herein, there is a construct included in the proposed model entitled Behavioral Intention to describe a person’s subjective probability that he/ she will purchase online via a mobile device.

2.2 Perceived Usefulness

Perceived usefulness has been a fundamental construct in many of the technology adoption models that have been proposed since 1989, when Davis first used this term [26]. It has been widely utilized in m-commerce area of study as well. For example, [53], [57], [59], [70], [71] confirmed that perceived usefulness is positively related to m-commerce behavioral intention. [5], [42], [90] proved that perceived usefulness positively affects mobile banking adoption and [53] confirmed that perceived usefulness affects mobile coupons adoption. Commonly, [115] confirmed that perceived usefulness positively impact on m-services’ behavioral intention and [19] approved its effect on m-payments. In this study, perceived usefulness refers to the degree which an individual believes that using a mobile device to shop online would enhance his/ her job performance. Thus, a mobile device with high perceived usefulness, sequentially, is one for which the m-user believes in the existence of a positive use-performance connection [26], [71]. Based on the aforementioned literature the following hypothesis can be stated:

H1: Perceived usefulness has a positive effect on m-shopping behavioral intention.

2.3 Perceived Ease of Use

Additionally to perceived usefulness, perceived ease of use has been a vital concept in many of technology adoption models regarding not only technologies and systems in general, but also m-shopping (e.g., [51], [92], [102]). For example, [24], [57], [71], [93] confirmed the significant impact of perceived ease of use on consumer’s intention to adopt m-commerce. Additionally, [115] presented its positively impact on m-services’ behavioral intention, [42], [66] and [103] its influence on m-banking adoption and [53] its influence on m-coupons adoption. However, a number of studies have confirmed that perceived ease of use can also indirectly impact behavioral intention through perceived usefulness. On these studies, researchers have proposed adoption intention models where perceived ease of use strongly impact on perceived usefulness. These investigations include a variety of mobile technology topics, such as m-commerce [7], [92], [106] and mobile data services [82], [115]. Furthermore, many researchers argued that perceived ease of use is not only considered as important factor for the adoption of a technology, but it affects long-term use of a technology as well [39], [50], [84]. Hence, it is a vital concern to examine the role of this construct to m-shopping.

Based on the aforementioned literature, the following hypotheses can be stated:

H2: Perceived ease of use has a positive effect on m-shopping behavioral intention.

H3: Perceived ease of use has a positive effect on perceived usefulness.

2.4 Anxiety

Several researchers have investigated anxiety as a significant inhibitor of IT and m-shopping adoption (e.g., [11], [12], [47], [64], [67]). Consumer anxiety about using IT focuses on individual’s ability and willingness to use technology-related tools [67]. Regarding m-commerce, according to [111], due to its newness and uniqueness in shopping encounters, m-commerce might provoke users’ anxiety to m-shopping adoption. [12] proved that anxiety impacts negatively on perceived ease of use of m-payments. Considering that m-commerce consists of data services that are not limited by temporal and spatial boundaries, consumer anxiety about using m-shopping is expected to exist and may be higher than anxiety about other shopping methods [111]. For example, while e-commerce demands a specific fixed local area network to access a website, m-commerce can be done everywhere via data services [44]. Furthermore, [11] mentioned that consumer anxiety about using data services is higher in the mobile environment because the responsibility for a failure or loss of a transaction may not be clear in this technology-mediated environment. [47] emphasized that anxiety increases since m-users are sensitive to services related to monetary transactions where they may lose both money and personal information. Taking into consideration the extremely personalized nature of mobile devices, it goes without saying that there is an increase in privacy and security concerns. Unfortunately, even skilled m-users may be frustrated while using m-commerce services because so many options have been placed into such a tiny device, and these “new capabilities often negatively affect users’ experience on other dimensions” [52] p. 7. As a consequence, individuals who are less anxious about using m-shopping also feel more comfortable utilizing it; so they are more likely to adopt m-commerce services. The volume of potential m-transactions is definitely expected to be higher in less anxious mobile users as well. Additionally, the more anxious the potential m-shopper is, the less easy its interaction with a mobile device for a monetary transaction will be. Thus, it is reasonably expected that he/ she will find it more difficult to shop via a mobile device than a free of anxiety individual. A stressful state of mind can undeniably disturb the easiness of m-shopping.

