The role of trust and risk perceptions in cloud archiving — Results from an empirical study

Journal of High Technology Management Research 25 (2014) 172–187

Contents lists available at ScienceDirect

Journal of High Technology Management Research

The role of trust and risk perceptions in cloud archiving — Results from an empirical study Daniel Burda ⁎, Frank Teuteberg Institute of Information Management and Information Systems, University of Osnabrueck, Katharinenstrasse 1, 49069 Osnabrueck, Germany

a r t i c l e

i n f o

Available online xxxx Keywords: Cloud Storage Cloud Archiving Cloud Computing Trust Risk SEM

a b s t r a c t This study presents and empirically validates a model that strives to explain end-user adoption of cloud storage as a means of personal archiving. Drawing from prior research on IT adoption, trust, risk and cloud computing, we develop a technology acceptance model that incorporates users' perceptions of risk and trust as well as major antecedents of trust. The research model is empirically tested with survey data collected from 229 cloud storage users. Our results show that trust can be conceived of as a factor that mitigates uncertainty and reduces the perception of risk, which is a significant inhibitor of the intention to use cloud storage for archiving. We find evidence that trust can be increased through both the provider's reputation and user satisfaction. Based on the results, we highlight important practical implications that can be used to inform marketing efforts of cloud storage providers and further suggest some opportunities for future research. © 2014 Elsevier Inc. All rights reserved.

1. Introduction Preserving digital data for the long-term is a challenging task in the light of rapidly changing technologies and the associated risk of obsolete soft-/hardware and media degradation (Burda & Teuteberg, 2013). In the private domain, all the valuable personal files, such as photographs, documents and music, are still primarily archived on traditional media such as local, external hard disks or DVDs (Ion, Sachdeva, Kumaraguru, & Capkun, 2011). From a user perspective, these files are often irreplaceable memories that money cannot buy. However, hard drives will fail eventually, which usually takes place at random and results in a loss of files (Top, 2013). To counteract those threats, consumer cloud storage solutions provide adequate means and have seen an increasing rise in demand and diffusion. Despite abundant headlines about privacy breaches or government surveillance programs, market analysts forecast that cloud storage will continue to grow at an aggressive pace in the next years and consumers are expected to increase their use (Verma, 2012). Using cloud storage, end-users can remotely store their data and use convenient on-demand storage services from a shared pool of highly reliable computing resources, without the burden of local data storage (C. Wang, Chow, Wang, Ren, & Lou, 2013). Compared to archiving on traditional media, archiving in the cloud, referred to as cloud archiving in this study, offers several advantages. It provides central and continuous availability of archived data that can be accessed simultaneously from various devices such as laptops or mobile devices anytime and anywhere. In addition, long-term data access can be preserved without the threat of media obsolescence that is usually extant and requires periodic replication of data onto newer storage media (Burda & Teuteberg, 2013).

⁎ Corresponding author. E-mail addresses: [email protected] (D. Burda), [email protected] (F. Teuteberg).

http://dx.doi.org/10.1016/j.hitech.2014.07.008 1047-8310/© 2014 Elsevier Inc. All rights reserved.

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

173

However, against the backdrop of the advantages of cloud archiving compared to traditional storage media, there are also risks associated with cloud archiving as users relinquish the ultimate control over the fate of their data (Ackermann, Widjaja, Benlian, & Buxmann, 2013; Wang et al., 2013). In fact, when archiving data on cloud storage, additional requirements regarding security, privacy and accountability need to be met in order to fulfill the expectations of the users, gain their trust and build long-term business relationships. The latter is particularly important in the archiving context because data is usually intended to be kept for an infinite period of time. Prior research provides some first empirical insights on user's perceptions of cloud storage and observed a general tendency to prefer local storage over cloud storage. Furthermore, it also showed the existence of latent mistrust, particularly for data considered sensitive by the user (Ion et al., 2011). Thus, it is the intention of this study to investigate the realm of cloud archiving at an individual level of analysis to provide an understanding about end-user acceptance of cloud archiving while focusing on the role of trust and risk. In particular, we strive to answer the following research questions: 1) How do trust and risk perceptions influence cloud archiving adoption decisions? 2) How can trust be established in the context of cloud archiving? We build on previous research that has addressed the issue of risk and uncertainty as important determiners of cloud adoption (Benlian, Hess, & Buxmann, 2009) and the importance of trust in both cloud adoption (Pearson, 2011; Walterbusch, Martens, & Teuteberg, 2013) as well as in the archiving domain (Burda & Teuteberg, 2013). Drawing on the technology acceptance model (TAM) (Davis, 1989), we propose a research model that incorporates trust and risk to investigate the adoption of cloud archiving in the private domain. Additionally, we integrate antecedents of trust to explain how trust can be built by cloud storage providers. The research model and underlying hypotheses are tested by using data collected through a survey of current cloud storage users that already had prior experience with cloud storage. This paper is structured as follows: We first review previous research and discuss the theoretical foundation of this study. Then, we derive our research model and hypotheses. Next, we delineate our research methodology followed by the results of the data analysis. The subsequent discussion section highlights the important findings as well as the practical implications and contributions of this study. Finally, we discuss the limitations and future research directions and conclude the paper. 2. Theoretical framework 2.1. Previous research In the advent of this research, we conducted a literature review. Therefore, we searched the databases of the top 20 MIS journals according to the AIS journal ranking list (AIS, 2013), the proceedings of major IS conferences (e.g. ICIS, ECIS) as well as the Digital Libraries of ACM and IEEE for relevant extant research.1 Our literature review shows a vast amount of research that has been published in pertinent information systems (IS) and computer science journals and conferences. Acknowledging this extant research, we find that studies have been conducted from an organizational, individual as well as technological perspective. While the technological literature focuses on addressing the issues of security, privacy and infrastructure performance by proposing new architectures, methods or prototypes (see, e.g., Brandt, Tian, Hedwig, & Neumann, 2012; Spillner et al., 2011; Wang et al., 2013), current organizational research on cloud computing primarily addresses the issues of opportunities and risks (Benlian & Hess, 2011) as well as decision making in cloud computing adoption. For example, Martens and Teuteberg (2012) propose a cost and risk based decision making model for cloud computing while Repschlaeger, Zarnekow, Wind, and Klaus (2012) suggest a framework to support organizations to systematically gather cloud computing requirements. Other authors examine perceived security risks in cloud computing and their measurement (Ackermann et al., 2013), the measurement of service quality in cloud computing (Benlian, Koufaris, & Hess, 2011), pricing models (Eaton, 2009) or risk/compliance management in cloud computing (Martens & Teuteberg, 2011). Most of the work that examines cloud computing from an individual end-consumer perspective focused on important determiners and inhibitors of cloud computing by applying commonly used IT adoption theories such as TAM or the theory of planned behavior (TPB). For example, Bhattacherjee and Park (2014) study the motivation of end-users to migrate from client-hosted computing to cloud computing and Ratten (2012) examines the impact of ethical and entrepreneurial orientation in cloud computing adoption. Behrend, Wiebe, London, and Johnson (2011) examine adoption behavior of students of software as a service (SaaS) solutions based on TAM3, while Giessmann and Stanoevska (2012) study end-user preferences in platform as a service (PaaS) solutions using a conjoint analysis. However, considering extant research, we only find the work of Ion et al. (2011) who specifically examined the realm cloud storage adoption from an end-user perspective. Ion et al. (2011) empirically investigated users' perceptions and privacy concerns with cloud storage providers. Based on interview and survey data collected from 402 participants, Ion et al. (2011) observed that 69% of all respondents preferred local storage over cloud storage and that they do not use cloud storage as their main storage medium for security and privacy reasons. They conclude that there is a general mistrust of the cloud based on a feeling that the internet is intrinsically insecure. Despite the above research and the work of Ion et al. (2011), there is a lack of research that investigates the realm of consumer cloud storage and in particular its usage as a means of personal archiving from an adoption perspective. Therefore, the present

1

More information on the applied literature review approach (e.g., used keywords/databases) are provided in Appendix A.

