Asia Pacific Journal of Marketing and Logistics Customer online shopping anxiety within the Unified Theory of Acceptance and Use Technology (UTAUT) framework Hakan Celik

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

Article information: To cite this document: Hakan Celik , (2016)," Customer online shopping anxiety within the Unified Theory of Acceptance and Use Technology (UTAUT) framework ", Asia Pacific Journal of Marketing and Logistics, Vol. 28 Iss 2 pp. 278 - 307 Permanent link to this document: http://dx.doi.org/10.1108/APJML-05-2015-0077 Downloaded on: 17 March 2017, At: 03:08 (PT) References: this document contains references to 125 other documents. To copy this document: [email protected] The fulltext of this document has been downloaded 1284 times since 2016*

Users who downloaded this article also downloaded: (2015),"The unified theory of acceptance and use of technology (UTAUT): a literature review", Journal of Enterprise Information Management, Vol. 28 Iss 3 pp. 443-488 http://dx.doi.org/10.1108/ JEIM-09-2014-0088 (2015),"Consumer attitudes towards online shopping: The effects of trust, perceived benefits, and perceived web quality", Internet Research, Vol. 25 Iss 5 pp. 707-733 http://dx.doi.org/10.1108/ IntR-05-2014-0146

Access to this document was granted through an Emerald subscription provided by emeraldsrm:602779 []

For Authors If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service information about how to choose which publication to write for and submission guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.

About Emerald www.emeraldinsight.com Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online products and additional customer resources and services. Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation. *Related content and download information correct at time of download.

The current issue and full text archive of this journal is available on Emerald Insight at: www.emeraldinsight.com/1355-5855.htm

APJML 28,2

278

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

Received 9 May 2015 Revised 14 August 2015 9 September 2015 Accepted 15 September 2015

Customer online shopping anxiety within the Unified Theory of Acceptance and Use Technology (UTAUT) framework Hakan Celik Department of Marketing, Bilecik Seyh Edebali University, Bilecik, Turkey Abstract Purpose – Few studies have investigated how anxiety operates within the Unified Theory of Acceptance and Use of Technology (UTAUT) framework. Consequently, the purpose of this paper is to explore the influence of anxiety on the customer adoption of online shopping based on the UTAUT. Design/methodology/approach – The UTAUT’s framework was extended by proposing new casual pathways between anxiety and its existing constructs (e.g. effort expectancy (EE), performance expectancy (PE) and behavioural intentions (BI)) within the contingencies of age, gender and experience. The partial least squares technique was employed to evaluate the statistical significance of the proposed pathways by analysing 483 sets of self-administrated survey responses. Findings – The results indicate that anxiety simultaneously exerts negative direct influences on PE, EE and BI constructs. While the moderating effects of age, gender and experience on the anxiety-intention link were found to be significant, there was no evidence suggesting that they moderate anxiety-PE and anxiety-EE relationships. Research limitations/implications – The limitations of the current study are inherent in its design and methodology, providing some directions for future research. Originality/value – This study contributes to the theory by including anxiety in the UTAUT and applying it to the online shopping context. The evidence about the significance of anxiety, with contingencies regarding age, gender and experience, supplies practical implications for online marketing strategies. Keywords Anxiety, UTAUT, PLS, Online shopping Paper type Research paper

Asia Pacific Journal of Marketing and Logistics Vol. 28 No. 2, 2016 pp. 278-307 © Emerald Group Publishing Limited 1355-5855 DOI 10.1108/APJML-05-2015-0077

1. Introduction Online retailing has noticeably impacted the world’s economy. Specifically, online retail sales reached $632 billion globally in 2012, up from $319 billion in 2008 (MarketLine, 2013). Some analysts estimate that the number of online shoppers worldwide will grow from 1.08 trillion in 2013 to 2.49 trillion in 2018; this represents a compound annual growth rate of almost 10 per cent (eMarketer, 2014). However, a recent international survey conducted in 39 countries shows that customers’ concerns about network security and privacy protection remain as serious obstacles to their uptake (Cole et al., 2013). Furthermore, the spatial and temporal separations between the buyer and the retailer, encompassed in the intangible nature of the virtual environment, heightened customer uncertainty and risk perceptions regarding online transactions. Thus, customer anxiety towards online shopping and its inherent uncertainty and risk has emerged as an important issue for retailers (Celik, 2011), as it decreases the degree of online shopping penetration and the amount of customer spending on online purchases. The increasing prominence of online retail has increased interest among researchers and practitioners in the customer adoption of this relatively new shopping medium. A growing number of studies have embedded online shopping adoption into various

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

theoretical perspectives from social psychology and the information systems fields to explore its determinants. While such models have facilitated substantial progress in the apprehension of the phenomenon over the years, an integrative approach to the current state of knowledge was not achieved until the introduction of the Unified Theory of Acceptance and Use of Technology (UTAUT), conceived by Venkatesh et al. (2003), to predict employee adoption of information technologies. It was also successfully applied in studying the adoption of other technologies in the non-work environment. Although few studies have extended the applicability of the UTAUT in the online shopping context, it is a valid and parsimonious model that can be used to explain customer purchase intentions and actual internet purchasing behaviour. Even UTAUT proponents have endorsed the systematic investigation of salient constructs (Baron et al., 2006). In addition, the theorization of alternative mechanisms based on these constructs fosters the utility of the UTAUT in the customer domain (Venkatesh et al., 2012). As such, anxiety becomes an especially relevant aspect in UTAUT online shopping research, because customers tend to consider their exposure to adverse outcomes when making purchases in an inherently risky environment (Featherman and Pavlou, 2003). Since the UTAUT was originally formulated and cross-validated in organizational contexts, it largely focused on employees’ cognitive and behavioural responses towards the new technology (Venkatesh et al., 2012). In the consumer context, unlike in the organizational context, the positive (e.g. enjoyment, fun and playfulness) and negative (e.g. fear, apprehension and anxiety) affective responses of the users play an important role in the acceptance and use of the technology (Celik, 2011). The integration of anxiety into the UTAUT will remedy its lack of an effective response focus. Consequently, this study aims to test the inhibiting effects of anxiety on two customer outcome beliefs (e.g. performance expectancy (PE) and effort expectancy (EE)) and the usage intentions of online shopping within the UTAUT framework, with contingencies regarding age, gender and experience. The remainder of this study is organized as follows. Section 2 introduces the theoretical background of this study, including a brief review of existing literature on UTAUT and anxiety within the online shopping context. Section 3 presents the research framework and hypotheses developed by employing the previous research in online shopping anxiety and the elements of UTAUT. Then, the research design is described in Section 4, followed by the presentation of analytical results in Section 5. The study findings are discussed from theoretical stand point in Section 6, followed by the limitations of the study and suggestions for future research in Section 7. Finally, the implications of these findings for practice are drawn in the last section. 2. Theoretical background 2.1 UTAUT and anxiety Since the proliferation of online retailing, several social psychology and information systems theoretical models applied to explain and predict consumer adoption of online shopping include the technology acceptance model (TAM) (O’Cass and Fenech, 2003; Singh et al., 2006), the TAM2 model (Zhang et al., 2006), motivation theory (Shang et al., 2005; Mohd Suki et al., 2008), the theory of planned behaviour (TPB) (Keen et al., 2004; Ramus and Nielsen, 2005), the decomposed TPB model (Lim and Dubinsky, 2005; Pavlou and Fygenson, 2006), the model of PC utilization (Klopping and McKinney, 2004), the innovation diffusion theory (Eastin, 2002; Hansen, 2005) and the social cognitive theory (SCT) (Foucault and Scheufele, 2002; Oyedele and Simpson, 2007). Previous studies have progressively contributed to the existing knowledge of the focal

Customer online shopping anxiety 279

APJML 28,2

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

280

behaviour; however, they fell behind in developing a more comprehensive view through the cross-theoretical integration of these diverse approaches. Consequently, Venkatesh et al. (2003) formulated the UTAUT based on a conceptual and empirical synthesis of the aforementioned models, which provided a coherent theoretical perspective in studying online shopping adoption. The UTAUT posits that an individual’s adoption of a new technology is a function of four core determinants: PE, EE, social influence (SI) and facilitating conditions (FC). According to the model’s structure, behavioural intentions (BI) to use technology are directly predicted by PE, EE and SI, whilst BI and FC directly determine actual use. The UTAUT also conceptualizes four individual difference variables (e.g. gender, age, experience and voluntariness) as moderators of these key relationships between the model’s constructs. It has been validated across a broad range of research settings, which illustrate that the model consistently explains over 50 per cent of the variance in BI and 30 per cent of the variance in technology use, even in different environments (e.g. organizational vs non-organizational), contexts (e.g. voluntary vs mandatory adoption) and cultures (e.g. individualist vs collectivist). A similar pattern of results regarding the model’s explanatory power has been also observed in online shopping behaviour studies applying the UTAUT framework. However, these studies indicate that the UTAUT has already reached its limits in terms of explaining the focal behaviour. As such, it needs to be extended by integrating new constructs into its structure for the advancement of the theory building progress in the online shopping adoption domain (Pahnila et al., 2011). A research effort to integrate the affect components into a cognition and behaviour-based UTAUT can alter the existing relationships between its constructs and enhance its generalizability to the consumer context. Anxiety, as a negative affective response of end users towards new technology, has received considerable attention in the technology adoption studies (Powell, 2013). The online shopping environment, unlike its organizational counterparts, is an extremely fertile atmosphere for customer anxiety, due to its inherent risks and dangers (e.g. privacy infringement, credit card fraud). 2.2 Reassessing the role of anxiety in online shopping In terms of SCT, anxiety is a negative affective reaction that adversely influences an individual’s determination to perform a specific act by both constraining his/her judgment about the personal capability to produce the performance through emotional arousal and lowering his/her expectations about the desired performance results (Compeau et al., 1999). The manifestation of anxiety can be classified into two broad categories: trait anxiety and state anxiety (Igbaria and Iivari, 1995). Trait anxiety refers to an individual’s personality characteristic reflecting his/her relatively stable negative attitudes towards certain external stimuli or situation. State anxiety corresponds to an individual’s temporary emotional distress to a particular external stimulus or situation (Gilbert et al., 2003; Saadé and Kira, 2006). Computer anxiety is a specific form of state anxiety manifesting itself as an individual’s transitory tendency of being fearful, apprehensive, intimidated, uneasy and aggressive when interacting with the functional (software) and mechanical (hardware) aspects of computers (Celik, 2011). Russell and Bradley (1997) state that computer anxiety stems from the worry of completing a computer-related task (task anxiety), the possibility of damaging computer equipment or losing important information (damage anxiety) and embarrassment due to the unexpected public exposure of computing incompetency (social anxiety). Task anxiety is closely related to online shopping, because it requires customers to interact

