Privacy in the shopping-street

Privacy in the shopping-street

A study about the role of the privacy calculus in the effect of firm

trustworthiness on the intention to use Location-Based Advertising.

Marketing Master Thesis

Faculty of Economics and Business

University of Groningen

Abstract

Because of the global growth in mobile Internet usage, location-based advertising (LBA) is perceived as a big opportunity for the marketing area. LBA is defined as any application, service, or campaign that uses geographic locations to deliver or enhance a marketing message/service. Despite the benefits of personalization and localization using LBA, consumers take the risks of information disclosure into account too. This so-called privacy calculus is the main focus of this study. Previous studies mainly focused on personal characteristics and the intention to use LBA, leaving the service provider out of scope. In this study it is investigated what role the privacy calculus has in the relationship between a firm’s perceived trustworthiness and user’s intention to use LBA. A survey about a fictional mobile application in a department store setting was conducted among 187 people. The type of LBA and information control was manipulated to check whether this influences the privacy calculus. Results showed that the privacy calculus has a significant role regarding the intention to use LBA. There were mixed results about the relationship between the trustworthiness proxies perceived reputation, positive attitude and purchase frequency on the one hand and the privacy calculus on the other hand. The main implication of this study is that providers of LBA should take the privacy calculus into account when introducing a new service.

Table of contents

2.2.1 Privacy disclosure benefits 7

2.2.2 Privacy disclosure risks 8

2.3 Trust in LBA provider 10

2.4 Information delivery mechanism 11

2.5 Information control 13 2.6 Control Variables 14 2.7 Conceptual Framework 16 3. Methodology 16

Appendix B: Factor loadings, AVE and Cronbach’s Alpha 41

Appendix C: Correlations 43

1. Introduction

Expecting the number of smartphones users to surpass the 2 billion worldwide in 2016 (eMarketer 2014), the use of mobile Internet will increase too (Ericsson mobility report 2015). From a marketing perspective, this development has led to a big opportunity in one of the latest mobile trends: location-based services (LBS). LBS are defined as network-location-based services that integrate a derived estimate of a mobile device’s location or position with other information so as to provide added value to the user (Barnes, 2003) as well to personalize the service (Snekkiness 2001). Literature suggests that LBS can be divided into several types (Dhar & Varshney 2011). Barkuus & Dey (2003) made a distinction between location-tracking and position-aware LBS, whereas others call them respectively pull- and push-based LBS (Xu & Tao 2010; Dhar & Varshney 2011; Lin et al. 2015). The main difference between these two types is that position-aware or pull-based services are based on the device’s own knowledge of its position (e.g. Taxi-service Uber, where people can order a taxi based on their current location), whereas location-tracking or push-based services are services relying on the tracking of peoples’ location by other parties such as mobile telephony service providers or retailers (e.g. Facebook’s push-messages, where Facebook sends its users notifications about events in their current neighborhood). Where marketers previously traced online browse history to target customers in an online setting, LBS make it possible to target customers based on their current location in the offline world. This is called location-based advertising (LBA). The Mobile Marketing Association (2011) defines LBA as any application, service, or campaign that uses geographic locations to deliver or enhance a marketing message/service. Despite the big opportunity for both consumer and merchant, many hesitate to use these services (Xu et al., 2011). The main reason for this hesitation of adopting LBA comes from the debate and controversy about potential threats to privacy (Zhou 2011; Dhar & Varshney 2011). While LBA provides mobile consumers the capability of being constantly reachable and accessing network services while “on the move”, they also introduce risks for mobile consumers who disclose location information to service providers or other parties (Xu & Gupta 2009). This dilemma is called the personalization privacy paradox or privacy calculus, where in the case of LBA, consumers disclose personal information and location data in return for added value such as personalized advertisements that will be delivered based on their context and location (Xu et al. 2011). This personalization privacy calculus is perceived to be interesting to further investigate, since the global market of location-based services is forecasted to increase with an annual growth rate of 22.5% from €10.3 billion in 2014 towards €34.8 billion in 2020 (Berg Insight, 2015).

focusing on the privacy concerns. Moreover, they take LBS as one single construct, but are clearly focusing on the position-aware LBS, with their pull-based ‘Send-a-Taxi’ application. This is a limitation since Barkhuus & Dey (2003) found that the level of concerns for privacy varied in the different types of LBS. The study of Xu et al. (2011) about the privacy calculus in location aware marketing is to my knowledge the first covering both the relation between privacy calculus and the decision-making process for location aware marketing and made a distinction between push and pull based services, mentioning them as covert versus overt services.