174

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

study sets out to take an initial step in filling this gap by examining end-users' intentions to use cloud archiving while acknowledging the users' trust and risk perceptions as well as important determinants of trust. To the best of our knowledge, this is the first study to investigate the cognitive factors that influence an end-user's cloud archiving behavior. 2.2. Hypotheses development and research model TAM has its origins in the theory of reasoned action (TRA) (Fishbein & Ajzen, 1975) and is often referred to as the most influential and commonly employed theory in IS research (Benbasat & Barki, 2007). Extant TAM research has shown that it is a parsimonious and robust model able to explain technology acceptance behaviors in a variety of IT contexts and settings (Paul A. Pavlou, 2003). While TAM was developed to predict individual adoption and use of new IT (Davis, 1989), it has been extended in studies such as Venkatesh and Davis (2000) (TAM2), Venkatesh and Bala (2008) (TAM3) or Gefen, Karahanna, and Straub (2003) by incorporating additional constructs as antecedents of TAM constructs or by integrating IT adoption with user satisfaction research (Wixom & Todd, 2005). TAM has been applied in both work-related and non-organizational research (Gefen et al., 2003) and consistently explains about 40% of the variance in individuals' intention to use an IT across various contexts and countries (Gefen et al., 2003; Venkatesh & Bala, 2008). We decided to base this study on TAM for the following reason. The robustness and parsimoniousness to explain IT adoption in various contexts and underlying relationships in TAM have already been proven multiple times and are arguably to be valid in the context of cloud archiving. Relying on these prior findings therefore gives us the freedom to place our focus on the identification of key drivers of trust in cloud archiving and the relationships with TAM constructs to eventually gain an understanding of the impact of trust in cloud archiving. That is, it is not the focus of this paper to reaffirm TAM relationships. Instead, we seek to understand the intertwining effects of trust and risk in the context of cloud archiving as well as important drivers of trust so that managerial recommendations can be provided which may aid cloud storage providers to increase usage and extend their market share. Subsequently, we elaborate our hypotheses and derive a research model which is depicted in Fig. 1. Following TAM, IT users behave rationally when they decide to use an IT. Based on this premise, a user's intention to use (USE) a new IT is determined by the perceived usefulness (PU) of using the IT and the perceived ease of use (PEOU) of the new IT. In the present study, PU refers to the subjective assessment of the utility offered by a cloud storage system for means of personal archiving (Gefen et al., 2003), while PEOU is viewed as the degree to which a person believes that using a cloud storage system would be free of effort (Davis, 1989). Based on TAM, we posit that the more useful and easy to use is the cloud storage system in enabling the users to archive their data assets in the long run, the more it will be employed. H1. PU will positively affect intended use of cloud archiving. H2. PEOU will positively affect intended use of cloud archiving. H3. PEOU will positively affect PU of cloud archiving. When adopting cloud archiving, uncertainty and different types of risks are present (Ackermann et al., 2013; Ion et al., 2011). Drawing on extant research, we can distinguish two forms of uncertainty that are relevant in the context of cloud archiving: environmental uncertainty and behavioral uncertainty (Pavlou, 2003). Environmental uncertainty stems from the unpredictable nature of the internet which is beyond the full control of the provider. Although the storage provider has a decisive influence on the availability and security of the storage service, there are still inevitable risks such as unintentional downtimes (outages), loss of data access or hacker attacks that threaten both continuity of the service and data privacy (Ackermann, Miede, Buxmann, & Steinmetz, 2011). On the other hand, behavioral uncertainty reflects the possibility that the storage provider may act in an opportunistic manner by taking advantage of the user's data which, in turn, creates additional risks for the user (Pavlou, 2003). For example, the provider may disclose private user information or manipulate and exploit the user's archived data (Ackermann et al., 2013). While behavioral and environmental types of uncertainty and resulting risks are considered interwoven, they are hard to measure as an objective reality. As such, prior research has primarily addressed the notion of perceived risk (Jarvenpaa, Tractinsky, & Vitale, 2000; Van der Heijden, Verhagen, & Creemers, 2003). Consistent with this perspective, we follow Pavlou (2003) and define perceived risk as a user's subjective belief that there is some probability of suffering a loss when archiving in the cloud. This definition comprises both dimensions

Satisfaction

H9 (+)

Trust H7 (+)

H10 (+)

Reputation

H11 (+)

H8(+)

H12(+)

H5 (+)

Perceived Usefulness H3 (+)

Familiarity

Risk

H6 (-)

Ease of Use

Fig. 1. Research model.

H2 (+)

H4 (-)

H1 (+)

Intention to Use

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

175

environmental and behavioral uncertainty and consequently may reflect both opportunistic behavior of the provider (e.g. by disclosing private information) and technology driven flaws (e.g. outages). Acknowledging prior research on SaaS adoption (Benlian & Hess, 2011) and e-commerce (Jarvenpaa et al., 2000; Pavlou & Gefen, 2004; Warkentin, Gefen, Pavlou, & Rose, 2002) that found perceived risk to reduce user's intentions, we formulate: H4. Perceived Risk (RSK) will negatively affect intention to use a cloud archiving. Generally, scholars from various research disciplines agree that whenever uncertainty, interdependence or fear of opportunism in exchange relations are present, trust has a vital influence in enabling these relations (Komiak & Benbasat, 2006; Mayer, Davis, & Schoorman, 1995; Pavlou & Gefen, 2004; Rousseau, Sitkin, Burt, & Camerer, 1998). That is, the need for trust primarily arises in risky situations (Mayer et al., 1995). While trust has been defined in various ways (Gefen et al., 2003; Rousseau et al., 1998), it is frequently defined according to Mayer et al. (1995, p. 726) as “willingness to be vulnerable to another party”. Based on this definition, trust refers to an aggregation of three dominantly cited beliefs about the characteristics of a trustee: ability, benevolence and integrity (Mayer et al., 1995; McKnight, Choudhury, & Kacmar, 2002). Collectively, these three characteristics have also been referred to as “trustworthiness” (Jarvenpaa et al., 2000). Consistent with this conceptualization of trust and extant IS literature that has examined trust in buyer–seller relationships in online environments or trust in recommendations agents, this study defines trust following Pavlou, Liang, and Xue (2007) and Rousseau et al. (1998) as: a user's intentions to accept vulnerability based on his/her beliefs that archiving data with a cloud storage provider will meet his/her confident expectations due to the provider's competence, integrity and benevolence. That is, trust reflects the perception to find what is expected rather than what is feared (McAllister, 1995). Previous studies consistently showed that trust exerts a positive direct effect on (purchase/adoption) intentions (Gefen et al., 2003; Kim, Ferrin, & Rao, 2009; Komiak & Benbasat, 2006; McKnight, Cummings, & Chervany, 1998; Wang & Benbasat, 2005). Contrarily, trust has been found to negatively affect perceived risk in e-commerce transactions (Jarvenpaa et al., 2000; Kim et al., 2009; Pavlou, 2003). Transferring these findings to the context of cloud archiving, where uncertainty and interdependence are present and fostered by the impersonal nature of exchange (Dimoka, 2010), we believe that trust helps users to overcome these concerns and encourages them to adopt cloud archiving by influencing their use intentions (H5) and perception of risk (H6). Moreover, prior empirical studies on trust and TAM in the e-commerce context have found a positive effect of trust on PU (Gefen et al., 2003; Pavlou, 2003; Wang & Benbasat, 2005). Drawing from these results, we posit that trust positively affects PU (H7). The reasoning behind this is that if a storage provider cannot convey trustworthiness and thus cannot be trusted to behave in line with the users' confident beliefs, then there is no reason for the user to expect any benefits from archiving data in the cloud (Pavlou, 2003). The provider may behave in an opportunistic manner by, for instance, disclosing private information, which will reduce the perception of usefulness. Contrarily, if the provider provides the promised benefits reliably and effectively (Ganesan, 1994), the user's perceptions of gaining utility from using cloud storage and, consequently, PU should increase. H5. Trust (TRT) will positively affect intention to use cloud archiving. H6. Trust will negatively affect perceived risk to use cloud archiving. H7. Trust will positively affect PU. In line with the integrated Trust–TAM model proposed by Gefen et al. (2003), we posit that PEOU also leads to higher levels of trust. Extant e-commerce research provides empirical evidence for this relationship arguing that when engaging with another party, people subconsciously use available information, such as appearance as a heuristic to assess trustworthiness of the other party (Gefen et al., 2003; Pavlou, 2003; Vance, Elie-Dit-Cosaque, & Straub, 2008; Wang & Benbasat, 2005). Gefen et al. (2003) assert that if more effort is placed in developing a web site so that it is usable and navigable, users will perceive ease of use and infer that an e-vendor is investing in the relationship thereby signaling a commitment to it. This argument also applies to the adoption of cloud archiving. If the cloud storage provider has expended effort in designing their websites and tools such as plug-ins, that help to integrate a user's desktop easily with the cloud storage (e.g. Dropbox), a user should perceive a higher ease of use which should positively influence a user's assessment of trustworthiness to the provider (H8). H8. PEOU will positively affect trust. Prior IS research has identified several variables that could influence trust. Drawing on this body of knowledge we focus on three specific influences, namely, satisfaction, reputation and familiarity in the interest of parsimony. We consider the selected factors relevant in the context of cloud storage adoption for archiving as described next. Satisfaction is considered a major driver of customer retention (Chiou & Droge, 2006) and defined based on Spreng, MacKenzie, and Olshavsky (1996) as an affective state that is the emotional reaction to the entire cloud storage experience of a user. Satisfaction results from customers' assessment of the perceived performance of a product or service vis-à-vis their initial expectation and the extent to which their expectation is confirmed (Bhattacherjee, 2001; Oliver, 1980). Following Ganesan (1994) in his study about buyer–seller relationships, buyers' satisfaction with outcomes will increase their perception of a vendor's trustworthiness because satisfaction indicates that the vendor is concerned about equitable outcomes and welfare of the buyer (indicating its benevolence). Moreover, satisfaction with a product or service indicates a certain level of performance and competence of the vendor in serving customer needs. This should lead to a higher perception of the vendor's ability, one dimension of trust, and in turn to an increase of the buyer's trust in the vendor. Extrapolating this logic to the context of cloud archiving, we argue that, if users positively perceive