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

with virtual storefronts via the internet’s communication infrastructure using various hardware (e.g. personal computers and mobile devices), software (e.g. operating systems and web browsers) and protocols (e.g. file transfer and transmission control), amid shopping tasks. Customers exhibited more anxiety about online retailer transactions and may refrain from shopping online if they experience uneasiness during shopping tasks due to access difficulties, navigational problems, inconvenient checkout procedures, poor interface designs and outdated information content (Vijayasarathy, 2004). Damage anxiety is also prevalent in online transactions (Kim and Forsythe, 2008; Perea y Monsuwé et al., 2004), perhaps due to the absence of interpersonal interactions, tangible inspection, information symmetry, mutual sales contracts and solid security in the online shopping environment (Featherman and Pavlou, 2003; Montoya-Weiss et al., 2003; Park et al., 2004; Suh and Han, 2003). Social anxiety is the most distant influencer of customer online purchase decisions, thus the working definition of anxiety used in this study refers to a customer’s tendency to experience some degree of fear, apprehension and aggression upon his/her impending or proceeding online purchase. A recent meta-analysis of 276 studies by Powell (2013) shows that anxiety directly and indirectly influences individual acceptance and use of information technology. However, empirical support for its direct effect is less significant than that of its indirect effect on adoption behaviour. The explanation for this comparative insignificance addresses the control-process perspective on anxiety, suggesting that people who experience self-doubt in handling the threatening situation exhibit either physical withdrawal (e.g. avoidance of using the system) or mental withdrawal (e.g. attention to task-irrelevant thinking during the system use) (Smith and Caputi, 2001). If the physical withdrawal is not possible, due to some circumstances (e.g. organizational rules mandating the technology use or expected benefits of technology use greater than its risks and dangers), the mental withdrawal goes into effect with its outcome: less effort for and attention to the computer-mediated task. Usage pattern research of self-service technologies, including online shopping, provides empirical support for the effect of mental withdrawal in the customer context. It states that the high level of anxiety causes low levels of customer engagement, but not total disengagement with self-service technologies (Meuter et al., 2003). The anxiety construct itself has been empirically illustrated to exert a direct negative influence on the BI (Chiu and Wang, 2008; Fillion et al., 2012). The BI is an additive function of individual- and social-related factors and a transition between the cognitive and evaluative products and technology use. Like other individual- and social-related expectancies and responses, BI mediates the influence of these anxiety outcomes on the actual behaviour. In addition, anxiety negatively influences individual perceptions of effort requirements and performance gains associated with the use of technology. According to the processing efficiency perspective introduced by Eysenck and Calvo (1992), anxiety increases the required effort to complete a computer-mediated task through the presence of worry about task performance. The interference of task-irrelevant thoughts makes individuals allocate and direct the extra processing resources to eliminate their negative effects and improve task performance. However, worry and intrusive thoughts eventually impede task performance; even so, additional individual resource allocations are conducted by suppressing their cognitive capacities to process and store task-relevant information and divert their attention to the self-deficit (Fakun, 2009; Mead and Drasgow, 1993; Smith and Caputi, 2001). Although derived from anxiety research on various information technologies, similar logic could be applied to online shopping as a task mediated by computer and communication technologies.

Customer online shopping anxiety 281

APJML 28,2

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

282

Individual differences, such as age, gender and experience, are found to influence the magnitude of the anxiety effect on an individual’s adoption and use of computer-related technologies. The literature suggests that older people tend to exhibit high levels of computer anxiety and resistance to adopt new information technology. An increased age brings difficulties in processing task-related information, allocating attentional resources for task performance and acquiring the required computing skills for task completion (Venkatesh et al., 2003; Arning and Ziefle, 2007). Gender is also associated with anxiety in the technology adoption literature, where females exhibit more computer anxiety than males (Todman, 2000). Men were found to exhibit more enthusiasm in exploring technology features (He and Freeman, 2010), get more support from their social environment about the technology (Busch, 1995) and feel more comfortable with the technology, because of their computing experience (Karsten and Schmidt, 2008). Contrary to expectations that the gender gap for computer anxiety will diminish over time due to the ubiquitous computing of daily life, it instead has widened (Todman, 2000). Computer anxiety and personal experience with computers have demonstrated an inverse relationship, meaning that as computer experience increases, the level of anxiety decreases (Igbaria and Chakrabarti, 1990). This is because computing experience significantly contributes to the development of individual computertechnology-related self-efficacy perceptions (Brown et al., 2004; Ong and Lai, 2006). Self-efficacy eventually counterbalances the negative emotional effect on the cognitive effort to allocate and process the attentional resources for the accomplishment of a computer-mediated task that, in turn, produces a more favourable EE and attitude towards computing (Venkatesh, 2000; Hackbarth et al., 2003; Schottenbauer et al., 2004). The study findings emphasize the relationships between computer anxiety and individual differences; as such, they can be used as the theoretical underpinnings for such relationships in the online shopping context. 3. Research framework and hypotheses The initial research model was based on the UTAUT, with some extensions and modifications. The model posits additional causal pathways by which anxiety influences EE, PE and BI constructs (Figure 1). These causal relationships between the new and existing constructs are moderated by age, gender and experience. Regarding the modifications, voluntariness and its intervening influences on the relationships between UTAUT constructs were excluded from the model as irrelevant constructs for online shopping. In the following sections, the related hypotheses were justified to provide the theoretical rationale for the research model. 3.1 PE PE is the degree to which individuals believe that using a technology will increase their task performance. This aspect was formulated for the UTAUT through the aggregation of five constructs: embodied perceived usefulness, job-technology fit, extrinsic motivation, relative advantage and outcome expectations in the different models (Venkatesh et al., 2003). Like its referent constructs, PE suggests that individuals evaluate their technology-mediated task performances in terms of the associated benefits (i.e. facilitation of efficiency, effectiveness and productivity in task performance) and costs (i.e. cognitive, behavioural or financial investments made for special tasks) (Perea y Monsuwé et al., 2004). If the cost is lower or the benefit is higher,

Customer online shopping anxiety

Gender Performance Expectancy

Behavioural Intentions

Anxiety

Use

283

Effort Expectancy

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

Age

Social Influence

Facilitating Conditions

Experience Dotted Lines: New Relationships Intact Lines: Existing Relationships

the utilitarian value of the technology will be greater, and the intention to use it will be positive. Organizational and non-organizational studies empirically tested and confirmed this proposition (Alwahaishi and Snásel, 2013; Brown et al., 2010; Sin Tan et al., 2013). A similar pattern is expected in online shopping use. Customer expectations or realizations of the utilitarian value associated with online shopping (e.g. time saving, bargain dealing, round-the-clock convenience, broad product availability and hassle-free shopping) significantly evoke online purchase intentions (Celik, 2011; Zhou et al., 2007). The UTAUT also conceives that PE influence is moderated by gender and age. Research has shown that the effect is particularly stronger for younger men, because they are more driven by instrumental benefits, concerned over performance achievement, desirous for task success and skilful in acquiring technology-related knowledge or handling the functions of technology than are women and older users (Arning and Ziefle, 2007; Morris et al., 2005). A recent study also highlights that the same holds true in the customer context (Venkatesh et al., 2012). Consequently, following hypothesis was formulated: H1. The influence of PE on BI to use online shopping channels is moderated by gender and age, where the effect is stronger for younger men. 3.2 EE EE is the individual assessments of the degree to which technology utilization is free of effort. This aspect was formed by integrating the effort-oriented constructs from the informing models (e.g. ease of use and complexity). In the UTAUT, the PE operates as an extrinsic motivator, representing the outcomes of technology use, while EE manifests as an intrinsic motivator, referring to the process facilitating the respected outcomes (Karahanna et al., 2006). Studies have found that the perceived cognitive and/or behavioural effort needed to learn and utilize an information technology artefact directly influences BI, especially in the exploratory period of technology use (Venkatesh and

Figure 1. The proposed research model

APJML 28,2

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

284

Davis, 2000; Gefen, 2003). Although its significance in determining BI was found to diminish over time with the user’s increasing technology experience, recent evidence suggests that its predictive power remains significant in the voluntary technology acceptance cases (e.g. online shopping), even after the initial experience period (Lin and Nguyen, 2011). Venkatesh et al. (2012) examined and reported the significance of the relationship between EE and BI in the customer context. According to Hansen (2006), the major motivation for a customer to choose the online shopping channel is the opportunity to maximize energy convenience by reducing the physical and mental effort needed to complete a shopping task not available from alternative channels. Likewise, Lim and Dubinsky (2004) stated that the problems in accessing retailer sites, the long duration of page downloads, the tedious navigational structures, the cluttered site content, the slow transaction speeds and the complex purchase procedures inhibit online store patronage intentions. UTAUT hypothesized that the importance of EE in determining intentions varies with gender, age and experience. However, much of the moderating effect of gender was found to be in conjunction with age and experience in a customer setting, such that the effect is stronger for older women in their early stages of experience with the technology (Venkatesh et al., 2012). In accordance with the discussion and postulations above, following hypothesis was developed: H2. The influence of EE on BI is moderated by gender, age and experience, such that the effect is stronger for older women in the early stages of experience with the online shopping channel. 3.3 SI SI is an individual’s perception that others think he/she should use an information technology artefact (Venkatesh et al., 2003). It comprises subjective norms, social factors and image constructs identified as conceptually similar and reflects the normative pressure involving an individual’s persuasion of approval about technology use from his/her social group and motivation to comply with the shared social meaning of it among the group members (Venkatesh et al., 2012). UTAUT, inheriting the common premise of theory of reasoned action and TPB, considers technology adoption as a volitional behaviour (see Ajzen, 1991). Thus, it suggests the deliberative intent mechanism in which social norm acts as a direct determinant of intention and intention mediates its relation with adoption behaviour. However, the regarded impact of normative pressure on focal behaviour has been the subject of much debate. While some argue that SI has a direct effect on BI in mandatory settings due to compliance resulting from potential social rewards and punishments for engagement or no engagement in the technology use, others suggest that it has a direct effect on the personal beliefs of the technology in voluntary settings due to internalization and identification resulting from the personal desire to maintain a favourable image and gain social status within the reference group by using the technology (Venkatesh and Morris, 2000; Venkatesh and Davis, 2000). Some equivocal results were also reported in the customer contexts, including online shopping. Since online shopping is a voluntary decision, SI is expected to have an influence on intention due to the internalization and identification effects. However, most research has found that the social norm is an insignificant agent in determining customer intentions, both directly and indirectly over their technology beliefs (Sin Tan et al., 2013; Pascual-Miguel et al., 2015). According to Zhang et al. (2006), since online shopping is not a socially motivated behaviour performed within the virtual environment’s privacy, without the reliance of others, because of the available online

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

information and the encumbrance of seeking public approval, shoppers may feel less pressure to comply with social norms (and/or alter their belief structures for social status gains). On the other hand, when examined in other voluntary use settings (e.g. mobile stock trading, internet banking and Facebook usage), where the target technology provides the same privacy, comfort and confidence as online shopping, social norms were found to exert a significant influence on customer intentions (AbuShanab et al., 2010; Im et al., 2011; Tai and Ku, 2013; Lallmahomed et al., 2013). Previous studies on online shopping adoption have provided consistent results regarding the impact of social norms on customer intentions (Alwahaishi and Snásel, 2013; Lim and Dubinsky, 2005; Slade et al., 2015; Yang, 2010). The current work also connects SI to intentions, thus following the framework of UTAUT. UTAUT suggests that the mentioned effect of SI is moderated by gender, age and experience, such that it is strongest for older women, particularly in their early stages of experience with the technology (Venkatesh et al., 2012). These contingencies were grounded on theoretical bases, indicating that women are more affected by others’ opinions when forming an intention to use new technology (Venkatesh et al., 2000). They also tend to seek more advice when they feel inexperienced about how to use it during the initial adoption period and have a stronger desire for the affiliation needs with their increasing age (Morris and Venkatesh, 2000; Venkatesh and Morris, 2000). Therefore, following hypothesis was developed: H3. The SI on BI will be moderated by gender, age and experience, such that the effect will be stronger for older women in the early stages of their experiences with the online shopping channel. 3.4 Anxiety In the current study, anxiety (ANX) refers to the degree to which an individual temporally experiences fear, apprehension and aggression when considering use of, or actually using, an online shopping channel. It is a concept-specific emotional distress dependent on a customer’s interactions with virtual storefronts via the internet’s communication infrastructure (Celik, 2011). It represents a different, although not entirely distinct, construct from computer anxiety, because its task and damage components are highly relevant to online shopping behaviour (Pahnila et al., 2011). Like the other computing activities, online shopping is a task-oriented activity requiring customers to accomplish various communication tasks by using the hardware, software and protocols when interacting with virtual internet stores. Prior research has emphasized that personal inferences about failure in attaining communication tasks and desired shopping outcomes due to the operation hurdles (e.g. navigational problems, inconvenient checkout procedures and poor interface designs) make customers anxious and hesitant to approach online shopping (Celik, 2011; Vijayasarathy, 2004). A number of studies also suggest that customer concerns about the implications of online shopping, such as identity thefts, credit card frauds, privacy infringements, unauthorized account accesses, misleading product promotions and demanding dispute resolutions, heighten anxiety levels about transactions with virtual vendors (Forsythe et al., 2006; Littler and Melanthiou, 2006). As anxiety grows, individuals demonstrate higher levels of uncertainty avoidance and lower levels of proclivity to engage with a computer-mediated task (Thatcher and Perrewe’, 2002). Furthermore, ANX was found to increase the effort required for task accomplishment and impede the cognitive capacity needed to produce the desired task outcomes (Brown et al., 2004;