The marketing literature suggests several consumer characteristics as useful in understanding how to reduce the negative effect of privacy calculus on intentions to use LBS. Xu and Gupta (2009) mention personal innovativeness as moderator of privacy concerns on adoption of LBS; Chen & Chang (2013) mention gender, age and experience as moderators of the attitude towards Near Field Communication on intention to use NFC; Xu & Teo (2010) mention industry self-regulation and government regulation and Okazaki et al. (2012) use social anxiety and situational involvement as moderators in the relation between privacy concern and the intention to use Quick Response promotions.

However, there is a lack of research about the influence of the LBA provider’s characteristics on the intention to use LBA and the role the privacy calculus in this relationship. These characteristics seem to be very relevant to study since the provider can influence these characteristics, while consumer characteristics cannot. Literature mentions the importance of firm trustworthiness in user acceptance of a new service (e.g. Jarvenpaa et al. 1999). There is to my knowledge no study about the role of the privacy calculus. Therefore, the purpose of this paper is to find out:

What is the role of the privacy calculus in the relationship between a firm’s trustworthiness and user’s intention to use LBA?

Perceived firm reputation, user satisfaction with the firm, loyalty and purchase frequency will be used as proxies for trustworthiness. These proxies stem from the study of Javenpaa et al. (1999), Jin et al. (2008) and Alhabeeb (2007). Besides this relationship, this paper will study the moderating role of both perceived information control and the delivery mechanism of the LBA too. These two possible moderators are mentioned in several other studies (Xu & Teo 2010; Taylor et al. 2009; Mothersbaugh et al. 2010).

2. Theory

2.1 Location-based advertising

Traditionally, LBS have been used in such areas as route-finding, roadside assistance and directory services. With the arrival of the mobile phones with GPS connectivity, location-based services made it possible to make use of new applications across different domains ranging from tracking and navigation systems to directory services, entertainment to emergency services, and various mobile commerce applications. Nevertheless, The location-tracking LBS were in the past mainly used by applications to locate friends by using the mobile phone as positioning device (Tekawade et al. 2013). This study will mainly focus on location-based advertising (LBA) as part of LBS. The Mobile Marketing Association (2011) defines location-based advertising (LBA) as any application, service, or campaign that uses geographic locations to deliver or enhance a marketing message/service. Bruner II & Kumar (2007) define LBA as marketer-controlled information customized for recipients’ geographic positions and received on mobile communication devices. They mention LBA as part of location-based services (LBS). Besides this, they state that it is important to make a distinction between LBA and mobile advertising. Mobile advertising is the broader of the two concepts and is used primarily to refer to ads that are sent to individual’s cell phones, while LBA is a very specific type of mobile advertising based on consumers’ current location.

Where previously consumer could only be located outdoors due to GPS limitations, new technologies make it possible to use locational information from sensors, RFID, Bluetooth and WLAN-networks. One of the latest developments is Bluetooth Low Energy (BLE). With BLE, the exact positioning became more precisely. It offers a range of broadcast advertising modes including sending general advertisements that can be detected by any phone with Bluetooth functionality. BLE consists of two separate devices, where the first device sends a unique identifier, a so-called beacon, whereas the second device is a device that can process this identifier, like mobile phones (Huth 2015). Retail stores are the largest users of these beacons.

Due to time and resource constraints and the novelty of the LBA concepts, the focus of this study will be on consumers’ intention to use LBA instead of actual behavior. Davis’ Technology Acceptance Model (1989) and Venkatesh’ Unified Theory of Acceptance and Usage of Technology (2003) take intention to use as dependent variable too. Venkatesh mentions that intention to use is a good predictor of actual behavior, and since actual behavior of a new technology like LBA is difficult to measure since there will be many non-users, intention to use will be used.