176

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

a provider's product and service quality, which implies that they are satisfied, the fear of opportunistic behavior will decrease and the users will have more confidence in the provider, which, in turn, will lead to a higher trust in the provider (Chiou & Droge, 2006). While a significant influence of satisfaction on trust has been shown in various IS contexts, such as e-commerce acceptance (Pavlou, 2003; Zhou, Lu, & Wang, 2009), application service providing/continuance (Kim, Hong, Min, & Lee, 2011), website loyalty (Flavián, Guinalíu, & Gurrea, 2006) or consumer participation in using online recommendation agents (Dabholkar & Sheng, 2012), there is also research reporting an effect of trust on satisfaction in the context of sales manager–salesperson (Brashear, Boles, Bellenger, & Brooks, 2003) or selling partner relationships (Smith & Barclay, 1997). However, in line with pertinent IS studies, we consider satisfaction as one of the important contributors of trust in a cloud storage provider and thus posit H9: H9. Satisfaction (SAT) with a cloud storage provider will positively affect trust. Reputation was found to exhibit a positive effect on trust in the context of internet stores (Jarvenpaa et al., 2000; Kim, Xu, & Koh, 2004) and identified as important in cloud provider selection research (Koehler, Anandasivam, Dan, & Weinhardt, 2010; Walterbusch et al., 2013). The reason for this is that the costs of untrustworthy behavior are perceived to be higher for firms that already attained a reputation for trustworthiness. Hence, trustworthiness can be conceived an asset which requires significant efforts by engaging in trustworthy behavior, and firms are reluctant to jeopardize their reputation through opportunistic behavior in order to protect the long-term benefits that reputation provides (Chiles & McMackin, 1996; Jarvenpaa et al., 2000). Defining provider reputation based on Doney and Cannon (1997) as the extent to which a user believes a cloud storage provider is honest and concerned about his customers, we hypothesize: H10. Reputation (REP) of a cloud storage provider will positively affect trust. Familiarity is a widely recognized prerequisite of trust and has been shown to predict trust in previous research (Bhattacherjee, 2002; Gefen, 2000; Gefen et al., 2003; Gulati, 1995; Komiak & Benbasat, 2006). Familiarity refers to one's understanding of an entity, often based on prior interactions, experience and learning of “the what, who, how, and when of what is happening” (Gefen et al., 2003, p. 63). As such, it reflects an understanding of present actions of other parties, while trust deals with beliefs about the future actions of others (Gefen, 2000). In the context of this study, user familiarity relates to an assessment of how well the user comprehends the cloud storage provider's procedures and infrastructure, e.g., how to upload/delete data or how to use the configuration interface (cf. Gefen et al., 2003). According to Gefen et al. (2003), familiarity creates trust by mitigating one's concerns that the other party might act opportunistic based on a reliance on previous joint activities when that did not happen. This assumption is likely to be valid in the context of this study. That is, if a user did not see any indication of opportunistic or dishonest behavior by the cloud storage provider previously, e.g., by disclosing or altering user information, then the user's trust should increase (H11). Another effect of familiarity relates to the perceived ease of use. Previous research showed that with increased familiarity, users find an IT easier to use as they have already obtained an understanding of the basic procedures und functionalities (Cho, Kwon, & Lee, 2007; Gefen et al., 2003). Extrapolating this finding to the context of this study, we posit H11 and H12 as follows. H11. Familiarity (FAM) with a cloud storage provider will positively affect trust. H12. Familiarity with a cloud storage provider will positively affect PEOU.

3. Research method In an effort to test our research model in a quantitative way that allows to statistically generalize the findings beyond the studied sample, we employed a cross-sectional survey (Pinsonneault & Kraemer, 1993), which finally was administered online. The measurement and analysis procedures are described subsequently. 3.1. Item development and pretesting To develop appropriate measurement items for this study, we first reviewed extant theoretical and empirical literature. Where possible, we adopted measurement items of the constructs based on existing scales and modified those to make them suitable for our context.2 All constructs in our model are operationalized as reflective constructs and were measured with multiple items on seven-point Likert scales. We developed an online questionnaire which was reviewed in two rounds of personal interviews with three different research colleagues. During the interviews, we presented and discussed all initial measurement items with the interviewees. Based on the consolidated feedback we obtained during the first round of interviews, several items were revised and simplified (MacKenzie, Podsakoff, & Podsakoff, 2011). Moreover, we improved the sequence of the questions which were purposefully randomized (Straub, Boudreau, & Gefen, 2004). The initial instrument was then subjected to a small-scale pretest which was conducted with a convenience sample of 19 respondents randomly drawn from IS faculty members and graduate students. The respondents completed the questionnaire and reported their feedback on the wording, length and concerns if they had any. Based on the resulting set of data, we examined the

2

The items used in this study are provided in the Appendix B.

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

177

validity and reliability of our measurement scales using SPSS. As a result of our data analysis and the comments provided by the respondents, we further revised some of the measurement items by, e.g., adding some negatively worded items. All final items measuring the various constructs were intentionally not related to a specific cloud storage provider to ensure that the items would be relevant for all participants and to allow a general rather than provider specific interpretation of the findings. 3.2. Data collection To test the proposed research model, we developed an online questionnaire and relied on convenience sampling in order to acquire study participants. That is, we invited students and staff from our university and another, affiliated German university via e-mail to participate in the study. Both universities combine various scientific disciplines in research and teaching, e.g., Social Sciences, Law and Business Administration/Economics. Though the use of student samples has often been criticized, we believe that it is an adequate target sample for our study for two reasons. First, extant research indicates a high degree of cloud storage adoption among students (Ion et al., 2011). Thus, it is reasonable to expect that students constitute a significant part of the target population (Compeau, Marcolin, Kelley, & Higgins, 2012) and therefore present good surrogates for cloud storage users with a sufficient level of cloud storage experience. Second, previous research about online behavior found that a student sample can foreshadow the direction in which the general population is moving since it typically represents early adopters of an innovation like cloud storage services (Gallagher, Parsons, & Foster, 2001). In an attempt to increase the response rate, we decided to employ a lottery which is deemed to exert a small, however positive influence on web survey response rates (Heerwegh, 2006; Porter & Whitcomb, 2003). In the invitation e-mail, we informed the potential participants that, if they respond to the survey, they will participate in a price draw of five 50, EUR Amazon gift certificates. Between June and July 2013, we collected 342 completed questionnaires out of which we excluded 80 respondents from our analysis as they indicated not to use any cloud storage services. On average, it took the respondents 13.2 min to complete the survey. In the introduction section of the survey, we briefed the participants that we were only interested in their personal opinion regarding cloud archiving without asking them to imagine a specific or hypothetical cloud storage provider while answering the survey questions. To ease the identification of outliers and unreliable responses during data screening (Marcoulides & Saunders, 2006), we used the Mahalanobis distance statistic (Mahalanobis, 1936) implemented in the software package AMOS. We further excluded another 33 from the remaining 262 cases because of conflicting answers (i.e., showing the same ratings on reverse/negated and positively worded items of the same scale) or unreliable responses (i.e., answering all questions with 7 or alternating 6 and 7). Eventually, a sample of 229 usable and completed questionnaires was used in the data analysis, which corresponds to an effective response rate of 67%. To test for sufficiently stable estimates (Goodhue, Lewis, & Thompson, 2006; Ringle, Sarstedt, & Straub, 2012), we performed a power analysis (Cohen, 1988) using G*Power 3.1.7 (Faul, Erdfelder, Buchner, & Lang, 2009) given four predictors (i.e., the largest number of independent latent variables impacting a particular dependent variable in the inner path model). The analysis showed that our sample size was adequate to detect a medium effect (i.e. f2 N =0.15) with a power of 0.99 (n = 229, p b 0.01). Moreover, a possible nonresponse bias was addressed by adopting the procedure recommended by Armstrong and Overton (1977). We conducted the non-parametric Mann–Whitney U-Test (Mann & Whitney, 1947) to test for differences between the first third and the last third of the respondents' data. The test revealed no significant differences in 27 out of 28 items, so we concluded that nonresponse bias was not an issue in this study. As shown in Table 1, of the 229 respondents in the final sample 33.6% were female and 69.0% were in the 18 to 24 age range. Further, 71.7% of the sampled respondents have been using cloud storage for more than one year, while 50.6% use between less than one and two gigabytes (GB) of cloud storage.3 However, only 34% of respondents employ cloud storage for the means of personal archiving. 3.3. Data analysis We used structural equation modeling (SEM) to test the measurement and structural models. The component-based partial least squares (PLS) approach was chosen and used for both the assessment of the measurement scales and the test of the proposed research hypotheses. While there is still an ongoing discussion about the strengths and weaknesses of different SEM methods, Goodhue, Lewis, and Thompson (2012) found that PLS does not perform worse than LISREL in terms of statistical power and avoidance of false positives. For this study, the PLS approach was selected since it is considered the preferred approach in early stage research and we seek to identify “driver” constructs of trust in cloud archiving rather than confirming established relationships (Goodhue et al., 2012; Hair, Ringle, & Sarstedt, 2011). To assess our model, we used the free of charge software application SmartPLS (version 2.0.M34) for data analysis and closely followed the recommendations given by Straub et al. (2004) for the validation of IS positivist research. 3.3.1. Assessment of reflective measurement model In a first step, we evaluated the individual item reliability and convergent validity of our constructs. Toward this end, we analyzed the factor loadings of the individual items on their hypothesized constructs and the average variance extracted (AVE). We note that all 3 4

A reference to a detailed overview of the participant sample can be found in the Appendix C. More information on SmartPLS can be found at: http://www.smartpls.de.