In this study, anxiety refers to negative emotions in cognitive states which evoked in actual or imaginary monetary online transactions via a mobile device [64]. Therefore, it is a vital concern to examine the role that negative emotional response plays in whether or not to adopt m-shopping.

Based on the aforementioned literature, the following hypotheses can be stated:

H4: Anxiety has a negative effect on m-shopping behavioral intention.

H5: Anxiety has a negative effect on perceived ease of use.

2.5 Mobile Skillfulness

Several studies have investigated skillfulness as a significant factor in influencing consumers’ behavior towards a specific technology (e.g., [25], [31]). Therefore, it is strongly recommended that skillfulness should also be examined as a driving factor in the area of mobile commerce (e.g., [64]). In this study, skillfulness refers to the individual’s assessment of his/ her ability to purchase online via a mobile device. It is assumed that the greater m-user’s capability belief the greater the chances to adopt and use mobile devices to shop online. For example, various studies found that the greater the user’s capability belief the greater user’s intrinsic motivation [22], [29], [64]. On the contrary, other researches have confirmed that people feel anxiety in their attempt to achieve particular behaviors that they do not perform skillfully [64]. Therefore, individual’s emotional arousal anxiety has been approved to negatively impact on his/ her perceived capability [22], [31], [64]. Furthermore, it is considered as important to understand the potential impact of mobile skillfulness to perceived enjoyment and perceived usefulness as well [9]. In specific, it is alleged that the greater the user’s capability belief the greater user’s gratification of m-shopping experience. Additionally, his/ her capability belief is expected to positively impact on perceived usefulness. Characteristically, [64] found that consumers’ lack of self-perception of mobile skillfulness can significantly reduce perceptions of enjoyment and usefulness. However, as far as it is concerned, both aforementioned factors have been slightly investigated by the academia regarding the m-commerce area of interest. Thus, based on the aforementioned literature, the following hypotheses can be stated:

H6: Mobile skillfulness has a positive effect on m-shopping behavioral intention.

H7: Mobile skillfulness has a negative effect on anxiety.

H8: Mobile skillfulness has a positive effect on perceived enjoyment.

H9: Mobile skillfulness has a positive effect on perceived usefulness.

2.6 Trust

In order to define and measure trust, there have been many suggestions in literature attributing it to meanings like privacy protection permitting a user to choose how his or her personal information is used [16], or perceived credibility showing that one partner believes that the other partner has the required expertise to perform a job effectively and reliably [20], [23]. [78] stated that trust in e-commerce is the belief that allows consumers to willingly become vulnerable to the online retailers after having considered the retailers’ characteristics including goodwill trust (benevolence) and credibility (honesty, reliability, and integrity). [70] divided the entity of trust in two sub-entities: trust in technology and trust in service providers. Trust in technology redirects to technical protocols, transaction standards, regulating policies, and payment systems [70], whereas, according to [16] analysis for e-commerce services, trust in service providers refers to ability -the user’s perception of provider’s competencies and knowledge salient to the expected behaviour, integrity -the user’s perception that the service providers will adhere to a set of principles or rules of exchange acceptable to the users during and after the exchange, and benevolence -the service provider is believed to intend doing good to the users, beyond its own profit motive.

In this study, the factor of trust refers to a) the security of mobile payments while shopping, b) the confidentiality of personal data (such as respecting consumers’ credit card details), c) the trustworthiness in the results after mobile purchases are conducted, and d) the full compliance with the terms & conditions of each mobile shop.

[70], [109] studying mobile commerce, [78] examining consumers’ acceptance of electronic commerce, [42], [90] investigating m-banking adoption, [21], [106] analyzing the m-commerce adoption, [91] contributing to e-banking, and [37] talking about Business-to-Customer (B2C) e-Services detected a positive influence of trust on consumers’ behavioral intention. In specific, [78] said that: trust reduces behavioral uncertainty related to the actions of the Web retailer, giving a consumer a perception of some control over a potentially uncertain transaction. On the other hand, [40] proved that there is an influence of trust on intention to e-shop through its influence on usefulness. This sense of overall control over their online transactions positively influences consumers' purchase intentions. Thus, it is reasonable to assume that there is, also, a positive relationship between trust and behavioral intention when it comes to the adoption of mobile shopping. So the following hypotheses can be stated:

H10: Trust has a positive effect on mobile shopping behavioral intention.