Customer online shopping anxiety 285

APJML 28,2

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

286

Saadé and Kira, 2006). Therefore, it is reasonable to expect that ANX has adverse effects on the BI, EE and PE constructs in the online shopping context. Moreover, age, gender and experience are associated with anxiety, such that older women in their early stages of experience with technology illustrate higher levels of anxiety towards the technologymediated tasks (Wang and Wang, 2008). Thus, following hypothesis was proposed: H4a. The influence of anxiety on BI will be moderated by gender, age and experience, such that the effect will be stronger for older women in the early stages of their experiences with the online shopping channel. H4b. The influence of anxiety on PE will be moderated by gender, age and experience, such that the effect will be stronger for older women in the early stages of their experiences with the online shopping channel. H4c. The influence of anxiety on EE will be moderated by gender, age and experience, such that the effect will be stronger for older women in the early stages of their experiences with the online shopping channel. 3.5 FC The FC construct is the degree to which an individual believes that the external support from both organizational and technical infrastructures is available when using an information technology artefact (Venkatesh et al., 2003). It encompasses three similar concepts from the previous TAM: perceived behavioural control, FC and compatibility (Venkatesh et al., 2012; Zhou, 2012). Ajzen (1991) presented the perceived behavioural control that involves the self-efficacy and resource facilitation concepts in TPB to capture the non-volitional aspects of focal behaviour. While self-efficacy represents the internal control factor, including personal judgment about the adequacy of knowledge, skill and will-power to accomplish a computer-mediated task, resource facilitation refers to the external control factor associated with the personal assessment about the availability of time, opportunity, compatible equipment and environmental support for the desired task performance (Taylor and Todd, 1995). Venkatesh et al. (2003) argued for the exclusion of self-efficacy from UTAUT because of its fading effect on BI when EE was present. They operationalized FC as containing certain elements (e.g. personal assessments about knowledge adequacy and assistance availability for technology use). Therefore, FC has been conceptualizing through the aggregation of internal and external support aspects in UTAUT. Online shopping has similar requirements for the presence of knowledge, resource and support empowering the customer to overcome certain constraints, such as the lack of tactile purchase experiences, the absence of direct personal contact with sales representatives, interactions with checkout interfaces instead of clerks and the need for shipment tracking (Song and Zahedi, 2005). Furthermore, the empirical results indicate that the FC influence on usage behaviour is moderated by experience and age, because with increased experience with technology, users know how to receive support from various sources to remove use resource constraints, and the cognitive/physical limitations associated with the age evokes the older users’ needs for assistance when using the respected technology (Venkatesh et al., 2008, 2012). Hence, following hypothesis is: H5. The influence of FC on usage will be moderated by age and experience, such that the effect will be stronger for older customers in the later stages of their experiences with the online shopping channel.

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

3.6 BI BI represents a transition between the individual- and social-related variables and the personal use of an information technology artefact in UTAUT (Venkatesh et al., 2003). It captures the motivation to enact the focal behaviour. Thus, UTAUT provides the sufficient presentation of belief-intention-behaviour relationships, because the influences of individual cognitive and evaluative responses towards the volitional use of technology are mediated by the information processing, underlying personal expectancies and SI (Armitage and Conner, 2001). However, BI does not mediate the effects of external variables on technology use totally so that the direct path between FC and technology use has been conceived in UTAUT (Venkatesh et al., 2003, 2012). BI is the most proximal determinant of technology use in previous UTAUT studies conducted in different settings, including diverse technologies such as mobile banking (Yu, 2012), tablet PCs (Anderson et al., 2006), decision support systems (Chang et al., 2007), online social support (Lin and Anol, 2008), collaboration technology (Brown et al., 2010), Facebook (Lallmahomed et al., 2013) and mobile internet (Venkatesh et al., 2012). There is also evidence suggesting that stronger customer intentions result in a higher determination to engage in online shopping. Therefore, following hypothesis was formulated: H6. BI will have a significant positive influence on usage. 4. Research methodology In order to test the hypothesized research model shown in Figure 1, a sample of Turkish online shoppers was surveyed with a self-administrated paper-and-pencil survey. A total of 80 voluntary students from a mid-size Western Turkish university were selected to serve as data collectors as part of their undergraduate course assignments. As an incentive, they received extra course credits in exchange for their services. The website of online mass merchant hepsiburada.com was used in this study to provide a familiar and uniform stimulus to the research participants. According to the latest report by Research AND Markets (2015), it is the largest domestic player on the booming Turkish B2C e-commerce market. The data collectors were instructed to survey customers regarding their shopping tasks on hepsiburada.com. They were provided specific instructions associated with the required demographic characteristics of research participants to ensure the sample’s representativeness of the Turkish online shopping population. No more than 60 per cent of those sampled could be of one gender, less than 25 years old, a university degree holder and from a monthly income level under $800. Finally, as seen in Table II, the participants were requested in the structured questionnaire to indicate their perceptions of the system aspects of hepsiburada.com, allowing them to perform the different stages of online purchase process (e.g. information search, price comparison, order placement, online payment and shipment tracking). In total, 506 questionnaires were collected during a period of six months; 23 were unusable, leaving 483 usable responses. One out of every 20 returning responses was selected for the authenticity check, in which participants were contacted via e-mail to verify that they had actually completed the survey. All of those contacted were verified. The sample was also split into early and late-respondent categories; t-tests found no difference in terms of their responses to the constructs (e.g. all t-values lower than 1.68 and p-values higher than 0.10). These results indicate that there is no risk of non-response bias in the sample.

Customer online shopping anxiety 287

APJML 28,2

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

288

The sample demographics are provided in Table I. The profiles were compared with those reported in the Stafford et al. (2006), Ergin and Akbay (2008) and Celik (2011). The sample was comprised of 55.3 per cent males and 44.7 per cent females (vs 53.7 and 46.3 per cent in the 2008 study). The average age was 27.8 years (vs the participants ranged in age 26-35 accounted for 54 per cent in the 2006 study). Among the participants, 59.7 per cent had a university degree (vs 49.3 per cent in the 2011 study). The largest number of participants, 79.7 per cent, earned less than $1,200 monthly (vs 76.3 per cent in the 2011 study). The average experience of participants with the targeted online store was 3.22 years. Finally, the participants made purchases from online stores approximately 2.72 times and spent an average of $267 on these purchases within the last six months (vs 54.5 per cent of participants who had used the store less than four times, and 38.8 per cent had spent more than $160 on these purchases within the same time period). The scales used to operationalize the constructs were taken from the original UTAUT study in which psychometric properties were validated across different time periods, settings (organizational and customer) and situations (voluntary and mandatory adoption of technology) by Venkatesh et al. (2003). The ANX scale measuring the task and damage anxieties was adopted from UTAUT’s four-item anxiety scale with modifications isomorphic to online shopping. Celik (2011) served as a Measure

Value

Gender

Female 1 Male (Continuous) 18 Junior high 1 school or less High school Vocational school Undergraduate school Graduate school o ₺10,000 1 ₺10,000₺20,000 ₺21,000₺30,000 ₺31,000₺40,000 W ₺40,000 (Continuous) 1

Age (years) Education

Annual income (TRY-₺) ($1≈₺2.23)

Table I. Descriptive statistics

Experience with “hepsiburada.com” (years) Number of online purchases made from (Continuous) “hepsiburada.com” within the last six months (Continuous) Amount spent in online shopping at “hepsiburada.com” within the last six months (TRY-₺) Notes: SD, standard deviation; na, not available

Min. Max. Mean SD Frequency % 2

na

na

55 5

27.8 na

7.1 na

5

na

na

216 267 na 19

44.7 55.3 na 3.9

86 90

17.8 18.6

260

53.9

28

5.8

140 154

29 31.9

91

18.8

60

12.3

13

3.22 2.19

35 na

7.3 na

1

10

2.72 1.99

na

na

15

5,000

596

na

na

802

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

basis for the regarded modifications (Table II). Actual shopping behaviour was measured through the self-reported online purchase frequency and amount spent for online purchases within the last six months. However, only the purchase frequency was used as the criterion variable during the model test phase. Age and experience was measured as continuous variables consistent with UTAUT. The final version was translated into Turkish and back-translated into English by three experienced bilingual translators to ensure consistency in item phrasing and the equivalency of their meanings. The draft questionnaire was subjected to critical review by 12 graduate students to detect confusing or ambiguous wording. A seven-point Likert scale, from 1 (strongly disagree) to 7 (strongly agree), was used to measure the constructs. The neutral point was not included in the scales to avoid courtesy bias from respondents. Previous research shows that the survey participants frequently chose the neutral option when asked to express their perceptions and attitudes (Nowlis et al., 2002). 5. Data analysis and results The partial least squares (PLS) method was used for the identification and evaluation of the research model. As a component-based modelling approach, PLS has been preferred to the covariance-based approaches (e.g. structural equation modelling (SEM) or multiple regressions) for incremental studies in the information systems research field that aim to build on an initial model by conceiving both new measures and structural paths (Hair et al., 2011). It also provides an advantage over SEM and regressions when working on research data under the conditions of non-normality, small sample size and multicollinearity (Compeau and Higgins, 1995). Finally, it allows for the simultaneous analysis of both the relationships among the latent variables (LVs) and the manifest variables (MVs), measuring their corresponding LVs (Haenlein and Kaplan, 2004). All MVs were plugged into their relevant LVs via a reflective measurement model in the current study. The two-step approach recommended by Anderson and Gerbing (1988) was adopted to implement the modelling strategy to assess the psychometric properties of the measurement model and estimate the path coefficients of the structural model with the influence of moderating variables afterwards. This sequence was followed to ensure the reliability and validity of the measurement model. SmartPLS 3 was used to perform the analyses (Ringle et al., 2014). The confirmatory factor analysis examined the loading patterns of MVs (measures) on their theoretically assigned LVs (constructs). The significance of item loadings was tested by using the bootstrap procedure with 500 subsamples that is consistent with prior UTAUT research (Venkatesh et al., 2003). The CFA results indicated that all loadings were significant ( p o 0.01) and above the conventional cut-off value of 0.70, recommended by Fornell and Larcker (1981). These results imply that more than 50 per cent (0.702) of variance in the observed MVs shared with their hypothetical LVs, thus serving as the baseline for internal consistency (Agarwal and Karahanna, 2000). As seen in Table II, the Cronbach’s α coefficients of all LVs was acceptable (Hair et al., 2006, p. 4), with the lowest being FC at 0.73. All other α coefficients were at least 0.83, providing additional support for the internal consistency. The reliability of measures was assessed by computing the composite scale reliability (CR) and the average variance extracted (AVE) scores for each construct. All CR values are greater than the suggested benchmark 0.70 and AVE scores (Table II) and compellingly exceed the common threshold value of 0.50 that provide additional evidence for the scale reliability (Gerbing and Anderson, 1988; Gefen, 2003). The convergent validity of measures