178

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

Table 1 Profile of respondents (n = 229). Gender

Female: 77 (33.6%)

Age

18–24 years: 158 (69.0%) Less than high school: 2 (0.87%) b3 months: 13 (5.68%) b1 GB: 50 (21.8%)

Level of education Period of cloud storage use Cloud storage used [GB]

Male: 152 (66.4%) 25–34 years: 65 (28.4%) Some college degree: 11 (4.8%) 3–6 months: 38 (16.6%) 1–2 GB: 66 (28.8%)

35–44 years: 4 (1.8%) High school degree: 156 (68.1%) 7–12 months: 54 (23.6%) 3–5 GB: 46 (20.1%)

45–65 years: 2 (0.88%) University degree: 58 (25.3%) 13–24 months: 59 (25.8%) 6–10 GB: 20 (8.7%)

PhD: 2 (0.87%) N2 years: 65 (28.4%) N10 GB: 47 (20.5%)

of the measurement items exhibit loadings that are significant at the 0.01 level on the hypothesized constructs and above the recommended minimum value of 0.707. In addition, each AVE value exceeds the accepted minimum of 0.50, which indicates that the latent construct accounts for at least 50% of the variance in the items. As such, both tests indicate an adequate degree of validity (Chin, 1998). Secondly, we assessed the discriminant validity of our constructs by comparing the square root of the AVE of each construct with all other inter-construct correlations. Our results indicate that our measurement model demonstrates sufficient discriminant validity. The square root of the AVE for each of the constructs is larger than all other inter-construct correlations (Fornell & Larcker, 1981). Following the recommendations given by Gefen and Straub (2005), we also examined the cross loadings of the individual items. The test revealed that each item loading is above 0.71 on the assigned target construct and at least 0.1 less on other constructs which indicates adequate convergent and discriminant validity. Thirdly, we evaluated the internal consistency and scale reliability by calculating the composite reliability (CR) and Cronbach's alpha values as can be seen from Table 2. The CR values for all of the constructs in our model are greater than 0.81 while the Cronbach's alpha values are greater than 0.66. This shows a satisfactory reliability for both criteria since all values are above the commonly accepted minimum thresholds of 0.6 or 0.7 respectively (Bagozzi & Yi, 1988; Gefen, Straub, & Boudreau, 2000). Table 2 summarizes the results of our assessment and shows the average variance extracted, Cronbach's alpha and composite reliability values of all constructs. Moreover, the occurrence of common method bias (CMB) is considered a threat to the validity and conclusions of a study. CMB can be understood as a bias that stems from the same method or part of method being used for multiple measurements (Burton-Jones, 2009). In an effort to reduce the likelihood of CMB, all items in the questionnaire were randomized making it more difficult for respondents to sense the inherent constructs, which in turn may influence their answers (Straub et al., 2004). However, as we used a questionnaire in which all of the data for both the predictor and criterion variables were obtained from the same person and in the same measurement context, CMB may arise from a variety of sources, such as social desirability or consistency motif (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). To encounter the threat of CMB, we performed three statistical tests. In a first step, we conducted a Harman's single-factor test in which all of the items were subjected to an unrotated exploratory factor analysis to determine whether (a) a single factor emerges or (b) one general factor accounts for the majority of the covariance among the measures (Podsakoff et al., 2003). The results of the test yielded 6 factors, the largest of which accounts for 34.06% of the variance, suggesting that CMB is not an issue in this study. For the second test, we incorporated a latent common methods variance factor (LCMVF) into our PLS model as described by Liang, Saraf, Hu, and Xue (2007). Using this approach, each indicator is converted into a single-indicator construct and linked with both the LCMVF and the theorized constructs thereby transforming them to second-order constructs. This enables the calculation of each indicator's variances as substantively explained by the theorized construct as well as by the method and thus enabling the measurement of the influence of CMB on the indicators. As a result of the test, we find that the average substantively explained variance by the major constructs is 0.72 while the average method related variance is 0.012. The ratio of substantive variance to method variance is about 60:1 and the majority of method factor loadings are not significant. While the above approach has also been criticized regarding its effectiveness to control for CMB, we also conducted a marker variable test as suggested by Rönkkö and Ylitalo (2011). A marker variable should be theoretically unrelated with the constructs under study (Bagozzi, 2011). Therefore, we a priori decided to include the construct “perception of causes of global warming” (Heath & Gifford, 2006) into our survey. The construct is measured with 3 items. The marker variable test reveals no difference in the significance of any path coefficients after including the marker variable in our model. Further, the results exhibit that all correlations between the marker variable and the focal constructs are less than 0.12. The maximum percentage of shared variance between the marker variable and the other constructs in the model is 1.3% indicating the amount of variance that can be attributed to method variance. Acknowledging this small magnitude and the results of the two former tests, we believe that CMB is unlikely to be a contaminant for the results of this study. 3.3.2. Assessment of structural model As discussed above, the structural model was estimated using the PLS approach. To test the significance of our loadings and coefficients, we employed the bootstrapping re-sampling technique with 229 cases and 1,000 samples. The obtained estimates from the PLS analysis, including standardized path coefficients, significance of the paths and the amount of variance explained (R2), are

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

179

Table 2 Reliabilities, AVE and latent variable correlations. CA: Cronbach's alpha, CR: Composite reliability, Shaded cells: Square root of AVE. Appendix B/D provides more details on indicator means, standard deviations, loadings and cross loadings.

CA

CR

AVE

PEOU

PEOU

0.80

0.88

0.71

0.85

FAM

USE

REP

RSK

SAT

TRT

FAM

0.79

0.88

0.71

0.54

0.84

USE

0.98

0.99

0.96

0.32

0.25

REP

0.89

0.92

0.75

0.28

0.27

0.14

0.86

RSK

0.66

0.81

0.59

–0.40

–0.32

–0.45

–0.32

0.77

SAT

0.80

0.87

0.62

0.66

0.40

0.30

0.44

–0.45

0.79

TRT

0.88

0.92

0.73

0.41

0.33

0.30

0.62

–0.44

0.54

0.85

PU

0.84

0.90

0.76

0.44

0.33

0.78

0.16

–0.39

0.36

0.34

PU

0.98

0.87

presented in Fig. 2. Acknowledging the R2 values, we see that our model accounts for 63.6% of the variance in intention to use, 21.9% of the variance in perceived usefulness, 29.6% of the variance in perceived ease of use, 19.2% of the variance in perceived risk and 48.7% of the variance in trust. While the TAM constructs including perceived risk and trust explain 63.6% of the variance in intention to use, the control variables general IT knowledge (CEX), period of experience with cloud storage (YEX), disposition to trust (DTT), gender, amount of cloud storage used (STU) altogether account for additional 0.3%. Notwithstanding, none of the path coefficients of our control variables on intention is significant, we performed the following two step analysis to examine the significance of the increase in R2. Toward this end, we calculated the effect size (f2) of the control variables in a first step (Chin, Marcolin, & Newsted, 2003). Secondly, a pseudo F-test was conducted by multiplying the effect size by (n–k–1) where n is the sample size and k is the number of independent variables of the full model, i.e., including the five control variables (Mathieson, Peacock, & Chin, 2001). We find an effect size of 0.01 indicating no significant effect (F = 0.36, p N 0.05) according to Cohen (1988). Acknowledging the significant path coefficients in Fig. 2, we see that intention is significantly determined by perceived usefulness (b = 0.74, p b 0.01) and perceived risk (b = −0.20, p b 0.01), therewith lending support for H1 and H4. Contrary to the TAM postulates and H2, ease of use shows no direct effect on intention (b = −0.08, p N 0.05). In an effort to explain this result, we conducted a mediation analysis following Baron and Kenny (1986) and a Sobel test to assess the significance of mediation effects (Sobel, 1982). In line with Baron and Kenny (1986), our analysis results indicate that the influence of PEOU on intention is significantly mediated by PU (z = 4.96, p b 0.01) denoting an indirect effect of 0.26 on intention to use. Trust does not directly affect intention and thus leads to a rejection of H5. However, the effect of trust is significantly mediated by both perceived risk (z = 3.51, p b 0.01) and PU (z = 2.57, p b 0.05) exhibiting a cumulative indirect effect of 0.23 on intention. We note that trust affects both perceived risk (b = 0.44, p b 0.01) and PU (b = 0.19, p b 0.01) providing support for H6 and H7. Identifying the antecedents of trust, we see that trust is significantly determined by satisfaction (b = 0.27, p b 0.01), reputation (b = 0.47, p b 0.01) thereby confirming H9 and H10. In contrast, familiarity and PEOU exert no effect on trust while familiarity significantly affects PEOU (b = 0.54, p b 0.01). Thus, H8 and H11 are rejected while H12 finds empirical support as shown in Fig. 2.

4. Discussion, implications and theoretical contribution 4.1. Discussion This study empirically examines a user's intention to archive his/her data using cloud storage based on TAM while acknowledging the role of trust and risk perceptions. In particular, we pay attention to the factors that influence a user's trust in cloud archiving which indirectly influences a user's intention through PU and perceived risk. Overall, we find good support for our theoretical model that is validated with data collected from 229 respondents who all were users of cloud storage solutions for at least three months and thus had some experience with cloud storage. Based on the empirical results of this study, we next discuss three important findings. First, our data show that the users' intention to use cloud archiving is influenced by both their risk perceptions as well as the TAM construct perceived usefulness. We find that PU exhibits the strongest influence on intention in our model as in many previous TAM studies (Gefen & Straub, 2000) indicating that cloud storage is considered an instrumental and appropriate IT for archiving one's digital belongings. As our mediation analysis shows, trust exerts indirect influence on intention to use via perceived risk and PU. Therewith, trust becomes an important component of cloud storage adoption decisions that significantly diminishes perceived risk on the one hand and increases the perceptions of usefulness of cloud storage as a means of personal archiving on the other hand. While this finding provides support for prior cloud adoption studies that identified the need for trust in cloud provider selection decisions (e.g., Walterbusch et al., 2013), it also parallels results from a meta-analysis on digital preservation that considers trust as one of the most important requirements in long-term archiving (Burda & Teuteberg, 2013). In line with other studies from the arena of SaaS and e-commerce adoption (e.g., Benlian & Hess, 2011; Pavlou, 2003), we find a negative influence of perceived risk on intention.

180

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

Fig. 2. Results of PLS analysis.