H11: Trust has a positive effect on mobile shopping perceived usefulness.

H12: Trust has a positive effect on mobile shopping perceived ease of use.

2.7 Relationship Drivers

[72] since 2000 has pointed out that one of the rules for a customer to feel closely related to a firm, or brand, or a process is the existence of distinctive competencies. Hence, we need to examine whether the distinctive competencies of the mobile shopping process cater to the specific interests and needs of the customers and lead to the development of solid relationships with them.

Experts in the marketing field have recognized some unique features of the mobile technology, which can help with the relationship building between the customers and the brand. To start with, the functions of SMS and MMS exchange give the mobile services the exclusiveness of ubiquitous and universal information accessibility, of information personalization and information dissemination [77]. [115] also mention that mobile and wireless technology do not set information access restrictions regarding time and space [13]. Mobile services’ location and time personalization feature enables individuals to ask for directions to the nearest gas station while traveling, or can receive last-minute alerts and flight delay notices through their mobile devices.

The fact that consumers can reach the information and use the applications from everywhere [77], [80] and anytime [58], [78], [114], gives a strong advantage for time-sensitive and location-based content and services to be received on consumers’ mobile devices at the point accustomed to their personal identity [30], [58], [77], [105]. Thus, personalized information and/or services are able to be achieved. [87] examining the success factors of mobile marketing, highlight ‘personalization’ as the key feature of mobility that drives consumers to shop. This statement is strongly reasoned, as they present scenarios, like the SMS advertising campaigns, which can be well-targeted to categorized groups, according to customers’ profiles and their previous actions, e.g. past leisure activities, music interests, or occupation (preferential treatment).

Also, it gets quite common for the brands to prefer a mobile application as their sales channel instead of a physical or an online store, as they can be in the customers’ hands anytime [79]. [87] concludes that a successful mobile advertising practice is based on the fact that some common mobile applications urge the recipients to act on the spot. For example, proximity marketing through Bluetooth (mainly in the past), or applications, which use iBeacons to transmit marketing content- e.g special discounts and coupons- to passers by, are common scenarios that trigger consumers and drive them to proceed with a purchase.

Additionally, mobile couponing is a common marketing strategy to increase mobile transactions [1], [76]. Mobile coupons can be delivered directly to the mobile device, e.g via an SMS and are stored in the mobile phone’s memory, or even to dedicated mobile applications (e.g. Apple Wallet). Hence, they are hard to get lost or forgotten, and can be easily redeemed. If consumers are subject to get rewarded with mobile coupons, discount prices, prizes, newsletters, free call time, etc. when they participate in a mobile transaction, they would be motivated to use the mobile services more frequently [9].

All the above features are used to drive the user to a close relationship with the mobile services as they make the shopping experience meaningful and increase engagement with mobile transactions. Hence, the user integrates them in his daily activities, gets attached to them, and spends a lot of time conducting such services. We hypothesize that such relationship drivers affect individuals’ mobile shopping intention, as well as mobile shopping perceived usefulness.

H13: Relationship Drivers have a positive effect on mobile shopping behavioral intention.

H14: Relationship Drivers have a positive effect on perceived usefulness of mobile shopping.

2.8 Enjoyment

Traditionally shopping is classified among the activities of fun and pleasure that leads to feelings of joy [55]. The retail literature has also focused on the notion of excitement [100]. However, [43] trying to define enjoyment claim that excitement corresponds to an emotional state that elicits high levels of pleasure and arousal, whereas enjoyment refers to a less emotional motivation correlated with consumers’ logic, which lets them decide their likings while shopping.

Non-store formats of shopping can offer satisfaction to profiles of shoppers who are internet literate and technology fans. Those types of consumers can correlate the shopping process with a shopping channel, which is both easy for them, as well as offers them pleasure. That is the reason that enjoyment has been found to be a factor with a positive impact on online shopping intention [83], [95]. Also, [64] in their study verify their hypothesis that enjoyment is a consumer’s intrinsic motivation that strongly affects the intention to engage in mobile shopping. [60] in their model verified that there is a significant effect of perceived post adoption enjoyment on the continued adoption of mobile services.