2.2 Privacy Calculus

of the consumer’s preferences and behavior. Van Doorn & Hoekstra (2013) argue that although the fit of the personalized advertisement may increase its relevance, resulting in positive behavioral consequences, the use of more personal information could lead to feelings of intrusiveness. Xu et al. (2011) mention this dilemma in the LBA setting. They state that the potential intrusion of privacy becomes an important concern for smartphone users. To consumers, on the one hand, they may identify great value in receiving personalized advertisements, while on the other hand privacy concerns about disclosing personal information in exchange for promotional messages could discourage them. Hagel & Rayport (1997) found that consumers are willing to disclose personal information when they know that they will receive substantial benefits in return. This dilemma is called the personalization privacy paradox. Xu & Teo (2009) argue that the decision whether or not disclosing personal information depends on a risk-benefit analysis in which consumers assess the outcomes they would receive as the result of providing personal information to LBS providers. Culnan and Bies (2003) argued that individuals will disclose personal information if they perceive that the overall benefits of disclosure are at least equal or greater than the risk of disclosure. This risk-benefit analysis is called the privacy calculus. The privacy calculus has often been used since it has been found in literature that the calculus perspective of privacy is the most useful framework for measuring consumer privacy concerns (Culnan & Bies 2003). In the same study, they mention privacy calculus as the second exchange, based on transaction theory where value in the form of goods or services is given in return for money or other goods. They added a second exchange where the consumer makes a non-monetary exchange of their personal information for value such as better service or bigger discounts. Implementing this theory in the LBA context, one could see the information disclosure in LBA as an exchange where consumers disclose their personal information and location data in return for added value such as personalized advertisements that will be delivered based on their context and location (Xu et al. 2011). They argue that the outcome of the cumulated effects of risk and benefits in this calculus will lead to the perceived value of the LBA. The definition of perceived value of information disclosure comes from the study of Xu et al. (2011) and is stated as the individual's overall assessment of the utility of information disclosure based on perceptions of privacy risks incurred and benefits received.

2.2.1 Privacy disclosure benefits

relevant information or services based on the user’s location, identity, activity, and time and mention two main benefits of LBS: They identify locatability as construct for the combination of time- and location-dependent benefit and personalization as construct for the user-dependent benefit.

Xu & Teo (2009) define locatability as the consumer’s perceived value of being able to access needed information/services at the right time in the right place. Therefore they state that locatability enabled by positioning and timeliness is a key advantage in LBA for consumers to exchange their personal information for getting access to needed information or services at anytime from anywhere. These positioning and timeliness dimensions must always come together before users can gain any value from LBA and are inherent with information disclosure in LBA. Lin et al. (2015) mention the advantages of targeting consumers in specific locations and time too, arguing that in the case of LBA this is especially the case since consumers nowadays carry their mobile devices anywhere and anytime.

The other advantage often mentioned in LBS literature is personalization. Personalization has been generally defined as the ability to provide content and services that are tailored to individuals based on knowledge about their preferences and behaviors (Adomavicius & Tuzhilin 2005). Xu & Teo (2009) adapt this definition to the LBS context, defining personalization as the extent to which the LBS can be tailored to consumers’ activity contexts, preferences, and needs. When LBS are personalized to individual customers’ interests, location, identity, activity and time, personalization is gained (Xu et al. 2011). In the LBA context, this is easily achievable since the advertising services are tied to the customer’s mobile phone, revealing the customer’s current location, activity and time. Moreover, Xu & Teo (2009) mention that with information about consumers’ geographical locations, service providers could use their marketing and advertising to tailor different messages for different consumers. In the personalization context, Gazley et al. (2015) mention perceived control of data as key factor of personalization. Through the development of new technologies such as mobile Internet for smartphones, consumers gained control over both content and timing of the advertising messages (McMillan and Hwang 2002). When customers have the perception of being able to make a choice about what product information to receive, the perceived value of the LBA will increase. Therefore is hypothesized that higher benefits of information disclosure will lead to higher perceived value of the LBA.

H1: The benefits of information disclosure are positively related towards the perceived value of the LBA.

2.2.2 Privacy disclosure risks

Gupta 2009; Clarke 2001; Xu et al. 2011). Although LBA offers great opportunities for customers in the form of locatability and personalization as already mentioned, these messages are often viewed as being unwelcome and irritating (Gazley et al. 2015). Van Doorn & Hoekstra (2013) and Edwards et al. (2002) mention this phenomenon as intrusiveness, what is stated as the interruption of editorial content. Consumers can get the feeling that the advertising message can interrupt their actions and that the advertiser is watching them (Greengard 2012). Therefore, a person’s privacy concern is the key factor in the privacy calculus. Although some studies take privacy concern as a single construct, most studies nowadays argue that privacy concern is a multi-dimensional concept (Zhou 2011). Most studies use the Concern For Informational Privacy (CFIP) scale by Smith et al. (1996). This scale has since served as some of the most reliable scales for measuring individuals’ concerns toward organizational privacy practices (Smith et al. 2011). This conceptualization defines privacy concern with four components: collection of personal information, improper access to personal information, errors in personal information and secondary use of personal information. Collection reflects the concern that too much individual personal information is being collected and stored in databases of the service provider. When service providers collect users’ personal information and current location, there is the risk that these providers may acquire more information than necessary, also called information over-collection, raising the privacy concern of consumers. Unauthorized secondary use reflects the concern that information is collected from individuals for one purpose but is used for other secondary purposes without consent (Smith et al. 1996). Due to the high value of the collected data, LBA-providers may share this information with third parties, or even sell it for profit. This decreases users’ benefits and brings great privacy risk to them. Errors reflect the concern that protections against deliberate and accidental errors in personal data are inadequate (Smith et al. 1996). Mobile service providers may not adopt measures such as verification to ensure information accuracy. This has a negative effect on the service quality of LBS and increases users’ concern about information errors. Improper access reflected the concern that data about individuals are readily available to people not duly authorized to view or work with data. Although most LBA-providers guarantee that only authorized personnel can access the collected data, it is impossible to ensure that the provider’s database will not be hacked or intruded (Xu & Gupta 2009; Zhou 2011). When a consumer’s privacy concern is high, the perceived value of the LBA will decrease. Therefore it is hypothesized that the higher the perceived risk of information disclosure, the lower the perceived value will be.