However, caution should be paid when interpreting this finding due to the relatively small path coefficient (r = −0.20, p b 0.01) and low explained variance in perceived risk (R2 = 0.192). A possible explanation for this finding may be a measurement issue that is indicated by the comparably low Cronbach's alpha value of the risk construct (0.66). Previous studies noted that risk is a complex construct that is difficult to measure. Weber, Blais, and Betz (2002) provide empirical support that risk attitudes exhibit cross-national, cultural and content domain-specific differences that are influenced by the differences in the perception of risk with respect to a particular behavior (Hsee & Weber, 1999). Other authors also report general difficulties in question based self-report measurement of risk with few reflective indicators. For instance, Pennings and Smidts (2000) conclude that it may be advisable to use measurement methods based on decision scenarios/gambles to infer risk preferences (see also, e.g., Qualls & Puto, 1989). Second, we see that the small effect of ease of use on intention is fully mediated by PU resulting in no direct influence of PEOU on intention. This finding has already been observed in previous studies and a possible explanation for this is provided by Karahanna, Straub, and Chervany (1999) who argue that after initial adoption and acquired experience with a system, ease of use concerns seem to be resolved and displaced by other considerations such as efficacy. This explanation is likely to be valid in our study since we explicitly selected a sample drawn from current cloud storage users. Acknowledging the index values of the PEOU construct, we find a high index value and low standard deviation (6.07 on a seven-point Likert-type scale, SD = 0.83) indicating that user's find cloud storage solutions extremely easy to use. And although ease of use is significantly correlated with trust (r = 0.414, p b 0.01), it does not influence trust in our model. Likewise, our data show that familiarity neither directly nor indirectly affects a user's trust in the cloud storage provider. That is, users may be familiar with the cloud storage solution and perceive it as easy to use, however, this does not affect their trust in the provider and, in turn, their intention to use. These results stand in contrast to prior ecommerce research (Gefen, 2000; Gefen et al., 2003) indicating that familiarity with a cloud provider and ease of use play a different role than in e-commerce contexts. This finding might be explained by Gefen (2000), who asserts that trust and familiarity are contextdependent and their cause and effect relationship may be contingent on the specific products and services under study. Third, we identified two direct antecedents of trust in the context of cloud storage adoption. Our results indicate satisfaction and reputation to be significant antecedents of trust. The formation of trust in the study's context depends on a user's overall satisfaction that is based on his/her experience with the provider and solution as well as his/her beliefs about the provider's reputation, i.e., beliefs about the provider's honesty and care for customer concerns. Acknowledging the relative high standardized path coefficient of reputation on trust (b = 0.47) highlights the importance of a provider's reputation in the formation of trust. This finding corroborates extant research on cloud service preferences that identified a provider's reputation as a key attribute by means of a conjoint analysis. Reputation was found to exert the highest relative importance compared to, e.g., pricing in cloud provider selection decisions (Koehler et al., 2010). 4.2. Implications for practice and theoretical contributions In summary, our results show that trust and risk are important elements in the acceptance and use of cloud archiving. Trust among current cloud storage users diminishes risk perceptions and impacts the perceptions of usefulness. Risk perceptions, in contrast to perceived usefulness, reduce a user's intention to employ cloud storage for personal archiving. Trust is influenced by user satisfaction and reputation. These findings offer some implications that are important for practitioners and particularly cloud storage providers striving to understand end-users cloud archiving adoption behavior. According to Ajzen (1991), an understanding of what factors influence a particular behavior is essential as each serves as a point of attack in an attempt to change behavior. Based on this understanding, effective interventions can be implemented and tailored so that they address the most salient factors which are deemed to provide substantial leverage in changing the behavior (Fishbein, 2008). This study uncovers some of these important elements that should be of interest for practitioners and particularly cloud storage providers. First, given the strong influence of reputation on trust, providers of cloud storage services need to pay attention and actively shape their reputation of being a reliable storage provider. Toward that end, cloud service providers could employ reputation advertising measures to foster and retain their reputation, such as public relations campaigns or the employment of evangelists who actively promote and advocate a provider in press articles, blogs or talks by disseminating positive information. Therewith, influence can be

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

181

exercised that increases the perception of a provider's reputation which is difficult to replicate for competitors and thus may constitute effective market entry barriers (Fombrun & Shanley, 1990). In addition, previous research provides empirical evidence that customers who perceive a firm to have a good reputation are more willing to engage in positive word of mouth than those that do not perceive the firm to have a good reputation (Walsh, Mitchell, Jackson, & Beatty, 2009). As such, a positive reputation may also support the acquisition of new customers. As an alternative strategy to build reputation, cloud providers could also rent reputation of other high-reputation infrastructure providers in order to signal customers about their credibility and attractiveness (Fombrun & van Riel, 1997). This strategy has already been employed in other contexts and is also applied by some cloud storage providers. For instance, the cloud storage provider Dropbox relies and actively signals the use of the Amazon's Simple Storage Service (S3) which is considered a highly reliable service and one of the big names in the market (Brodkin, 2011). Second, acknowledging the TAM construct usefulness as the most important predictor of a user's intention to use cloud archiving, cloud storage provider should clearly convey the value proposition of cloud archiving by highlighting the advantages in comparison to archiving on traditional media. In this attempt, storage provider could use real-life scenarios that showcase the threats of media/hardware obsolescence and media degradation that can be avoided using cloud storage. These scenarios could, e.g., be illustrated by means of readily understandable video clips or other materials that inform about prominent examples where archived data was lost due to obsolescence issues (see, e.g., NASA Goddard Space Flight Center, 2001). Promoting cloud storage as a sort of insurance that guarantees the long-term accessibility of all digital personal belongings should significantly increase the perception of usefulness and in turn the intention to use cloud archiving. In these efforts, focus should be placed on reminding people that conventional storage media such as hard disks or DVDs are likely to fail eventually and become obsolete which presents a significant risk to the accessibility of their data. Third, we find that user satisfaction contributes to the formation of trust. While this finding might not be surprising, it points toward both the need to retain customers and to cross- and up-selling potentials, i.e., additional sources of revenue for the cloud provider. To satisfy customers, cloud storage providers are called to provide high quality customer support that goes beyond the provision of FAQ sections on the provider's website. In terms of system quality, providers should provide different ways to access the cloud storage, e.g., by means of mobile applications or other innovative features which can increase customer satisfaction. Satisfied users tend to trust more and in turn are more willing to extend their current use by employing cloud storage for archiving. As archiving of important personal belonging, such as photographs, should usually require far more storage than most of the free accounts offer, there is an up-selling potential that could be exploited by storage providers. From a theoretical point of view, this study contributes to the body of research that investigates the adoption of cloud computing from an individual perspective acknowledging the role of trust and risk perceptions in adoption decisions. To the best of the author's knowledge, this study is the first that sheds light on the use of cloud storage as a means of personal archiving specifically. While the topic of cloud archiving has gained little attention from the IS community to date, our proposed model may open new venues for future research that investigates how people handle their digital belongings, that they desire to retain in the long-run, in face of the rise in virtualized hardware and storage solutions. 5. Limitations and future research Like all studies, this one is not free of limitations. One limitation of the paper is the drawn sample. We surveyed experienced cloud storage users who were mainly IS students. Although students are a significant part of the target population, external validity might be limited due to student's comparably lower age, higher education, technical affinity, less disposable income and less crystallized attitudes compared to the general population (Sears, 1986). However, previous research noted that when studying an innovation such as cloud archiving, a student sample may more appropriately represent users than a randomly drawn sample from a general population. Given that our sample consists of experienced cloud storage users rather than the late majority of adopters, the results may anticipate the direction in which the general population is stirring (Gallagher et al., 2001). Moreover, we collected data from two European countries only. That is, caution should be taken when these findings are to be generalized to other regions. This is because national culture has not only been found to substantially affect risk perceptions, but also technology acceptance (Leidner & Kayworth, 2006). Acknowledging the sample demographics, it further becomes obvious that approximately 34% of the subjects currently archive their data using cloud storage which may also affect the results given that this subgroup may exhibit differences compared to the remaining 66% of participants. To investigate this issue, future studies could apply multigroup analyses with our model (Sarstedt, Henseler, & Ringle, 2011) to elicit potential differences between the “archivers” and “non-archivers”. However, apart from the limitations related to the drawn sample, we follow the notion of Compeau et al. (2012) who argue that establishing generalizability is about process rather than the design of a single study. Thus, in an effort to generalize the findings, future research is required to replicate the study using different samples that better approximate a general, non-student population. Besides that, alternative SEM approaches, such as LISREL, could be used to validate the proposed research model. Another limitation lies in the measurement of trust. In this study, trust has been measured as a unidimensional construct. This preempts a more detailed examination of the effects and relative importance of the underlying factors of trust. As such, future research could proceed by conceptualizing trust as a second order construct to trace back the importance of its constituents and employ measurement methods that recently have been proposed to measure trust perceptions specifically with regard to information technology and security respectively (e.g., Ackermann et al., 2013; Mcknight, Carter, Thatcher, & Clay, 2011). Finally, there is a limitation associated with the cross-sectional design of the study which precludes a more dynamic view for understanding the adoption of cloud archiving over time. Although we based our research model on established theories and a comprehensive literature review, a longitudinal design may have provided more confidence for the results. In particular, our results might be

182

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

biased due to a vast amount of headlines about U.S. intelligence programs focusing on the surveillance of internet user activities. During the period of data collection, the issue of surveillance of internet users and internet companies was intensively discussed in various media. Some respondents explicitly referred to that issue in optional free text fields in our online-questionnaire indicating that at least some of the respondent's ratings might have been biased due to that fact. As such, future research could proceed by surveying a group of individuals across time. Using this longitudinal design, scholars could investigate, e.g., trust and risk perceptions across time and also observe the effects of publicly reported issues (e.g. data privacy infringements and breaches, headlines related to specific cloud storage providers, topics such as data loss incidents). Further, future research could acknowledge the role of privacy in cloud storage adoption. While extant research on information privacy provides empirical evidence for the intertwining of information privacy concerns and risk perceptions (e.g., Malhotra, Kim, & Agarwal, 2004), it would be interesting to see how these cause and effect relationships applied to the field of cloud archiving. 6. Conclusion In this study, we investigate the acceptance of cloud storage as a means of personal archiving. Drawing on TAM, we examine the role of trust and risk perceptions and identify important antecedents of trust. Our research model is empirically tested with survey data collected from 229 cloud storage users. Our results show that trust can be thought of as a factor that mitigates uncertainty and reduces the perception of risk which again reduces intention to use. On the other hand, trust indirectly influences intention through perceived usefulness. These findings suggest the importance of risk reduction and trust in cloud archiving adoption decisions. In addition, we show that trust is influenced by a user's satisfaction with the cloud storage provider and a provider's reputation. While our findings shed some initial light on the realm of cloud archiving from a technology acceptance angle, more research is needed to examine other important determinants in cloud archiving adoption decisions as well as to generalize the findings beyond the drawn sample in this study. Appendix A. Literature review approach To examine extant research, we queried the following databases that provided us with access to a wide range of publications from several Computer Science disciplines and IS comprising major journals and conference proceedings: (1) ACM Digital Library, (2) IEEE Xplore Digital Library, (3) EBSCOhost (Business Source Premier Database) and (4) Association for Information Systems Electronic Library (AISel). To focus our review on high quality literature, we restricted the queries to the top 20 MIS journals according to the AIS journal ranking (AIS, 2013) and the proceedings of major IS conferences such as ICIS or ECIS. We queried for the general terms