Additionally, apart from the studies that verify the direct effect of enjoyment on the intention to use factor of the TAM model, there are analyses that prove the correlation between enjoyment and perceived usefulness and perceived ease of use. To be more specific, [60] verify that if SMS services are perceived as enjoyable, the value perception of SMS as useful and easy to be used is increased. Also, [96] conducted a study in Singapore proving that perceived enjoyment is a significant predictor of perceived usefulness, perceived ease of use, and intention to use technology. Anyway, [27] observed the positive interaction between enjoyment and perceived usefulness, while studying the motivations to use computers in the workplaces.

H15: Enjoyment has a positive effect on mobile shopping behavioral intention.

H16: Enjoyment has a positive effect on perceived usefulness of mobile shopping.

H17: Enjoyment has a positive effect on perceived ease of use of mobile shopping.

2.9 Innovativeness

Innovativeness as a factor has been integrated in many previous technology acceptance studies. In information technology terminology innovativeness is used as the “willingness of an individual to try out any new information technology” [34], [69]. In free interpretation in the field of technology acceptance, innovation refers to the degree of interest in trying a new thing, new concept, or innovative product or service [85]. [2] first introduced innovativeness construct and its impact on information technology in 1998. Individuals, who are considered as competent technically by their social environment, tend to find that using new technology is less troublesome suggesting a direct positive effect of innovativeness on perceived ease of use [113].

There are more frameworks verifying this impact of innovativeness on both factors: behavioral intention and perceived ease of use. To be more specific, [101] concluded that innovativeness has a positive impact on perceived ease of use and negative impact on perceived usefulness regarding the adoption process of IT from service employees. The study by [63] showed that innovativeness has a positive impact on perceived ease of use of 3G mobile value-added services, whereas the influence on perceived usefulness of 3G mobile value-added services is insignificant. Additionally, [57] confirmed that innovativeness impact positively on perceived ease of use of m-commerce. [18] discussed the importance of innovation for the mobile telecom industry, as well as, [3] confirmed the positive impact of personal innovativeness on behavioral intention to use mobile services in libraries. Furthermore, [56] showed that innovativeness has a positive effect on both perceived ease of use and continuance intention to use m-commerce.

The positive results of studies around the direct effect of innovativeness on mobile commerce [65], [110] and on mobile shopping [7] lead the authors to hypothesize that:

H18: Innovativeness has a positive effect on mobile shopping behavioral intention.

H19: Innovativeness has a positive effect on perceived ease of use of mobile shopping.

All the hypotheses formed above, are summarized in Table 1.

Figure 1: Hypothesized model (Model 1) 

3.1 Operationalization of Variables

Operational definitions of the study variables are shown in Table 2. For each variable, a multiple-item scale was developed based on literature review; where each item was measured based on a 5-point Likert scale, ranging from 1-Completely Disagree to 5-Completely Agree. Three items were used to measure anxiety, skillfulness and enjoyment, four items were used to measure perceived ease of use, trust, innovativeness, relationship drivers and behavioral intention, whereas five items were used to measure perceived usefulness.

3.2 Data Collection and Sample Characteristics

Data were collected through a structured electronic questionnaire which was uploaded on a website for three month, from February to April 2016. Additionally, contacts from various mailing lists were asked to fill in the questionnaire, as well as members of two popular social networks-Facebook and Twitter-were encouraged to participate in the survey. The questionnaire was based on prior surveys approved for their validity and reliability and was pretested before being widely distributed. A pilot study using a sample of thirty responses helped to identify possible problems in terms of clarity and accuracy Thus, comments and feedback from respondents improved the final presentation of the items. Thirty two participants gave incomplete answers and their results were dropped from the study. As a result, the final sample consisted of 473 respondents.

Table 1: Research hypotheses in the current study 

Table 2: The operational definition of research variables 

The demographic profile of the respondents, presented in Table 3, indicates that 49.3% were male and 50.7% female. The vast majority (87.7%) was between 18-34 years old, whereas only 2.7% was above the age of 44. In terms of their educational background, about 80% had received a higher education degree, whereas with respect to their occupation more than half (54%) were students, followed by private employees (21.6%) and public servants (8.9%). Finally, the majority of respondents reported a monthly income less than 900 €, whereas about one out of four (26%) did not reveal their income.