Customer online shopping anxiety 289

APJML 28,2 Constructs (LVs)/Measures (MVs)

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

290

Performance expectancy (PE) PE1: I find the system useful for shopping tasks PE2: using the system enables me to accomplish shopping tasks more quickly PE3: using the system increases my productivity in accomplishing shopping tasks PE4: if I use the system, I will increase my chances of getting better deals Effort expectancy (EE) EE1: my interaction with the system would be clear and understandable EE2: it would be easy for me to become skilful at using the system EE3: I find the system easy to use EE4: learning to operate the system is easy for me Social influence (SI) SI1: people who influence my behaviour think that I should use the system SI2: people important to me think that I should use the system SI3: people very close to me have been helpful in the use of the system SI4: in general, people very close to me supported the use of the system Facilitating conditions (FC) FC1: I have the resources necessary to use the system FC2: I have the knowledge necessary to use the system FC3: the system is not compatible with other systems I usea FC4: a specific person (or group) is available for assistance with the difficulties of using the system

Table II. Convergent validity, internal consistency and reliability

Anxiety (ANX) ANX1: I feel apprehensive about making purchase through the system ANX2: it scares me to think that I could lose my personnel and credit card information using the system for shopping

n ¼ 483 Factor Composite Mean SD loadings Cronbach’s α reliability AVE

5.66 1.40

0.86

5.58 1.37

0.88

5.37 1.37

0.88

5.49 1.43

0.79

5.38 1.34

0.79

5.54 1.40 5.90 1.30

0.82 0.86

6.01 1.34

0.80

4.25 1.74

0.86

4.21 1.70

0.88

4.23 1.86

0.88

4.40 1.75

0.79

6.21 1.15

0.74

6.17 1.33

0.71

5.55 1.45

0.79

5.07 1.71

0.72

2.93 1.85

0.85

3.33 2.08

0.89

0.87

0.91

0.73

0.83

0.89

0.67

0.88

0.92

0.73

0.74

0.83

0.55

0.89

0.92

0.75

(continued )

n ¼ 483 Factor Composite Mean SD loadings Cronbach’s α reliability AVE

Constructs (LVs)/Measures (MVs)

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

ANX3: I hesitate to use the system for fear of making costly mistakes I cannot correct 3.03 1.96 ANX4: the system is somewhat intimidating to me 2.58 1.79 Behavioural intention to use the system (BI) BI1: I intend to make purchase(s) through the system in the next six months BI2: I predict I would make purchase(s) through the system in the next six months BI3: I plan to make purchase(s) through the system in the next six months Notes: AVE, average variance extracted; SD,

0.88

Customer online shopping anxiety 291

0.84

4.98 1.50

0.93

5.02 1.48

0.96

0.95

0.97

0.90

5.01 1.50 0.96 standard deviation. athe reverse coded item

Table II.

was verified by examining their cross-loadings. Finally, the inter-construct correlations were compared with the square root of AVE scores to assess discriminant validity. Each square root of AVE (the italic elements on the diagonal) surpasses the intercorrelations of the construct with every other construct (the numbers off the diagonal), in support of discriminant validity (Table III) (Gefen and Straub, 2005). Based on the sound measurement model, the original UTAUT and two extended models tested the research hypotheses. A PLS analysis was carried out, further assessing and validating each direct and moderated mode of the original and extended models. Age and experience ranges were broken up into grouping (e.g. young vs old; inexperienced vs experienced). Gender, age and experience were dummy variables (0, 1). The dichotomization for the resulting information loss, reduced correlations among the collinear predictors and lower statistical power of linear models has been criticized, but is commonly and successfully used (MacCallum et al., 2002; Quinones and Kakabadse, 2015). Two separate analyses were conducted with continuous and Mean

SD

1

2

3

4

5

6

7

8

9

10

1. PE 5.52 1.19 0.85 2. EE 5.65 1.16 0.53 0.82 3. SI 4.26 1.52 0.25 0.19 0.85 4. FC 5.72 1.08 0.52 0.54 0.11 0.74 5. ANX 2.93 1.68 −0.24 −0.22 −0.04 −0.30 0.87 6. BI 5.00 1.42 0.27 0.22 0.27 0.18 −0.20 0.95 7. GDR D D 0.03 0.06 −0.01 0.11 −0.15 −0.01 na 8. AGE D D −0.03 −0.06 −0.06 0.01 −0.07 0.03 0.22 na 9. EXP D D 0.17 0.14 −0.13 0.22 −0.10 0.08 0.19 0.17 na 10. USE 2.72 1.99 0.09 0.08 0.17 0.08 −0.10 0.14 0.11 0.06 0.21 na Notes: “Use” has the single indicator, the purchase frequency; “D” indicates the dummy-coded variable; diagonal italic elements are square roots of AVEs and off-diagonal elements are interTable III. construct correlations Discriminant validity

APJML 28,2

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

292

dichotomized moderators, revealing that the cross-correlations between the moderators and focal variables were almost identical, further justifying the median dichotomization (Iacobucci et al., 2015; Krauth, 2003). The path estimates of BI and USE are reported in Tables IV and V. After computing the path estimates in the structural models, the bootstrap procedure, with 500 subsamples, was applied to assess the significance of the respected coefficients. The common recommended number of bootstrap subsamples is at least 500 to reliably examine the stability of path estimates (Chin, 1998; Venkatesh et al., 2003). Gender, age and experience were mean centred before forming the interaction terms in D+I models to reduce the multicollinearity (Venkatesh et al., 2003, 2012). The nomological validity of each model was evaluated using R 2 values of BI and USE constructs; R 2 should be at least 0.10 for these endogenous constructs to be considered to have adequate nomological validity (McKenna et al., 2013). The differential effect sizes of ANX and interaction terms were inspected through f 2 values calculated as follows; (R2Model When Interaction Terms Included−R 2 Model When Interaction Terms Excluded/1−R2Model When Interaction Terms Excluded) (Chin et al., 2003). The higher the f 2 value, the greater the effect of the interaction term (or exogenous construct), whereby values of f 2 were defined as small (0.02), medium (0.15) and large (0.35) (Chin, 1998). The global criterion of goodness of fit (GoF) (Tenenhaus et al., 2005) was used to evaluate the overall predictive performance of the models. It was estimated as the geometric mean of average communality (AVE) and average R2 of the endogenous constructs, whereby the baseline values of 0.1, 0.25 and 0.36 were regarded, respectively, as small, medium and large (Wetzels et al., 2009). The R2 for original UTAUT with direct-effects-only predicting BI was 0.12, and the parameter estimates for PE and SI were significant (Table IV). The direct impact of SI was significantly stronger in magnitude compared to that of PE. However, the main effects of EE and FC were not significant. PLS analysis was repeated after removing EE from the baseline model to test the notion that FC would be non-significant in predicting BI when EE was present in the model because its effect was largely captured by EE (Venkatesh, 2000; Venkatesh et al., 2003). However, there was no change in the R2 (adjusted R2 for both models ¼ 0.12). As for the path from FC to BI, the result was still non-significant ( β ¼ 0.08, p W 0.10). The effects of PE and SI on BI only increased marginally ( β ¼ 0.17, p o 0.05 and β ¼ 0.22, p o 0.01). Therefore, these two constructs were not overlapping, and FC was not a significant agent for BI. When PLS analysis was run once again with BI as the dependent variable and PE, EE and SI as the predictors, the results demonstrated that EE was not significantly associated with BI ( β ¼ 0.10, p W 0.10), whereas the remaining relationships involving PE-BI ( β ¼ 0.18, p W 0.05) and SI-BI ( β ¼ 0.21, p W 0.01) appeared to be almost identical to the relationships found in the previous analysis. The absence of EE’s significance in the baseline model was not in line with the past UTAUT research (Venkatesh et al., 2003). The R2 for this three-component model explaining the variation in BI was 0.12, barely exceeding the threshold for adequate nomological validity. A GoF value of 0.24, exceeding the suggested cut-off for moderate effect size, was obtained. As such, the original UTAUT with direct effects did not perform well for predicting BI in the online shopping context. Some of the most telling effects were observed when the direct-effects-only model predicting the actual use of online shopping system (USE) was tested. The results illustrated that the paths from BI and FC to USE were not significant ( β ¼ 0.06, pW0.10 and β ¼ 0.04, pW0.10, respectively). The analysis was repeated after removing Experience (EXP) to test whether the finding of no significant effects of these constructs

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

UTAUT (Original) D Only D+I (a) Dependent variable: behavioural intention (BI) 0.13 R2 0.12 Adjusted R2 Performance expectancy (PE) 0.15** Effort expectancy (EE) 0.09 Social influence (SI) 0.21*** Facilitating conditions (FC) 0.04 Anxiety (ANX) Gender (GDR) −0.02 Age (AGE) 0.05 Experience (EXP) 0.03 PE × GDR PE × AGE GDR × AGE PE × GDR × AGE EE × GDR EE × AGE EE × EXP GDR × AGE (included earlier) GDR × EXP AGE × EXP EE × GDR × AGE EE × GDR × EXP EE × AGE × EXP GDR × AGE × EXP EE × GDR × AGE × EXP SI × GDR SI × AGE SI × EXP GDR × AGE (included earlier) GDR × EXP (included earlier) AGE × EXP (included earlier) SI × GDR × AGE SI × GDR × EXP SI × AGE × EXP GDR × AGE × EXP (included earlier) SI × GDR × AGE × EXP ANX × GDR ANX × AGE ANX × EXP GDR × AGE (included earlier) GDR × EXP (included earlier) AGE × EXP (included earlier) ANX × GDR × AGE ANX × GDR × EXP ANX × AGE × EXP GDR × AGE × EXP (included earlier) ANX × GDR × AGE × EXP

0.74 0.72 0.14** 0.10 0.24*** 0.05 −0.04 0.06 0.03 0.05 −0.09 −0.01 −0.06 −0.12 0.15* −0.43*** Earlier 0.02 0.12 −0.12 0.14* −0.15* 0.09 0.46*** −0.11 −0.13 −0.14 Earlier Earlier Earlier 0.08 0.06 0.12 Earlier −0.44***

UTAUT (Extended I) D Only D+I 0.15 0.14 0.14** 0.07 0.21*** 0.01 −0.15*** −0.04 0.04 0.04

0.70 0.67 0.13** 0.09 0.22*** 0.04 −0.13** −0.05 0.06 0.05 0.06 −0.07 −0.01 −0.08 −0.11 0.16* −0.36*** Earlier 0.00 0.14 −0.12 0.15* −0.06 0.12 0.33*** −0.12 −0.13 −0.13 Earlier Earlier Earlier 0.08 0.05 0.12 Earlier −0.32*** 0.06 0.06 0.11 Earlier Earlier Earlier 0.06 −0.12 0.00 Earlier −0.20**

(continued )

Customer online shopping anxiety 293

Table IV. Structural model tests I

APJML 28,2

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

294

Table IV.