Table 3 Literature review results. Database

ACM

IEEE

EBSCOhost

AISel

Limiters

Limit to ACM and affiliated organizations-publications 12.03.2013 6 4 (−2) 2 (0)

Limit to IEEE publications 12.03.2013 40 34 (−6) 6 (−28)

Limit to publications of the top 20 MIS journals 12.03.2013 19 19 (0) 4 (−15)

Limit to publications of the top 20 MIS journals 12.03.2013 41 41 (0) 15 (−26)

Search date Results Available download After analyzing abstract After backward/forward and additionally exploratory search

∑

106 98 (−8) 27 (−71) 48 (+21)

“cloud” or “*aas” or “archive” or “retention” or “retain” or “preserv*” or “backup*” or “adopt*” or “diffu*” or “acceptance” or “usage” or “intention” included in the abstract of the paper and did not restrict the timeframe of the search to gain exhaustive results.5 The search resulted in 127 sources of which we were able to download 125 due to access restrictions. We evaluated these 127 papers regarding their relevance to the topic of cloud archiving by reading the papers' abstracts. This led to a reduction of relevant sources from 127 to 27. Next, we conducted both backward and forward references search and obtained additional 21 papers leading to a total of 48 papers ranging from year 1984 to 2013. Table 3 summarizes our results.

5 Example search-term used to query the IEEE database: ((((“Abstract”:cloud) or (p_Abstract:“*aas”)) and ((p_Abstract:“presev*”) or (p_Abstract:“archiv*”) or (p_Abstract:retain) or (p_Abstract:retention) or (p_Abstract:backup))) and ((p_Abstract:adopt*) or (p_Abstract:diffu*) or (p_Abstract:acceptance) or (p_Abstract:usage) or (p_Abstract:intention))).

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

183

Appendix B. Item sources, item wording, item means, SD, loadings

Table 4 Constructs and measurement items. Construct

Source

Item wording

Mean

SD

Loading

Intention to use

(Gefen et al., 2003)

Ease of use

(Wixom & Todd, 2005)

Usefulness

(Davis, 1989; Gefen et al., 2003)

USE1: I would use cloud storage to archive my personal data. USE2: I am very likely to archive my personal data using cloud storage. USE3: I intend to use cloud storage for personal archiving in the future. PEOU1: Cloud storage is easy to use. PEOU2: It is easy to get cloud storage to do what I want it to do. PEOU3: Learning to operate cloud storage is easy. PU1: Cloud storage enables me to archive and retrieve my personal data faster. PU2: Cloud storage enhances my effectiveness in archiving and retrieving my personal data. PU3: Overall, I find cloud storage useful for archiving my personal data. RSK1: There is a high potential for loss involved in using cloud provider for personal archiving. RSK2: There is a considerable risk involved in using cloud storage for personal archiving. RSK3: A decision to use cloud storage for personal archiving is risky. TRT1: Overall, my cloud storage provider is trustworthy. TRT2: My cloud storage provider wants to be known as one who keeps promises and commitments. TRT3: I trust my cloud storage provider keeps my best interests in mind. TRT4: Even if not monitored, I'd trust my cloud storage provider to do the job right. SAT1: I am satisfied in general with my past transactions with my current cloud storage provider. SAT2: I feel frustrated about my overall experience with your cloud storage provider.a SAT3: Overall, I am pleased with my cloud storage provider. SAT4: I am dissatisfied with the products and services I have received from my cloud storage provider in the past.a REP1: My cloud storage provider is known to be dependable. REP2: My cloud storage provider has a good reputation in the market. REP3: My cloud storage provider has a reputation for dependability. REP4: My cloud storage provider has a poor reputation in the market.a FAM1: I am familiar with storing data in the cloud. FAM2: I am familiar with cloud storage. FAM3: I am familiar with organizing my personal data on cloud storage.

5.03 5.16 5.00 6.10 6.07 6.03 5.07

1.77 1.72 1.81 0.87 0.80 0.82 1.68

0.989 0.971 0.983 0.819 0.887 0.829 0.848

5.12

1.37

0.879

5.44 2.59

1.45 1.22

0.880 0.707

3.37

1.53

0.737

3.83 5.39 5.47

1.54 1.04 0.96

0.849 0.874 0.849

5.32 5.55

0.95 0.90

0.822 0.872

6.00

0.73

0.880

6.19

0.86

0.745

5.45 6.23

1.05 0.96

0.762 0.756

5.36 5.84 5.39 5.76 6.00 5.69 5.41

1.05 0.97 1.05 1.26 1.02 1.28 1.44

0.915 0.835 0.901 0.800 0.729 0.893 0.895

Risk

Trust

Satisfaction

(Pavlou & Gefen, 2004)

(Gefen, 2000; Jarvenpaa et al., 2000)

(Pavlou, 2003; Spreng et al., 1996)

Reputation

(Pavlou, 2003)

Familiarity

(Gefen, 2000)

a

Reverse coded items.

Appendix C. Detailed participant profiles Table 5 Detailed participant profiles.

Gender Male Female Age 18–24 years 25–34 years 35–44 years 45–54 years 55–65 years Education High school degree Some College degree University degree Other Occupation Student Clerk Official

Frequency

Percentage

152 77

66.38% 33.62%

158 65 4 1 1

69.00% 28.38% 1.75% 0.44% 0.44%

156 11 58 4

68.12% 4.80% 25.33% 1.74%

217 5 3

94.76% 2.18% 1.31%

184

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

Table 5 (continued) Frequency Manager Other Knowledge of computers Low Intermediate High Very high Amount of cloud storage used b1 GB 1–2 GB 3–5 GB 6–10 GB N10 GB Duration of experience with cloud storage b3 months 3–6 months 7–12 months 13–24 months N2 years Purpose of use (multiple choice) Temporary/short-term storage Archiving Backup File sharing Collaboration Synchronization with other devices

Percentage

1 3

0.44% 1.31%

1 17 120 91

0.44% 7.42% 52.40% 39.74%

50 66 46 20 47

21.83% 28.82% 20.09% 8.73% 20.52%

13 38 54 59 65

5.68% 16.59% 23.58% 25.76% 28.38%

115 78 116 196 148 138

50.22% 34.06% 50.66% 85.59% 64.63% 60.26%

Appendix D. Item loadings and cross loadings Table 6 Item loadings and cross loadings.

PEOU1 PEOU2 PEOU3 FAM1 FAM3 FAM5 USE1 USE2 USE3 REP1 REP2 REP3 REP4* RSK1 RSK3 RSK4 SAT1 SAT2* SAT3 SAT4* TRT1 TRT2 TRT3 TRT4 PU1 PU2 PU4

Ease of use

Familiarity

Intention to use

Reputation

Risk

Satisfaction

Trust

Usefulness

0.819 0.886 0.829 0.403 0.480 0.488 0.311 0.328 0.303 0.256 0.208 0.270 0.218 −0.424 −0.211 −0.286 0.588 0.468 0.563 0.422 0.368 0.353 0.300 0.388 0.287 0.444 0.390

0.470 0.457 0.453 0.729 0.893 0.895 0.236 0.269 0.219 0.280 0.204 0.258 0.168 −0.397 −0.115 −0.231 0.374 0.259 0.395 0.179 0.246 0.378 0.200 0.296 0.236 0.353 0.272

0.274 0.291 0.244 0.143 0.189 0.279 0.989 0.971 0.983 0.156 0.090 0.143 0.086 −0.234 −0.206 −0.506 0.269 0.218 0.282 0.161 0.275 0.234 0.267 0.235 0.563 0.646 0.792

0.217 0.268 0.215 0.323 0.154 0.211 0.128 0.157 0.128 0.915 0.835 0.901 0.800 −0.273 −0.143 −0.298 0.364 0.293 0.331 0.382 0.570 0.542 0.490 0.519 0.210 0.122 0.103

−0.379 −0.326 −0.298 −0.236 −0.211 −0.359 −0.460 −0.416 −0.442 −0.302 −0.238 −0.290 −0.285 0.711 0.742 0.843 −0.433 −0.310 −0.342 −0.294 −0.396 −0.364 −0.328 −0.405 −0.339 −0.276 −0.387

0.530 0.661 0.472 0.315 0.337 0.358 0.299 0.308 0.285 0.369 0.358 0.402 0.378 −0.434 −0.256 −0.341 0.880 0.745 0.762 0.756 0.464 0.449 0.417 0.524 0.278 0.353 0.299

0.333 0.364 0.353 0.273 0.235 0.322 0.292 0.297 0.280 0.609 0.506 0.567 0.449 −0.349 −0.295 −0.362 0.511 0.299 0.487 0.353 0.873 0.849 0.822 0.872 0.317 0.274 0.295

0.364 0.406 0.330 0.194 0.284 0.349 0.751 0.763 0.778 0.161 0.097 0.143 0.156 −0.227 −0.142 −0.440 0.316 0.221 0.386 0.150 0.337 0.254 0.311 0.258 0.848 0.879 0.880

References Ackermann, T., Miede, A., Buxmann, P., & Steinmetz, R. (2011). Taxonomy of technological IT outsourcing risks: Support for risk identification and quantification. Paper presented at the European Conference on Information Systems (ECIS 2011), Helsinki, Finland. Ackermann, T., Widjaja, T., Benlian, A., & Buxmann, P. (2013). Perceived IT security risks of cloud computing: Conceptualization and scale development. Paper presented at the 33rd International Conference on Information Systems (ICIS 2012), Orlando, FL, US. AIS (2013). MIS journal rankings. Retrieved July 3, 2013, from. http://ais.affiniscape.com/displaycommon.cfm?an=1&subarticlenbr=432 Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179–211.