3.3 Data Analysis Plan

Structural equation modeling (SEM) using maximum likelihood estimation was used to test the hypothesized model of Figure 1. SEM is appropriate to test hypothesized structural relations between multiple independent and dependent variables (through the use of estimated regression parameters). Although the use of SEM in cross-sectional designs cannot lead to establishing causal pathways, it is useful for testing the relative fit of a given causal model to an existing data set. SEM is comprised of a measurement model and a structural model.

The measurement part of the model allows for the estimation of latent variables reflecting constructs of interest. The measurement model for all nine constructs was first assessed using Confirmatory Factor Analysis (CFA). CFA assumes each manifest variable to be a distinct indicator of an underlying latent construct, whereby different constructs are permitted to be inter-correlated. Further, reliability, convergent and discriminant validity of the constructs were evaluated. Convergent validity uses three recommended standards to assess the measuring model: (1) all indicator factor loading values should exceed 0.5 [41]; (2) composite reliability (CR) should exceed 0.6 [10]; and (3) the average variance extracted (AVE) of each construct should exceed 0.5 [35]. Discriminant validity was also assessed by comparing the correlations between constructs with the square roots of AVE values [35]. In order to detect common method bias, post hoc Harman’s one factor analysis was conducted to check whether variance in the data can be largely attributed to a single factor [81].

After the constructs have met the required measurement standards, the relationships between the constructs were estimated. This constitutes the structural model, which is tested to investigate the strength and direction of the relationships between the theoretical constructs. A model’s overall goodness-of-fit can be assessed using a combination of measures. The basis for an adequately fitted model was a chi-square/df ratio less than five [15], adjusted goodness-of-fit index (AGFI), comparative fit index (CFI), normed fit index (NFI), incremental fit index (IFI) and Tucker-Lewis index (TLI) values greater than 0.90 [48], a root mean square error of approximation (RMSEA) less than 0.05 [17] and a standardized root-mean square residual (SRMR) less than 0.08 [48]. The stability of the model estimates was tested via a bootstrap resampling procedure (500 sub-samples). All SEM analyses were conducted in Mplus version 7.31 [73].

Table 3: Demographic characteristics of the study participants (n = 473) 

4.1 Measurement Model

The measurement model (i.e. Confirmatory Factor Analysis or CFA model) included 34 items describing the nine latent constructs presented in Table 3. The model’s overall goodness-of-fit was satisfactory (Table 4): chi square/df, Standardized Root Mean Square Residual (SRMR), and (Root Mean Square Error of Approximation) RMSEA are less than the recommended values while Goodness of Fit Index (GFI), Adjusted Goodness of Fit (AGFI), Comparative Fit Index (CFI), Normed Fit Index (NFI), Incremental Fit Index (IFI) and Tucker-Lewis Index (TLI) are greater than the suggested threshold. The standardized indicator factor loadings for all observed indicators, shown in Table 5, were all statistically significant and exceeded 0.5. Skewness and kurtosis values for all items were less than three, which revealed no significant departures from normality [61]. Composite reliability (CR) of constructs ranged from 0.76 to 0.95, above the recommended threshold of 0.6. Cronbach's alpha values were also high in all cases (>0.80). Average variance extracted (AVE) values were greater than the recommended 0.5 cut-off, ranging from 0.51 to 0.86. Therefore, all the conditions for convergent validity were met. Table 6 shows that the the square roots of AVE were all greater than the inter-scale correlations (off-diagonal elements) in the corresponding rows and columns, and therefore, discriminant validity was supported. Furthermore, an un-rotated principal component factor analysis revealed nine factors with eigenvalues greater than 1. The first factor accounted for 29.7% of covariance among the measures at maximum. The other eight factors contributed 37.5% of the remaining variance, each accounting for 3-8%. Thus, common method bias is not likely a contaminant of our results. In addition, an inspection of modification indices did not suggest any model revisions that were substantially meaningful.

Table 4: Evaluation of model goodness-of-fit 

a[15], c[17], b[48]

Table 5: Standardized factor loadings and individual item reliability 

Table 6: Inter-correlations and square roots of average variance extracted 

Note: The diagonal elements show the square root of the average variance extracted. The off diagonal elements show the correlations between the constructs, all significant at 0.01 level.