UTAUT (Original) D Only D+I

UTAUT (Extended I) D Only D+I

(b) Depended variable: use R2 0.09 0.19 0.09 0.28 Adjusted R2 0.08 0.17 0.08 0.26 BI 0.06 0.05 0.06 0.06 FC 0.04 0.05 0.04 0.03 AGE 0.09 0.08 0.09 0.09 EXP 0.13 0.12 0.13 0.14 FC × AGE −0.21** −0.21** FC × EXP 0.07 −0.10 AGE × EXP −0.12 −0.10 FC × AGE × EXP 0.21** 0.18** BI × AGE 0.04 BI × EXP 0.14 AGE × EXP (included earlier) Earlier BI × AGE × EXP 0.30*** Notes: D Only, direct-effects-only; D+I, direct effects and interaction terms; “Earlier” refers to the coefficients that have been listed in the table before, but need to be listed again for the complete assessment of the higher-order interaction terms; greyed out cells are not applicable for the specific column. *,**,***Significant at the 0.10, 0.05 and 0.01 levels, respectively

on USE could be attributed to its inclusion. However, there was no evidence in this initial analysis to suggest that FC ( β ¼ 0.08, pW0.10) and BI ( β ¼ 0.07, pW0.10) have direct effects on USE, even after controlling for the effect of EXP. These results were surprising and not in accordance with the findings of previous UTAUT research. The amount of variance in USE explained by FC, BI, AGE and EXP was 0.08, indicating very lowpredictive ability. GoF value could not be computed at this stage because the endogen construct, USE, has been measured formatively (Hair et al., 2014, p. 185). Interaction terms were included in the original UTAUT to test the moderating effects of AGE, Gender (GDR) and EXP on PE-BI, EE-BI and SI-BI relationships. The predictive power of the model increased from 0.12 to 0.72 after computing the regarded effects (Table V). The f 2 test of the difference between the direct-effects-only and the interaction terms added models was significant with very large effect size ( f 2 ¼ 0.68). A GoF value of 0.6, exceeding the threshold for high-global validity, also suggested that the model performs better in predicting BI than the direct-effects-only model. H1 proposed that the effect of PE on BI would be moderated by gender and age. As reported in Table IV, the three-way interaction effect of PE × GDR × AGE on BI was not significant (B ¼ −0.06, p W 0.10). However, the positive direct influence of PE on BI remained significant (β ¼ 0.14, p o 0.05). The two-way interaction effects of EE × AGE ( β ¼ 0.15, p W 0.10) and EE × EXP ( β ¼ −0.43, p o 0.01) on BI showed that EE’s role was more important among the older and less-experienced customers. Furthermore, the effect of EE was in the form of a four-way interaction (EE × GDR × AGE × EXP), suggesting it was strongest for older women with less experience with shopping channels ( β ¼ 0.46, p W 0.01). Thus, H2 was confirmed. H3, considering GDR, AGE and EXP to moderate the effect of SI on BI, such that it was more important among the older women in their early stages of experience with the shopping channel, was partially accepted. Contrary to our expectation, the relationship involving SI-BI was strongest in less-experienced younger women by a four-way

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

UTAUT (Extended II) (a) Depended variable: PE R2 Adjusted R2 ANX GDR AGE EXP ANX × GDR ANX × AGE GDR × AGE ANX × GDR × AGE (b) Depended variable: EE R2 Adjusted R2 ANX GDR AGE EXP ANX × GDR ANX × AGE GDR × AGE ANX × GDR × AGE

D Only

D+I

0.09 0.08 −0.24*** −0.01 −0.03 0.18***

0.10 0.09 −0.24*** −0.04 −0.03 0.17*** 0.05 −0.10 0.03 0.07

0.08 0.07 −0.24*** 0.03 −0.05 0.14***

0.09 0.08 −0.24*** 0.06 −0.01 0.13*** 0.04 −0.09 −0.07 0.04

(c) Depended variable: BI (with the direct effects of ANX on PE and EE) R2 0.15 Adjusted R2 0.14 PE 0.13** EE 0.07 SI 0.21*** FC 0.02 ANX −0.15*** GDR −0.04 AGE 0.04 EXP 0.04 PE × GDR PE × AGE GDR × AGE PE × GDR × AGE EE × GDR EE × AGE EE × EXP GDR × AGE (included earlier) GDR × EXP AGE × EXP EE × GDR × AGE EE × GDR × EXP EE × AGE × EXP GDR × AGE × EXP EE × GDR × AGE × EXP

0.70 0.67 0.12** 0.10 0.22** 0.04 −0.13** −0.05 0.06 0.05 0.06 −0.08 −0.01 −0.08 −0.11 0.16* −0.35*** Earlier −0.01 0.14 −0.12 0.15* −0.07 0.12 0.33***

(continued )

Customer online shopping anxiety 295

Table V. Structural model tests II

APJML 28,2

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

296

Table V.

UTAUT (Extended II) D Only

D+I

SI × GDR −0.12 SI × AGE −0.13 SI × EXP −0.13 GDR × AGE (included earlier) Earlier GDR × EXP (included earlier) Earlier AGE × EXP (included earlier) Earlier SI × GDR × AGE 0.08 SI × GDR × EXP 0.05 SI × AGE × EXP 0.12 GDR × AGE × EXP (included earlier) Earlier SI × GDR × AGE × EXP −0.33*** ANX × GDR 0.06 ANX × AGE 0.06 ANX × EXP 0.11 GDR × AGE (included earlier) Earlier GDR × EXP (included earlier) Earlier AGE × EXP (included earlier) Earlier ANX × GDR × AGE 0.06 ANX × GDR × EXP −0.12 ANX × AGE × EXP 0.00 GDR × AGE × EXP (included earlier) Earlier ANX × GDR × AGE × EXP −0.20** Notes: D Only, direct-effects-only; D+I, direct effects and interaction terms; “Earlier” refers to the coefficients that have been listed in the table before, but need to be listed again for the complete assessment of the higher-order interaction terms; greyed out cells are not applicable for the specific column. *,**,***Significant at the 0.10, 0.05 and 0.01 levels, respectively

interaction effect of SI × GDR × AGE × EXP on BI (β ¼ −0.44, p o0.01). Finally, the R2 value for USE was 0.17, an improvement in the result of the direct-effects-only model of 0.08. The f 2 test result indicated that the improvement in the model’s predictive power was small but significant ( f 2 ¼ 0.10). The effect of FC on USE was found to vary across AGE and EXP, such that it was more salient to younger customers less experienced with the shopping channel (β ¼ 0.21, p o 0.05). Therefore, H5 was rejected and H6 was not supported ( β ¼ 0.05, p W 0.10). The first extended model differs from original UTAUT by including an additional direct path from anxiety (ANX) to BI. When this extended model with direct-effectsonly was tested, the paths from PE, SI and ANX to BI were significant ( β ¼ 0.14, p o 0.05, β ¼ 0.21, p o 0.01 and β ¼ −0.15, p o 0.01, respectively). The significant negative impact of ANX on BI illustrated that it was as important as PE and SI in accounting for BI to use the online shopping channel. The impacts of EE ( β ¼ 0.07, p W 0.10) and FC ( β ¼ 0.01, p W 0.10) on BI (Table IV) were insignificant. The results regarding the paths from FC and BI to USE were not significant. There were no major increments in the R2 values for BI (0.14) and USE (0.09) as compared to the original UTAUT with the direct-effects-only model (0.12 and 0.09, respectively). The GoF value of 0.27 indicated that the model for BI does not perfectly fit the observed data. Comparison of the direct-effects-only and interaction terms added models revealed that the moderating effects had a significant impact on R2 increasing it (0.15 in direct effects vs 0.70 in interaction terms). The f 2 was found to be 0.65, suggesting

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

that the regarded improvement in R2 was highly significant. Furthermore, the model has a GoF value of 0.59 that could be considered highly satisfactory. However, its predictive power seemed to slightly decrease over the original UTAUT with interaction terms due to the negative effects of ANX on the other focal variables. As theorized, the effect of ANX on BI varied across GDR, AGE and EXP (ANX × GDR × AGE × EXP); it was more salient to older women in their early stages of individual experience with online shopping channel ( β ¼ −0.20, p o 0.05). Therefore, H4a was confirmed based on this evidence. The results demonstrated that the effects of EE and SI were moderated by GDR, AGE and EXP via the four-way interactions that were in the same direction and magnitude as the interaction effects. Likewise, the effect of FC on USE was found to vary across AGE and EXP. The modified model provided slightly weaker paths regarding these interaction terms when compared with the original UTAUT. The interaction effects of AGE and EXP on BI-USE link were detected and tested. The results obtained in Table IV show that AGE and EXP jointly affect the relationship between BI and USE ( β ¼ 0.30, p o 0.01). After constraint for AGE and EXP, the R2 for USE generated by BI and FC enhanced up to 0.26. The f2 value of 0.20 indicated that this improvement was moderate in size, but still significant. For the second extended model, the possibility that ANX could be an endogenous construct influencing PE, EE and BI directly was examined. Table V shows the path coefficients and R2 for the direct-effects-only and the interaction terms added models. The amount of variances in BI explained by the second extended model with and without interaction terms were similar to the first extended model with and without interaction terms. There were no changes in the properties of the causal paths as compared to that of the first extended model’s modes. On the other hand, ANX was found to exert significant negative influences on both PE and EE ( β ¼ −0.24, p o 0.01 and β ¼ −0.24, p o 0.01, respectively). An interesting pattern emerged in the support for the negative direct effect of ANX on BI ( β ¼ −0.15, po0.01), even when its direct effects on PE and EE were present. There was no evidence in the analysis to suggest that ANX would have any significant indirect effect on BI via PE or EE. The effect of ANX on BI was via a four-way interaction once again, providing additional support for H4b. Regarding the moderating effects of GDR, AGE and EXP on the ANX-PE and ANX-EE relationships considered in H4b and H4c, the results failed to confirm the existence of significant variations in the relationships across the moderators. Therefore, H4b and H4c were rejected. As seen in Table V, EXP was treated as a predictor, rather than a moderator, due to its significant effects on both criterion variables (PE and EE) during the regarded analyses. 6. Discussion The present study developed a new theory by grounding anxiety in the UTAUT model and applying it to the online shopping context. Therefore, it provides new interrelationships among anxiety, PE and EE constructs associated with the intentions to use online shopping channels. Another contribution of this study is the addition of anxiety, an affect component, into the mainly cognition- and behaviour-based UTAUT framework to remedy an omitted variable problem and enhance its generalizability to the consumer environment. Moreover, there is an ongoing debate about the causal impact of anxiety in information systems research field. The role of anxiety in online shopping is now shown to exert inhibiting effects on consumer intentions, PE and EE with consideration for the intervening variables (gender, age and experience) and the influence of anxiety on consumer beliefs and intentions.