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

185

Armstrong, J., & Overton, T. (1977). Estimating nonresponse bias in mail surveys. Journal of Marketing Research, 14, 396–402. Bagozzi, R. P. (2011). Measurement and meaning in information systems and organizational research: methodological and philosophical foundations. MIS Quarterly, 35(2), 261–292. Bagozzi, R. P., & Yi, Y. (1988). On the evaluation of structural equation models. Journal of the Academy of Marketing Science, 16(1), 74–94. Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 1173–1182. Behrend, T. S., Wiebe, E. N., London, J. E., & Johnson, E. C. (2011). Cloud computing adoption and usage in community colleges. Behaviour & Information Technology, 30(2), 231–240. Benbasat, I., & Barki, H. (2007). Quo vadis TAM? Journal of the Association for Information Systems, 8(4), 7. Benlian, A., & Hess, T. (2011). Opportunities and risks of software-as-a-service: Findings from a survey of IT executives. Decision Support Systems, 52(1), 232–246. Benlian, A., Hess, T., & Buxmann, P. (2009). Drivers of SaaS-adoption — An empirical study of different application types. Business & Information Systems Engineering, 1(5), 357–369. Benlian, A., Koufaris, M., & Hess, T. (2011). Service quality in software-as-a-service: Developing the SaaS-qual measure and examining its role in usage continuance. Journal of Management Information Systems, 28(3), 85–126. Bhattacherjee, A. (2001). Understanding information systems continuance: An expectation–confirmation model. MIS Quarterly, 25(3), 351–370. Bhattacherjee, A. (2002). Individual trust in online firms: Scale development and initial test. Journal of Management Information Systems, 19(1), 211–242. Bhattacherjee, A., & Park, S.C. (2014). Why end-users move to the cloud: A migration-theoretic analysis. European Journal of Information Systems, 23(3), 357–372. Brandt, T., Tian, Y., Hedwig, M., & Neumann, D. (2012). Autonomic management of software as a service systems with multiple quality of service classes. Paper presented at the 20th European Conference on Information Systems (ECIS 2012), Barcelona, Spain. Brashear, T. G., Boles, J. S., Bellenger, D. N., & Brooks, C. M. (2003). An empirical test of trust-building processes and outcomes in sales manager–salesperson relationships. Journal of the Academy of Marketing Science, 31(2), 189–200. Brodkin, J. (2011). Amazon bests Microsoft, all other contenders in cloud storage test. Retrieved December 2, 2013, from. http://arstechnica.com/business/2011/12/ amazon-bests-microsoft-all-other-contenders-in-cloud-storage-test/ Burda, D., & Teuteberg, F. (2013). Sustaining accessibility of information through digital preservation: A literature review. Journal of Information Science, 39(4), 439–455. Burton-Jones, A. (2009). Minimizing method bias through programmatic research. MIS Quarterly, 33(3), 445–471. Chiles, T. H., & McMackin, J. F. (1996). Integrating variable risk preferences, trust, and transaction cost economics. Academy of Management Review, 21(1), 73–99. Chin, W. W. (1998). The partial least squares approach to structural equation modeling. In G. A. Marcoulides (Ed.), Modern methods for business research (pp. 295–336). Mahwah, NJ, US: Lawrence Erlbaum Associates. Chin, W. W., Marcolin, B.L., & Newsted, P. R. (2003). A partial least squares latent variable modeling approach for measuring interaction effects: Results from a Monte Carlo simulation study and an electronic-mail emotion/adoption study. Information Systems Research, 14(2), 189–217. Chiou, J. -S., & Droge, C. (2006). Service quality, trust, specific asset investment, and expertise: Direct and indirect effects in a satisfaction–loyalty framework. Journal of the Academy of Marketing Science, 34(4), 613–627. Cho, D. -Y., Kwon, H. J., & Lee, H. -Y. (2007). Analysis of trust in internet and mobile commerce adoption. Paper presented at the 40th Annual Hawaii International Conference on System Sciences (HICSS 2007), Waikoloa, Hawaii. Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale, N.J.: Lawrence Erlbaum. Compeau, D., Marcolin, B., Kelley, H., & Higgins, C. (2012). Research commentary — Generalizability of information systems research using student subjects — A reflection on our practices and recommendations for future research. Information Systems Research, 23(4), 1093–1109. Dabholkar, P. A., & Sheng, X. (2012). Consumer participation in using online recommendation agents: Effects on satisfaction, trust, and purchase intentions. The Service Industries Journal, 32(9), 1433–1449. Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319–340. Dimoka, A. (2010). What does the brain tell us about trust and distrust? Evidence from a functional neuroimaging study. MIS Quarterly, 34(2), 373–396. Doney, P.M., & Cannon, J. P. (1997 Aprill). An examination of the nature of trust in buyer–seller relationships. The Journal of Marketing, 61, 35–51. Eaton, B.D. (2009). Bid price control and dynamic pricing in clouds. Paper presented at the 17th European Conference on Information Systems (ECIS 2009), Verona, Italy. Faul, F., Erdfelder, E., Buchner, A., & Lang, A. -G. (2009). Statistical power analyses using G* power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41(4), 1149–1160. Fishbein, M. (2008). A reasoned action approach to health promotion. Medical Decision Making, 28(6), 834–844. Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An introduction to theory and research. Reading, MA: Addison-Wesley. Flavián, C., Guinalíu, M., & Gurrea, R. (2006). The role played by perceived usability, satisfaction and consumer trust on website loyalty. Information & Management, 43(1), 1–14. Fombrun, C., & Shanley, M. (1990). What's in a name? Reputation building and corporate strategy. Academy of Management Journal, 33(2), 233–258. Fombrun, C., & van Riel, C. (1997). The reputational landscape. Corporate Reputation Review, 1(1), 5–13. Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, XVII, 39–50. Gallagher, K., Parsons, J., & Foster, K. D. (2001). A tale of two studies: Replicating ‘advertising effectiveness and content evaluation in print and on the web’. Journal of Advertising Research, 41(4), 71–82. Ganesan, S. (1994). Determinants of long-term orientation in buyer–seller relationships. The Journal of Marketing, 58(2), 1–19. Gefen, D. (2000). E-commerce: The role of familiarity and trust. Omega, 28(6), 725–737. Gefen, D., Karahanna, E., & Straub, D. W. (2003). Trust and TAM in online shopping: An integrated model. MIS Quarterly, 27(1), 51–90. Gefen, D., & Straub, D. W. (2000). The relative importance of perceived ease of use in IS adoption: A study of e-commerce adoption. Journal of the Association for Information Systems, 1(8), 1–28. Gefen, D., & Straub, D. (2005). A practical guide to factorial validity using PLS-graph: Tutorial and annotated example. Communications of the Association for Information Systems, 16(1), 5. Gefen, D., Straub, D., & Boudreau, M. C. (2000). Structural equation modeling and regression: Guidelines for research practice. Communications of the Association for Information Systems, 4(1), 7. Giessmann, A., & Stanoevska, K. (2012). Platform as a service — A conjoint study on consumers' preferences. Paper presented at the International Conference on Information Systems (ICIS 2012), Orlando, FL, US. Goodhue, D., Lewis, W., & Thompson, R. (2006). PLS, small sample size, and statistical power in MIS research. Paper presented at the 39th Annual Hawaii International Conference on System Sciences (HICSS'06), Hawaii. Goodhue, D. L., Lewis, W., & Thompson, R. (2012). Does PLS have advantages for small sample size or non-normal data? MIS Quarterly, 36(3), 891–1001. Gulati, R. (1995). Does familiarity breed trust? The implications of repeated ties for contractual choice in alliances. Academy of Management Journal, 38(1), 85–112. Hair, J. F., Ringle, C. M., & Sarstedt, M. (2011). PLS-SEM: Indeed a silver bullet. Journal of Marketing Theory and Practice, 19(2), 139–152. Heath, Y., & Gifford, R. (2006). Free-market ideology and environmental degradation: The case of belief in global climate change. Environment and Behavior, 38(1), 48–71. Heerwegh, D. (2006). An investigation of the effect of lotteries on web survey response rates. Field Methods, 18(2), 205–220. Hsee, C., & Weber, E. (1999). Cross-national differences in risk preference and lay predictions. Journal of Behavioral Decision Making, 12, 165–179. Ion, I., Sachdeva, N., Kumaraguru, P., & Capkun, S. (2011). Home is safer than the cloud!: Privacy concerns for consumer cloud storage. Paper presented at the 7th Symposium on Usable Privacy and Security, Pittsburgh, Pennsylvania. Jarvenpaa, S. L., Tractinsky, N., & Vitale, M. (2000). Consumer trust in an internet store. Information Technology and Management, 1(12), 45–71.