4.2 Structural Model

Once the measurement model was correctly specified, the structural model of Figure 1 was estimated to provide an empirical measure of the hypothesized relationships among the research variables and constructs (standardized parameters are shown; solid lines represent significant pathways and correlations between constructs are not shown for clarity). Inspection of the goodness-of-fit measures also provided evidence of a good model fit (see Table 4).

Figure 2 shows the path diagram with the resulting standardized regression coefficients that are used to indicate the direction and magnitude of relationships among variables. The estimation of the structural model indicates that 14 out of 19 hypotheses were supported (Table 7). The exceptions are the direct effects of perceived ease of use, anxiety, mobile skillfulness and trust on behavioral intention (H2, H4, H6 and H10), as well as the effect of trust on perceived ease of use (H12). On the contrary, as expected, perceived usefulness exerts a strong positive effect on behavioral intention (H1; β = 0.245, t = 3.237, p = 0.001) and perceived ease of use has a strong positive impact on perceived usefulness (H3; β = 0.434, t = 5.794, p = 0.001). Anxiety was found to have a strong negative impact on perceived ease of use (H5; β = -0.354, t = -5.045, p < 0.001), thereby supporting our hypothesis. Mobile skillfulness was shown to negatively influence anxiety (H7; β = -0.738, t = -15.718, p < 0.001), and to positively affect perceived enjoyment (H8; β = 0.573, t = 14.013, p < 0.001) and perceived usefulness (H9; β = 0.372, t = 2,810, p = 0.005). As anticipated, trust showed a positive relationship towards perceived usefulness (H11; β = 0.179, t = 2.145, p = 0.028). Relationship drivers were found to have a positive impact on behavioral intention (H13; β = 0.174, t = 2.128, p = 0.033) and a strong positive impact on perceived usefulness (H14; β = 0.294, t = 5.017, p < 0.001). Moreover, enjoyment has a positive effect on behavioral intention (H15; β = 0.277, t = 3.647, p < 0.001), perceived usefulness (H16; β = 0.511, t = 8.537, p < 0.001) and perceived ease of use (H17; β = 0.215, t = 3.985, p < 0.001). Finally, innovativeness was found to have marginally significant positive effects on behavioral intention (H18; β = 0.200, t = 2.016, p = 0.044) and perceived ease of use (H19; β = 0.186, t = 2.085, p = 0.037). All in all, over 80% of the variance in mobile behavioral intention was explained collectively by perceived usefulness, mobile skillfulness, enjoyment, relationship drivers and innovativeness.

Table 7: Path coefficients (standardized regression coefficients) 

Note: *p < 0.05; **p < 0.01; ***p < 0.001; ns = non-significant

*p < .05, **p < 0.01, ***p < 0.001.

Figure 2 Results of structural model 

5.1 Limitations and Future Work

Despite the fact that the aforementioned results provide meaningful implications, the research can be further improved by overtaking some important limitations. It would be of high interest, to investigate the behavior of the model after the addition of other constructs suggested by the literature, such as social influence drivers. Social networking might have a positive influence on behavioral intention regarding mobile shopping and affect the relationships between existing factors. Moreover, further evidence from heterogeneous samples is needed to ensure that the observed relationships are similar across different gender, education or age groups. Additionally, it is to be emphasized that these findings are limited to a sample of Greek consumers. Similar studies could be conducted in other countries, taking into consideration the different cultural notions in order to get comparative data for a cross-cultural study or a meta-analysis of existing studies. Such a comparison could be beneficial for the mobile marketing strategy targeting to multicultural mobile products and services. As the consumer’s cultural background is one of the aspects which can influence the creation of a favorable climate for developing and consolidating electronic transactions, we consider that another interesting line of research would be to contrast the validity of the proposed behavioral model with samples of consumers from other cultures and compare the results obtained.

Furthermore, regarding future research, we think it would be very useful to complement this study with the development and validation of a scale to measure m-shopping motivations and perceived risk, taking into consideration attitudinal variables such as motivations and barriers on mobile shopping decisions.

There is no doubt that the success of this new direct sales channel will be conditioned by both the group of product/ service offered and their cost.