Customer online shopping anxiety 297

APJML 28,2

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

298

The study provides cross-cultural validation of the original and extended UTAUT models with direct effects and interaction terms by examining a sample of online shoppers from Turkey. The significant negative influence of anxiety on consumer intentions was observed in all cases during the data analysis. The regarded effect was found to be moderated by gender, age and experience; it was more significant for older women having less experience with online shopping channels. There are a number of possible explanations for these results. One is that older people face major difficulties in processing complex information related to a new computing task, allocating the cognitive resources for its performance and acquiring the required computing skills for its completion (Arning and Ziefle, 2007; Venkatesh et al., 2003, 2012). Previous studies have repeatedly observed that females tend to feel less comfortable with computing because they react more somatically to the emotion and uncertainty and thus are more reluctant to develop adequate computing self-efficiency (Beyer, 2008; He and Freeman, 2010; King et al., 2002). In addition, personal experience with computers is inversely related to computer anxiety; increasing experience significantly contributes to the development of self-efficacy perception about computing (Brown et al., 2004; Igbaria and Chakrabarti, 1990; Ong and Lai, 2006), thus leading to a decrease in computer anxiety. The same relationship holds true in the online shopping context. As experience decreases, an individual perceives the online shopping as particularly challenging and thus experiences anxious feelings. The results also illustrated that anxiety had significant negative impacts on both PE and EE constructs, in addition to its direct effect on consumer intentions. More anxious individuals perceive computers as less useful (Thorpe and Brosnan, 2007). This is also the case for online shopping channels. Anxious individuals try to allocate and direct additional processing resources to eliminate the negative effects of computer anxiety and improve their performances on computing tasks (Fakun, 2009; Mead and Drasgow, 1993; Smith and Caputi, 2001). In this study, unexpectedly, the negative effects of anxiety on both performance and EE constructs did not vary across gender, age and experience groups. Anxiety makes both genders equally focus on the magnitude of effort for using shopping channels and processes to achieve shopping tasks. In terms of age, Kim and Forsythe (2010) pointed out that younger individuals have as much technology-related stress as older ones, because they are increasingly exposed to and forced to learn new technologies every day. Some studies failed to detect the difference in computer anxiety between experienced and novice users, suggesting that the performance deficit and the excessive process expectancy caused by anxious feelings during a computing task is independent of the prior level of individuals’ computing experiences (Mahar et al., 1997; McInerney et al., 1994). Moreover, Beckers and Schmidt (2003) noted that the positive experience, not the amount of experience, relieves the individuals’ anxious feelings when they engage in a computing task. Thus, the conceptualization of experience as an engagement amount with online shopping channel could result in the discussed finding about the insignificant moderating effect of this construct. PE was positively related to consumer intentions in all models tested in this study. However, regarding the moderating effects of gender and age, the results failed to confirm the significant variations between gender and age groups. Consumers evaluated the performance of online shopping in terms of the associated benefits (i.e. its facilitation of efficiency, effectiveness and productivity in accomplishment of shopping tasks) and costs (i.e. cognitive, behavioural or financial investments made by them to utilize online channel for their shopping tasks). If the cost was lower or the benefit was higher, the utilitarian

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

value of the technology (its perceived performance) would be greater. However, the perceived utilitarian value of online shopping seems to be almost counterbalanced by its costs in this study. Contrary to expectations, EE was not found to be a significant determinant of consumer intentions to use online shopping channels. SI was found to be the strongest determinant of BI to use the shopping channel. This result contradicts Venkatesh and Davis (2000) by showing that the compliance mechanism causing SI has a direct effect on intentions works, not only in mandatory contexts, but also in voluntary contexts. The strength of this relationship appeared to vary with gender, age and experience, such that it was more significant for younger women less experienced with shopping online. Younger women could be more motivated to comply with others’ expectations and satisfy their affiliation needs, especially in collectivist and feminine cultures, like the national culture of Turkey. This result is in accordance with that of the original UTAUT study (Venkatesh et al., 2003). The effect was moderated by gender and age, such that it is more significant for younger customers in their early stages of experience with online shopping channels. It contradicts the results of Venkatesh et al. (2012) and Morris et al. (2005), who indicated that the effect is more salient in older consumers who are novice shopping medium users. Finally, the effect of intentions on usage was found to be significant only when examined with the moderating effects of age and experience. It was more pronounced for older and more experienced customers. Contrary to the conclusions of previous studies (Kim et al., 2005; Morris et al., 2005), the strength of the relationship between intentions and use could increase as experience with the shopping medium increases, because its routine use becomes habitually reinforced more by the associated cues possessed during the customers’ exposures to this technology. 7. Limitations This study has some limitations to its application to real-world situations. First, the study sample might not represent the general population of online shoppers, because most participants are customers of a particular online store. The study was also limited to examining a wide array of online stores (e.g. specialty shops, general merchandisers, online retailers, digital auction sites or consignment shops). A multi-group comparison between the store types could further improve the generalizability of study findings about the role of anxiety in online shopping. The conceptualization of anxiety as a unitary construct is another source of limitation. As discussed previously, damage and task anxiety components emanate from different sources. While the personal worry about the failure of completing the online shopping task causes task anxiety, damage anxiety stems from personal perceptions of the possible risks and dangers of online shopping fostered by the intangible and uncertain nature of the shopping environment. Future studies could examine investigate the components of anxiety and their effects on customer perceptions and intentions separately. Finally, it is recommended that researchers and practitioners be aware of cultural differences when applying the findings of this study to consumers in other countries. Previous research has shown that differences in cultural dimensions across nations have significant effects on individual perceptions and adoptions of online shopping (Ashraf et al., 2014). Hence, examining the effects of cross-cultural differences (i.e. possible comparisons between the cultures high and low in uncertainty avoidance or power distance) on the causal relationships in the extended models is another recommended future avenue.

Customer online shopping anxiety 299

APJML 28,2

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

300

8. Implications for practice The study findings offer important implications for marketers, particularly practitioners and e-tailers in Turkey. The study results support the findings by Powell (2013), demonstrating that anxiety is critical in negatively influencing customer perceptions of and intentions towards shopping online. Previous research illustrates that anxiety stems from customer concerns about either the failure in attaining desired shopping outcomes due to operation hurdles (e.g. navigational problems, inconvenient checkout procedures and poor interface designs) or the implications of online shopping (e.g. identity thefts, credit card frauds, privacy infringements, unauthorized account accesses, misleading product promotions and demanding dispute resolutions) (Celik, 2011). Despite the drastic advancements in electronic commerce infrastructures and online security technologies, further improvements are needed to reduce customer anxiety regarding online shopping. Many online stores now have highly interactive interface designs, very responsive navigational tools, convenient checkout processes, detailed privacy statements, trusted third-party security seals, improved authentication procedures, neatly designed return policies and effectively managed warehouse operations aimed at enhancing customer confidence. However, the information about products and services provided by site management is still pretty much technical, heavily worded and somehow unclassified. The absence of the physical access to and evaluation of products in online store environments also increases risk perceptions and the anxiety levels of online shoppers. Various multimedia elements, such as close-up photographs, videos and animations, have been embedded in many shopping sites to provide customers with an overall experience and evaluation of the products. However, product photos are not visually appealing and adequately detailed in many cases. Likewise, the uploaded video content often consists of reproduced versions of TV commercials or new product launch videos. In developing visual product presentations to enable customers to experience products virtually, practitioners must recognize the importance of artistic proficiency and proper creativity, providing a multifaceted view and deeper understanding of the products (Demangeot and Broderick, 2010; Jiang and Benbasat, 2005). Since online apparel shopping is common practice among female customers, new technologies, such as 3D presentations or virtual try-on tools, could be very useful tools for the reduction of repurchase evaluation stresses and risk perceptions, due to product appearance and fit issues. Previous studies have shown that technologies facilitating the customer experiment with various clothing items may also encourage them to spend more time and purchase more items on the site (Kim and Forsythe, 2008). Many online stores try to minimize customer concerns about transaction improprieties, such as credit card fraud or fraudulent sales resulting in financial loss, by offering payment options through various trusted intermediaries that guarantees a refund if customers are not fully satisfied. PayPal is one of these institutional intermediaries integrated into many online shopping sites. However, customers still perceive it as a distrustful foreign subsidiary, because of its unclear, ineffective and unattended endorsement, especially in Turkey. Moreover, inexperienced older customers tend to be more prone to urban myths about fraudulent activities in the online environment and are prejudiced against online transactions, often with no basis in reality, only exposure to adverse media reports about these activities. Therefore, online marketers must reach out through different communication platforms, including social media, to demonstrate the security features of their stores and the transaction safety provisions of these intermediaries.

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

Finally, the customer anticipation of bearing return shipping costs for failed or defective products also increases their risk perceptions and anxiety levels. Previous research illustrates that customers paying for their own product returns generally decrease their repurchases from the same online store (Bower and Maxham, 2012; Petersen and Kumar, 2009). Therefore, it could be beneficial to online retailers to offer a free return shipping policy if they detect its positive effects on customer relationships and purchases when they consult their own customer data. References AbuShanab, E., Pearson, J.M. and Setterstrom, A.J. (2010), “Internet banking and customers’ acceptance in Jordan: the unified model’s perspective”, Communications of the Association for Information Systems, Vol. 26 No. 1, pp. 493-524. Agarwal, R. and Karahanna, E. (2000), “Time flies when you’re having fun: cognitive absorption and beliefs about information technology usage”, MIS Quarterly, Vol. 24 No. 4, pp. 665-694. Ajzen, I. (1991), “The theory of planned behavior”, Organizational Behavior & Human Decision Processes, Vol. 50 No. 2, pp. 179-211. Alwahaishi, S. and Snásel, V. (2013), “Acceptance and use of information and communications technology: a UTAUT and flow based theoretical model”, Journal of Technology Management and Innovation, Vol. 8 No. 2, pp. 61-73. Anderson, J.C. and Gerbing, D.W. (1988), “Structural equation modelling in practice: a review and recommended two-step approach”, Psychological Bulletin, Vol. 103 No. 3, pp. 411-423. Anderson, J.E., Schwager, P.H. and Kerns, R.L. (2006), “The drivers for acceptance of tablet PCs by faculty in a college of business”, Journal of Information Systems Education, Vol. 17 No. 4, pp. 429-440. Armitage, C.J. and Conner, M. (2001), “Efficacy of the theory of planned behaviour: a meta‐ analytic review”, British Journal of Social Psychology, Vol. 40 No. 4, pp. 471-499. Arning, K. and Ziefle, M. (2007), “Understanding age differences in PDA acceptance and performance”, Computers in Human Behavior, Vol. 23 No. 6, pp. 2904-2927. Ashraf, A.R., Thongpapanl, N.T. and Auh, S. (2014), “The application of the technology acceptance model under different cultural contexts: the case of online shopping adoption”, Journal of International Marketing, Vol. 22 No. 3, pp. 68-93. Baron, S., Patterson, A. and Harris, K. (2006), “Beyond technology acceptance: understanding consumer practice”, International Journal of Service Industry Management, Vol. 17 No. 2, pp. 111-135. Beckers, J.J. and Schmidt, H.G. (2003), “Computer experience and computer anxiety”, Computers in Human Behavior, Vol. 19 No. 6, pp. 785-797. Beyer, S. (2008), “Gender differences and intra-gender differences amongst management information systems students”, Journal of Information Systems Education, Vol. 19 No. 3, pp. 301-310. Bower, A.B. and Maxham, J.G. (2012), “Return shipping policies of online retailers: normative assumptions and the long-term consequences of fee and free returns”, Journal of Marketing, Vol. 76 No. 4, pp. 110-124. Brown, S.A., Dennis, A.R. and Venkatesh, V. (2010), “Predicting collaboration technology use: integrating technology adoption and collaboration research”, Journal of Management Information Systems, Vol. 27 No. 2, pp. 9-54. Brown, S.A., Fuller, R.M. and Vician, C. (2004), “’Who’s afraid of the virtual world? Anxiety and computer-mediated communication”, Journal of the Association for Information Systems, Vol. 5 No. 2, pp. 79-107.