186

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

Karahanna, E., Straub, D. W., & Chervany, N. L. (1999). Information technology adoption across time: A cross-sectional comparison of pre-adpotion and post-adoption beliefs. [Article]. MIS Quarterly, 23(2), 183–213. Kim, D. J., Ferrin, D. L., & Rao, H. R. (2009). Trust and satisfaction, two stepping stones for successful e-commerce relationships: A longitudinal exploration. Information Systems Research, 20(2), 237–257. Kim, J., Hong, S., Min, J., & Lee, H. (2011). Antecedents of application service continuance: A synthesis of satisfaction and trust. Expert Systems with Applications, 38, 9530–9542. Kim, H. -W., Xu, Y., & Koh, J. (2004). A comparison of online trust building factors between potential customers and repeat customers. Journal of the Association for Information Systems, 5(10), 392–420. Koehler, P., Anandasivam, A., Dan, M., & Weinhardt, C. (2010). Cloud services from a consumer perspective. Paper presented at the 16th Americas Conference on Information Systems (AMCIS), Lima, Peru. Komiak, S. Y., & Benbasat, I. (2006). The effects of personalization and familiarity on trust and adoption of recommendation agents. MIS Quarterly, 30(4), 941–960. Leidner, D. E., & Kayworth, T. (2006). A review of culture in information systems research: Toward a theory of information technology culture conflict. Management Information Systems Quarterly, 30(2), 357–399. Liang, H., Saraf, N., Hu, Q., & Xue, Y. (2007). Assimilation of enterprise systems: The effect of institutional pressures and the mediating role of top management. MIS Quarterly, 31(1), 59–87. MacKenzie, S. B., Podsakoff, P.M., & Podsakoff, N.P. (2011). Construct measurement and validation procedures in MIS and behavioral research: Integrating new and existing techniques. MIS Quarterly, 35(2), 293–334. Mahalanobis, P. C. (1936). On the generalised distance in statistics. Paper presented at the National Institute of Science of India. Malhotra, N. K., Kim, S. S., & Agarwal, J. (2004). Internet Users' Information Privacy Concerns (IUIPC): The construct, the scale, and a causal model. Information Systems Research, 15(4), 336–355. Mann, H. B., & Whitney, D. R. (1947). On a test of whether one of two random variables is stochastically larger than the other. Annals of Mathematical Statistics, 18(1), 50–60. Marcoulides, G. A., & Saunders, C. (2006). PLS: A silver bullet? MIS Quarterly, 30(2), iii–ix. Martens, B., & Teuteberg, F. (2011). Risk and compliance management for cloud computing services: Designing a reference model. Paper presented at the 17th Americas Conference on Information Systems (AMCIS), Detroit, US. Martens, B., & Teuteberg, F. (2012). Decision-making in cloud computing environments: A cost and risk based approach. Information Systems Frontiers, 14(4), 871–893. Mathieson, K., Peacock, E., & Chin, W. W. (2001). Extending the technology acceptance model: The influence of perceived user resources. ACM SIGMIS Database, 32(3), 86–112. Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. McAllister, D. J. (1995). Affect- and cognition-based trust as foundations for interpersonal cooperation in organizations. Academy of Management Journal, 38(1), 24–59. Mcknight, D. H., Carter, M., Thatcher, J. B., & Clay, P. F. (2011). Trust in a specific technology: An investigation of its components and measures. ACM Transactions on Management Information Systems (TMIS), 2(2), 12. McKnight, D. H., Choudhury, V., & Kacmar, C. (2002). Developing and validating trust measures for e-commerce: An integrative typology. Information Systems Research, 13(3), 334–359. McKnight, D. H., Cummings, L. L., & Chervany, N. L. (1998). Initial trust formation in new organizational relationships. Academy of Management Review, 23(3), 473–490. NASA Goddard Space Flight Center (2001). Disposition of original — IUE tapes at the national space science data center. Retrieved January 12, 2014, from. http://nssdc. gsfc.nasa.gov/astro/iuepaper.html Oliver, R. L. (1980). A cognitive model of the antecedents and consequences of satisfaction decisions. Journal of Marketing Research, 17(4), 460–469. Pavlou, P. A. (2003). Consumer acceptance of electronic commerce: Integrating trust and risk with the technology acceptance model. International Journal of Electronic Commerce, 7(3), 101–134. Pavlou, P. A., & Gefen, D. (2004). Building effective online marketplaces with institution-based trust. Information Systems Research, 15(1), 37–59. Pavlou, P. A., Liang, H., & Xue, Y. (2007). Understanding and mitigating uncertainty in online exchange relationships: A principal–agent perspective. MIS Quarterly, 31(1), 105–136. Pearson, S. (2011). Toward accountability in the cloud. IEEE Computer Society, 15(4), 64–69. Pennings, J. M., & Smidts, A. (2000). Assessing the construct validity of risk attitude. Management Science, 46(10), 1337–1348. Pinsonneault, A., & Kraemer, K. L. (1993). Survey research methodology in management information systems: An assessment. Journal of Management Information Systems, 10(2), 75–105. Podsakoff, P.M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N.P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879–903. Porter, S. R., & Whitcomb, M. E. (2003). The impact of lottery incentives on student survey response rates. Research in Higher Education, 44(4), 389–407. Qualls, W. J., & Puto, C. P. (1989). Organizational climate and decision framing: An integrated approach to analyzing industrial buying decisions. Journal of Marketing Research, XXVI, 172–192. Ratten, V. (2012). Entrepreneurial and ethical adoption behaviour of cloud computing. The Journal of High Technology Management Research, 23(2), 155–164. Repschlaeger, J., Zarnekow, R., Wind, S., & Klaus, T. (2012). Cloud requirement framework: Requirements and evaluation criteria to adopt cloud solutions. Paper presented at the 20th European Conference on Information Systems (ECIS), Barcelona, Spain. Ringle, C. M., Sarstedt, M., & Straub, D. (2012). A critical look at the use of PLS-SEM in MIS quarterly. MIS Quarterly, 36(1), iii–xiv. Rönkkö, M., & Ylitalo, J. (2011). PLS marker variable approach to diagnosing and controlling for method variance. Paper presented at the 32nd International Conference on Information Systems (ICIS 2011), Shanghai, China. Rousseau, D.M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. Academy of Management Review, 23(3), 393–404. Sarstedt, M., Henseler, J., & Ringle, C. M. (2011). Multigroup analysis in partial least squares (PLS) path modeling: Alternative methods and empirical results. Advances in International Marketing, 22, 195–218. Sears, D. O. (1986). College sophomores in the laboratory: Influences of a narrow data base on social psychology's view of human nature. Journal of Personality and Social Psychology, 51(3), 515–530. Smith, J. B., & Barclay, D. W. (1997). The effects of organizational differences and trust on the effectiveness of selling partner relationships. The Journal of Marketing, 61(1), 3–21. Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equation models. Sociological Methodology, 13, 290–312. Spillner, J., Bombach, G., Matthischke, S., Muller, J., Tzschichholz, R., & Schill, A. (2011). Information dispersion over redundant arrays of optimal cloud storage for desktop users. Paper presented at the 4th IEEE International Conference on Utility and Cloud Computing (UCC), Melbourne, Australia. Spreng, R. A., MacKenzie, S. B., & Olshavsky, R. W. (1996). A reexamination of the determinants of consumer satisfaction. Journal of Marketing, 60(3), 15–32. Straub, D., Boudreau, M. C., & Gefen, D. (2004). Validation guidelines for IS positivist research. Communications of the Association for Information Systems, 13(24), 380–427. Top (2013). Why use online backup? Retrieved July 2, 2012, from. http://www.thetop10bestonlinebackup.com Van der Heijden, H., Verhagen, T., & Creemers, M. (2003). Understanding online purchase intentions: Contributions from technology and trust perspectives. European Journal of Information Systems, 12(1), 41–48. Vance, A., Elie-Dit-Cosaque, C., & Straub, D. W. (2008). Examining trust in information technology artifacts: The effects of system quality and culture. Journal of Management Information Systems, 24(4), 73–100. Venkatesh, V., & Bala, H. (2008). Technology acceptance model 3 and a research agenda on interventions. Decision Sciences, 39(2), 273–315. Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the technology acceptance model: Four longitudinal field studies. Management Science, 46(2), 186–204. Verma, S. (2012). Forecast: Consumer digital storage needs, 2010–2016. Stamford: Gartner, Inc.

D. Burda, F. Teuteberg / Journal of High Technology Management Research 25 (2014) 172–187

187

Walsh, G., Mitchell, V. -W., Jackson, P. R., & Beatty, S. E. (2009). Examining the antecedents and consequences of corporate reputation: A customer perspective. British Journal of Management, 20(2), 187–203. Walterbusch, M., Martens, B., & Teuteberg, F. (2013). Exploring trust in cloud computing: A multi-method approach. Paper presented at the 21st European Conference on Information Systems (ECIS), Utrecht, Netherlands. Wang, W., & Benbasat, I. (2005). Trust in and adoption of online recommendation agents. Journal of the Association for Information Systems, 6(3), 72–101. Wang, C., Chow, S., Wang, Q., Ren, K., & Lou, W. (2013). Privacy-preserving public auditing for secure cloud storage. IEEE Transactions on Computers, 62(2), 362–375. Warkentin, M., Gefen, D., Pavlou, P. A., & Rose, G. M. (2002). Encouraging citizen adoption of e-government by building trust. Electronic Markets, 12(3), 157–162. Weber, E. U., Blais, A.R., & Betz, N. E. (2002). A domain-specific risk-attitude scale: Measuring risk perceptions and risk behaviors. Journal of Behavioral Decision Making, 15(4), 263–290. Wixom, B. H., & Todd, P. A. (2005). A theoretical integration of user satisfaction and technology acceptance. Information Systems Research, 16(1), 85–102. Zhou, T., Lu, Y., & Wang, B. (2009). The relative importance of website design quality and service quality in determining consumers' online repurchase behavior. Information Systems Management, 26(4), 327–337.