Customer online shopping anxiety 301

APJML 28,2

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

302

Busch, T. (1995), “Gender differences in self-efficacy and attitudes toward computers”, Journal of Educational Computing Research, Vol. 12 No. 2, pp. 147-158. Celik, H. (2011), “Influence of social norms, perceived playfulness and online shopping anxiety on customers’ adoption of online retail shopping: an empirical study in the Turkish context”, International Journal of Retail & Distribution Management, Vol. 39 No. 6, pp. 390-413. Chang, I.C., Hwang, H.G., Hung, W.F. and Li, Y.C. (2007), “Physicians’ acceptance of pharmacokinetics-based clinical decision support systems”, Expert Systems with Applications, Vol. 33 No. 2, pp. 296-303. Chin, W.W. (1998), “The partial least squares approach to structural equation modelling”, in Marcoulides, G.A. (Ed.), Modern Methods for Business Research, Lawrance Erlbaum Associates, Hillsdale, NJ, pp. 295-336. Chin, W.W., Marcolin, B. and Newsted, P. (2003), “A partial least squares latent variable modelling approach for measuring interaction effects: results from a Monte Carlo simulation study and an electronic-mail emotional adoption study”, Information Systems Research, Vol. 14 No. 2, pp. 189-217. Chiu, C.M. and Wang, E.T. (2008), “Understanding web-based learning continuance intention: the role of subjective task value”, Information & Management, Vol. 45 No. 3, pp. 194-201. Cole, J.I., Schramm, P., Zhou, L. and Reyes-Sepulveda, E. (2013), The World Internet Project: International Report, (5th ed.), Center for the Digital Future, Annenberg School of Comminication and Journalism, University of Southern California, available at: www. digitalcenter.org/wp-content/uploads/2013/12/2013worldinternetreport.pdf (accessed 23 December 2014). Compeau, D., Higgins, C.A. and Huff, S. (1999), “Social cognitive theory and individual reactions to computing technology: a longitudinal study”, MIS Quarterly, Vol. 23 No. 2, pp. 145-158. Compeau, D.R. and Higgins, C.A. (1995), “Computer self-efficacy: development of a measure initial test”, MIS Quarterly, Vol. 19 No. 2, pp. 189-211. Demangeot, C. and Broderick, A.J. (2010), “Exploration and its manifestations in the context of online shopping”, Journal of Marketing Management, Vol. 26 Nos 13/14, pp. 1256-1278. Eastin, M.S. (2002), “Diffusion of e-commerce: an analysis of the adoption of four ecommerce activities”, Telematics & Informatics, Vol. 19 No. 3, pp. 251-267. eMarketer (2014), “Retail sales worldwide will top $22 trillion this year”, available at: www. emarketer.com/Article/Retail-Sales-Worldwide-Will-Top-22-Trillion-This-Year/1011765 (accessed 15 January 2015). Ergin, E.A. and Akbay, H.O. (2008), “An empirical investigation of Turkish consumers’ online shopping patterns”, Journal of Global Business &Technology, Vol. 4 No. 2, pp. 54-59. Eysenck, M.W. and Calvo, M.G. (1992), “Anxiety and performance: the processing efficiency theory”, Cognition & Emotion, Vol. 6 No. 6, pp. 409-434. Fakun, D. (2009), “How to mitigate the significant negative influence of computer anxiety on ease of use perceptions”, Behaviour & Information Technology, Vol. 28 No. 3, pp. 223-238. Featherman, M. and Pavlou, P.A. (2003), “Predicting e-services adoption: a perceived risk facets perspective”, International Journal of Human-Computer Studies, Vol. 59 No. 4, pp. 451-474. Fillion, G., Braham, H. and Ekionea, J.P.B. (2012), “Testing UTAUT on the use of ERP systems by middle managers and end-users of medium-to large-sized Canadian enterprises”, Academy of Information & Management Sciences Journal, Vol. 15 No. 2, pp. 1-28. Fornell, C. and Larcker, D.F. (1981), “Evaluating structural equation models with unobservable variables and measurement error”, Journal of Marketing Research, Vol. 18 No. 1, pp. 39-50.

Forsythe, S., Liu, C., Shannon, D. and Gardner, L.C. (2006), “Development of a scale to measure the perceived benefits and risks of online shopping”, Journal of Interactive Marketing, Vol. 20 No. 2, pp. 55-75. Foucault, B.E. and Scheufele, D.A. (2002), “Web vs campus store? Why students buy textbooks online”, Journal of Consumer Marketing, Vol. 19 No. 5, pp. 409-423.

Customer online shopping anxiety

Gefen, D. (2003), “TAM or just plain habit: a look at experienced online shoppers”, Journal of End User Computing, Vol. 15 No. 3, pp. 1-13.

303

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

Gefen, D. and Straub, D. (2005), “A practical guide to factorial validity using PLS-graph: tutorial and annoted example”, Communications of the Association for Information Systems, Vol. 16 No. 1, pp. 91-109. Gerbing, D.W. and Anderson, J.C. (1988), “An updated paradigm for scale development incorporating unidimensionality and its assessment”, Journal of Marketing Research, Vol. 25 No. 2, pp. 186-192. Gilbert, D., Lee-Kelley, L. and Barton, M. (2003), “Technophobia, gender influences and consumer decision-making for technology related products”, European Journal of Innovation Management, Vol. 6 No. 4, pp. 253-263. Hackbarth, G., Grover, V. and Yi, M.Y. (2003), “Computer playfulness and anxiety: positive and negative mediators of the system experience effect on perceived ease of use”, Information & Management, Vol. 40 No. 3, pp. 221-232. Haenlein, M. and Kaplan, A.M. (2004), “A beginner’s guide to partial least squares analysis”, Understanding Statistics, Vol. 3 No. 4, pp. 283-297. Hair, J.F., Ringle, C.M. and Sarstedt, M. (2011), “PLS-SEM: indeed a silver bullet”, Journal of Marketing Theory and Practice, Vol. 19 No. 2, pp. 139-152. Hair, J.F., Hult, G.T.M., Ringle, C.M. and Sarstedt, M. (2014), A Primer on Partial Least Squares Structural Equation Modelling (PLS-SEM), Sage Publications, Inc., Thousand Oaks, CA. Hair, J.F., Black, W.C., Babin, B.J., Anderson, R.E. and Tatham, R.L. (2006), Multivariate Data Analysis, (Vol. 6), Pearson Prentice Hall, Upper Saddle River, NJ. Hansen, T. (2005), “Consumer adoption of online grocery buying: a discriminant analysis”, International Journal of Retail and Distribution Management, Vol. 33 No. 2, pp. 101-121. Hansen, T. (2006), “Determinants of consumers’ repeat online buying of groceries”, International Review of Retail Distribution & Consumer Research, Vol. 16 No. 1, pp. 93-114. He, J. and Freeman, L. (2010), “Are men more technology-oriented than women? The role of gender on the development of general computer self-efficacy of college students”, Journal of Information Systems Education, Vol. 21 No. 2, pp. 203-212. Iacobucci, D., Posavac, S.S., Kardes, F.R., Schneider, M.J. and Popovich, D.L. (2015), “Toward a more nuanced understanding of the statistical properties of a median split”, Journal of Consumer Psychology, Vol. 25 No. 4, pp. 652-665, available at: www.sciencedirect.com/ science/article/pii/S1057740814001107 (accessed 22 December 2014). Igbaria, M. and Chakrabarti, A. (1990), “Computer anxiety and attitudes towards microcomputer use”, Behaviour & Information Technology, Vol. 9 No. 3, pp. 229-241. Igbaria, M. and Iivari, J. (1995), “The effects of self-efficacy on computer usage”, Omega, Vol. 23 No. 6, pp. 587-605. Im, I., Hong, S. and Kang, M.S. (2011), “An international comparison of technology adoption testing the UTAUT model”, Information & Management, Vol. 48 No. 1, pp. 1-8. Jiang, Z. and Benbasat, I. (2005), “Virtual product experience: effects of visual and functional control of products on perceived diagnosticity and flow in electronic shopping”, Journal of Management Information Systems, Vol. 21 No. 3, pp. 111-147.

APJML 28,2

304

Karahanna, E., Agarwal, R. and Angst, C.M. (2006), “Reconceptualizing compatibility beliefs in technology acceptance research”, MIS Quarterly, Vol. 30 No. 4, pp. 781-804. Karsten, R. and Schmidt, D. (2008), “Business student computer self-efficacy: ten years later”, Journal of Information Systems Education, Vol. 19 No. 4, pp. 445-451. Keen, C., Wetzels, M., de Ruyter, K. and Feinberg, R. (2004), “E‐tailers versus retailers: which factors determine consumer preferences”, Journal of Business Research, Vol. 57 No. 7, pp. 685-695. Kim, J. and Forsythe, S. (2008), “Adoption of virtual try-on technology for online apparel shopping”, Journal of Interactive Marketing, Vol. 22 No. 2, pp. 45-59.

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

Kim, J. and Forsythe, S. (2010), “Factors affecting adoption of product virtualization technology for online consumer electronics shopping”, International Journal of Retail & Distribution Management, Vol. 38 No. 3, pp. 190-204. Kim, S.S., Malhotra, N.K. and Narasimhan, S. (2005), “Two competing perspectives on automatic use: a theoretical and empirical comparison”, Information Systems Research, Vol. 16 No. 4, pp. 418-432. King, J., Bond, T. and Blandfort, S. (2002), “An investigation of computer anxiety by gender and grade”, Computers in Human Behaviour, Vol. 18 No. 1, pp. 69-84. Klopping, I.M. and McKinney, E. (2004), “Extending the technology acceptance model and the task-technology fit model to consumer e-commerce”, Information Technology, Learning & Performance Journal, Vol. 22 No. 1, pp. 35-48. Krauth, J. (2003), “Median dichotomization in CFA: is that allowed?”, Psychology Science, Vol. 45 No. 2, pp. 324-329. Lallmahomed, M.Z.I., Ab.Rahim, N.Z., Ibrahim, R. and Abdul Rahman, A. (2013), “Predicting different conceptualizations of system use: acceptance in hedonic volitional context (Facebook)”, Computers in Human Behavior, Vol. 29 No. 6, pp. 2776-2787. Lim, H. and Dubinsky, A.J. (2004), “Consumers’ perceptions of e-shopping characteristics: an expectancy-value approach”, Journal of Services Marketing, Vol. 18 No. 7, pp. 500-513. Lim, H. and Dubinsky, A.J. (2005), “The theory of planned behavior in e‐commerce: making a case for interdependencies between salient beliefs”, Psychology & Marketing, Vol. 22 No. 10, pp. 833-855. Lin, C.P. and Anol, B. (2008), “Learning online social support: an investigation of network information technology based on UTAUT”, CyberPsychology & Behavior, Vol. 11 No. 3, pp. 268-272. Lin, C.P. and Nguyen, C. (2011), “Exploring e-payment adoption in Vietnam and Taiwan”, Journal of Computer Information Systems, Vol. 51 No. 4, pp. 41-52. Littler, D. and Melanthiou, D. (2006), “Consumer perceptions of risk and uncertainty and the implications for behaviour towards innovative retail services: the case of internet banking”, Journal of Retailing & Consumer Services, Vol. 13 No. 6, pp. 431-443. McInerney, V., McInerney, D.M. and Sinclair, K.E. (1994), “Student teachers, computer anxiety and computer experience”, Journal of Educational Computing Research, Vol. 11 No. 1, pp. 27-50. McKenna, B., Tuunanen, T. and Gardner, L. (2013), “Consumers’ adoption of information services”, Information & Management, Vol. 50 No. 5, pp. 248-257. MacCallum, R.C., Zhang, S., Preacher, K.J. and Rucker, D.D. (2002), “The practice of dichotomization of quantitative variables”, Psychological Methods, Vol. 7 No. 1, pp. 19-40. Mahar, D., Henderson, R. and Deane, F. (1997), “The effects of computer anxiety, state anxiety, and computer experience on users’ performance of computer based tasks”, Personality & Individual Differences, Vol. 22 No. 5, pp. 683-692.

MarketLine (2013), “Global online retail”, MarketLine Industry Profile, Reference Code: 0199-2344, London, available at: www.marketline.com (accessed 12 August 2014). Mead, A.D. and Drasgow, F. (1993), “Equivalence of computerized and paper-and-pencil cognitive ability tests: a meta-analysis”, Psychological Bulletin, Vol. 114 No. 3, pp. 449-458. Meuter, M.L., Ostrom, A.L., Bitner, M.J. and Roundtree, R.I. (2003), “The influence of technology anxiety on consumer use and experience with self-service technologies”, Journal of Business Research, Vol. 56 No. 11, pp. 899-906.

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

Mohd Suki, N., Ramayah, T. and Mohd Suki, N. (2008), “Internet shopping acceptance: examining the influence of intrinsic versus extrinsic motivations”, Direct Marketing: An International Journal, Vol. 2 No. 2, pp. 97-110. Montoya-Weiss, M.M., Voss, G.B. and Grewal, D. (2003), “Determinants of online channel use and overall satisfaction with a relational, multi-channel service provider”, Journal of Academy of Marketing Science, Vol. 31 No. 4, pp. 448-459. Morris, M.G. and Venkatesh, V. (2000), “Age differences in technology adoption decisions: implications for a changing workforce”, Personnel Psychology, Vol. 53 No. 2, pp. 375-403. Morris, M.G., Venkatesh, V. and Ackerman, P.L. (2005), “Gender and age differences in employee decisions about new technologies: an extension to the theory of planned behavior”, IEEE Transactions on Engineering Management, Vol. 52 No. 1, pp. 69-84. Nowlis, S., Kahn, B. and Dhar, R. (2002), “Coping with ambivalence: the effect of removing a neutral option on consumer attitude and preference judgments”, Journal of Consumer Research, Vol. 29 No. 3, pp. 319-334. O’Cass, A. and Fenech, T. (2003), “Web retailing adoption: exploring the nature of internet usersweb retailing behaviour”, Journal of Retailing and Consumer Services, Vol. 10 No. 2, pp. 81-94. Ong, C.S. and Lai, J.Y. (2006), “Gender differences in perceptions and relationships among dominants of e-learning acceptance”, Computers in Human Behavior, Vol. 22 No. 5, pp. 816-829. Oyedele, A. and Simpson, P.M. (2007), “An empirical investigation of consumer control factors on intention to use selected self-service technologies”, International Journal of Service Industry Management, Vol. 18 No. 3, pp. 287-306. Pahnila, S., Siponen, M. and Zheng, X. (2011), “Integrating habit into UTAUT: the Chinese eBay case”, Pacific Asia Journal of the Association for Information Systems, Vol. 3 No. 2, pp. 1-30. Park, J., Lee, D. and Ahn, J. (2004), “Risk-focused e-commerce adoption model: a cross-country study”, Journal of Global Information Management, Vol. 7 No. 2, pp. 6-30. Pascual-Miguel, F.J., Agudo-Peregrina, Á.F. and Chaparro-Peláez, J. (2015), “Influences of gender and product type on online purchasing”, Journal of Business Research, Vol. 68 No. 7, pp. 1550-1556. Pavlou, P.A. and Fygenson, M. (2006), “Understanding and predicting electronic commerce adoption: an extension of the theory of planned behavior”, MIS Quarterly, Vol. 30 No. 1, pp. 115-143. Perea y Monsuwé, T., Dellaert, B.G.C. and de Ruyter, K. (2004), “What drives consumers to shop online? A literature review”, International Journal of Service Industry Management, Vol. 15 No. 1, pp. 102-121. Petersen, J.A. and Kumar, V. (2009), “Are product returns a necessary evil? Antecedents and consequences”, Journal of Marketing, Vol. 73 No. 3, pp. 35-51. Powell, A.L. (2013), “Computer anxiety: comparison of research from the 1990s and 2000s”, Computers in Human Behavior, Vol. 29 No. 6, pp. 2337-2381.

Customer online shopping anxiety 305

APJML 28,2

306

Quinones, C. and Kakabadse, N.K. (2015), “Self-concept clarity, social support and compulsive internet use: a study of the US and the UAE”, Computers in Human Behavior, Vol. 44, pp. 347-356. Ramus, K. and Nielsen, N.A. (2005), “Online grocery retailing: what do consumers think?”, Internet Research, Vol. 15 No. 3, pp. 335-352. Research AND Markets (2015), “Turkey B2C e-commerce market 2015”, Research AND Markets, ID No. 3161733, available at: www.researchandmarkets.com/reports/3161733/turkey-b2c-ecommerce-market-2015ID:3161733 (accessed 31 July 2015). Ringle, C.M., Wende, S. and Becker, J.M. (2014), SmartPLS 3, SmartPLS, Hamburg, available at: www.smartpls.com

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

Russell, G. and Bradley, G.L. (1997), “Teachers’ computer anxiety: implications for professional development”, Education & Information Technologies, Vol. 2, pp. 17-30. Saadé, R.G. and Kira, D. (2006), “The emotional state of technology acceptance”, Issues in Informing Science & Information Technology, Vol. 3, pp. 529-539. Schottenbauer, M.A., Rodriguez, B.F., Glass, C.R. and Arnkoff, D.B. (2004), “Computers, anxiety, and gender: an analysis of reactions to the Y2K computer problem”, Computers in Human Behavior, Vol. 20 No. 1, pp. 67-83. Shang, R., Chen, Y. and Shen, L. (2005), “Extrinsic versus intrinsic motivations for customer to shop on-line”, Information & Management, Vol. 42 No. 3, pp. 401-413. Sin Tan, K., Chong, S.C. and Lin, B. (2013), “Intention to use internet marketing: a comparative study between Malaysians and South Koreans”, Kybernetes, Vol. 42 No. 6, pp. 888-905. Singh, N., Fassott, G., Chao, M.C.H. and Hoffman, J.A. (2006), “Understanding international web site usage: a cross-national study of German, Brazilian, and Taiwanese online consumers”, International Marketing Review, Vol. 23 No. 1, pp. 83-97. Slade, E.L., Dwivedi, Y.K., Piercy, N.C. and Williams, M.D. (2015), “Modeling consumers’ adoption intentions of remote mobile payments in the United Kingdom: extending UTAUT with innovativeness, risk, and trust”, Psychology & Marketing, Vol. 32 No. 8, pp. 860-873. Smith, B. and Caputi, P. (2001), “Cognitive interference in computer anxiety”, Behaviour & Information Technology, Vol. 20 No. 4, pp. 265-273. Song, J. and Zahedi, F.M. (2005), “A theoretical approach to web design in e-commerce: a belief reinforcement model”, Management Science, Vol. 51 No. 8, pp. 1219-1235. Stafford, T.F., Turan, A.H. and Khasawneh, A.M. (2006), “MIDDLE-EAST.COM: diffusion of the internet and online shopping in Jordan and Turkey”, Journal of Global Information Technology Management, Vol. 9 No. 3, pp. 43-61. Suh, B. and Han, I. (2003), “Effect of trust on customer acceptance of internet banking”, Electronic Commerce Research & Applications, Vol. 1 No. 3, pp. 247-263. Tai, Y.M. and Ku, Y.C. (2013), “Will stock investors use mobile stock trading? A benefit-risk assessment based on a modified UTAUT model”, Journal of Electronic Commerce Research, Vol. 14 No. 1, pp. 67-84. Taylor, S. and Todd, P. (1995), “Assessing IT usage: the role of prior experience”, MIS Quarterly, Vol. 19 No. 1, pp. 561-570. Tenenhaus, M., Vinzi, V.E., Chatelin, Y.M. and Laura, C. (2005), “PLS path modeling”, Computational Statistics & Data Analysis, Vol. 48 No. 1, pp. 159-205. Thatcher, J.B. and Perrewe’, P.L. (2002), “An empirical examination of individual traits as antecedents to computer anxiety and computer self-efficacy”, MIS Quarterly, Vol. 26 No. 4, pp. 381-396. Thorpe, S.J. and Brosnan, M.J. (2007), “Does computer anxiety reach levels which conform to DSM IV criteria for specific phobia?”, Computers in Human Behavior, Vol. 23 No. 3, pp. 1258-1272.

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

Todman, J. (2000), “Gender differences in computer anxiety among university entrants since 1992”, Computers & Education, Vol. 34 No. 1, pp. 27-35. Venkatesh, V. (2000), “Determinants of perceived ease of use: integrating perceived behavioral control, computer anxiety and enjoyment into the technology acceptance model”, Information Systems Research, Vol. 11 No. 4, pp. 342-365. Venkatesh, V. and Davis, D. (2000), “A theoretical extension of the technology acceptance model: four longitudinal field studies”, Information Systems Research, Vol. 46 No. 2, pp. 186-204. Venkatesh, V. and Morris, M.G. (2000), “Why ’don’t men ever stop to ask for directions? Gender, social influence, and their role in technology acceptance and usage behavior”, MIS Quarterly, Vol. 24 No. 1, pp. 115-139. Venkatesh, V., Morris, M.G. and Ackerman, P.L. (2000), “A longitudinal field investigation of gender differences in individual technology adoption decision-making processes”, Organizational Behavior and Human Decision Processes, Vol. 83 No. 1, pp. 33-60. Venkatesh, V., Thong, J.Y.L. and Xu, X. (2012), “Consumer acceptance and use of information technology: extending the unified theory of acceptance and use of technology”, MIS Quarterly, Vol. 36 No. 1, pp. 157-178. Venkatesh, V., Brown, S.A., Maruping, L.M. and Bala, H. (2008), “Predicting different conceptualizations of system use: the competing roles of behavioral intention, facilitating conditions, and behavioral expectation”, MIS Quarterly, Vol. 32 No. 3, pp. 483-502. Venkatesh, V., Morris, M.G., Davis, G.B. and Davis, F.D. (2003), “User acceptance of information technology: toward a unified view”, MIS Quarterly, Vol. 27 No. 3, pp. 425-478. Vijayasarathy, L.R. (2004), “Predicting consumer intentions to use on-line shopping: the case for an augmented technology acceptance model”, Information & Management, Vol. 41 No. 6, pp. 747-762. Wang, H.Y. and Wang, Y.S. (2008), “Gender differences in the perception and acceptance of online games”, British Journal of Educational Technology, Vol. 39 No. 5, pp. 787-806. Wetzels, M., Odekerken-Schröder, G. and van Oppen, C. (2009), “Using PLS path modeling for assessing hierarchical construct models: guidelines and empirical illustration”, MIS Quarterly, Vol. 33 No. 1, pp. 177-195. Yang, K. (2010), “Determinants of US consumer mobile shopping services adoption: implications for designing mobile shopping services”, Journal of Consumer Marketing, Vol. 27 No. 3, pp. 262-270. Yu, C.S. (2012), “Factors affecting individuals to adopt mobile banking: empirical evidence from the UTAUT model”, Journal of Electronic Commerce Research, Vol. 13 No. 2, pp. 104-121. Zhang, X., Prybutok, V.R. and Koh, C.E. (2006), “The role of impulsiveness in a TAM based online purchasing model”, Information Resources Management Journal, Vol. 19 No. 2, pp. 54-68. Zhou, L., Dai, L. and Zhang, D. (2007), “Online shopping acceptance model: a critical survey of consumer factors in online shopping”, Journal of Electronic Commerce Research, Vol. 8 No. 1, pp. 41-62. Zhou, T. (2012), “Examining location-based services usage from the perspectives of unified theory of acceptance and use of technology and privacy risk”, Journal of Electronic Commerce Research, Vol. 13 No. 2, pp. 135-144. Corresponding author Hakan Celik can be contacted at: [email protected]

For instructions on how to order reprints of this article, please visit our website: www.emeraldgrouppublishing.com/licensing/reprints.htm Or contact us for further details: [email protected]

Customer online shopping anxiety 307

This article has been cited by:

Downloaded by UNIVERSITY OF INDONESIA At 03:08 17 March 2017 (PT)

1. Anil Gupta, Neelika Arora. 2017. Understanding determinants and barriers of mobile shopping adoption using behavioral reasoning theory. Journal of Retailing and Consumer Services 36, 1-7. [CrossRef]