H2: The risks of information disclosure are negatively related towards the perceived value of the LBA.

H3: The privacy calculus is positively related towards intention to use LBA.

2.3 Trust in LBA provider

sequences, preferences over time and proportion over time. In his study he mentions that many other studies mention two different aspects of loyalty: the behavioural (purchase) loyalty, which results in repeated purchases, and the attitudinal loyalty, which represents a commitment to a brand or firm. In the current study, these two types of loyalty are separated as well. The behavioural purchase loyalty will from now on be mentioned as purchase frequency, whereas the attitudinal loyalty will be referred to as loyalty to the firm.

Perceived trustworthiness is positively related to citizen’s intention to use an e-government service (Carter & Belanger 2015). It is expected that the higher the trustworthiness of a LBA provider, the lower the consumer’s privacy concern will be. Therefore, the perceived value of the LBA will increase, leading to higher intentions to use LBA. Since perceived reputation, satisfactions with the firm, loyalty to the firm and purchase frequency are antecedents of firm trustworthiness; the same relations are expected as the relation between trustworthiness and intention to use. Therefore it is hypothesized:

H4a: Reputation of the LBA provider is positively related towards the benefits of information disclosure.

H4b: Reputation of the LBA provider is negatively related towards the risks of information disclosure.

H5a: Satisfaction with the LBA provider is positively related towards the benefits of information disclosure.

H5b: Satisfaction with the LBA provider is negatively related towards the risks of information disclosure.

H6a Loyalty to the LBA provider provider is positively related towards the benefits of information disclosure.

H6b Loyalty to the LBA provider is negatively related towards the risks of information disclosure.

H7a Purchase frequency at the LBA provider provider is positively related towards the benefits of information disclosure.

H7b Purchase frequency at the LBA provider is negatively related towards the risks of information disclosure.

2.4 Information delivery mechanism

but most studies mention these two different approaches as push- and pull-based LBA (e.g. Bruner II & Kumar 2007; Dhar & Varshney 2011; Lin et al. 2015 Xu & Teo 2009). This push versus pull distinction will be used in this paper. In push-based LBA, marketers send relevant advertisements to consumers by observing their behavior through tracking the current locations of their mobile devices. With these data, personalization systems personalize the advertisements based on the user's distance to a store or merchant (Xu et al. 2011), encouraging them to purchase certain products or services (Dhar & Varshney 2011). Bruner II & Kumar (2007) divide push-based LBA in opt-in and opt-out services. In the opt-in approach LBA providers will only send advertisements when the user previously indicated their interest in a specific store. In the opt-out approach, LBA providers are able to send their advertisements to everyone in the specified area, until users mention that they do not want to receive the advertisements anymore. In the pull-based LBA, the providers only locate users’ mobile devices when they initiate the requests. This type of LBA is often referred to as ‘on-demand services’. In this approach, location information is useful only to complete the requested transaction (e.g. sending special offers of the nearest merchant or store to the user). Barkhuus & Dey (2003) found out that location-tracking services (push-based LBA) generate more privacy concerns than position-based services (pull-based LBA). Xu & Teo (2009) state that compared to the pull-based mechanism, consumers in push-based LBS need to disclose a larger volume of personal information and location data to gain the benefits of locatability and personalization, which would represent the consumer’s additional input, but represent a positive outcome for the firm, thereby lowering consumers’ perceptions of distributive justice. Therefore it is expected that the effects of firm trustworthiness on the privacy calculus will be stronger in push- versus pull-based services. Therefore it is hypothesized:

H8a: The effects of firm reputation on the benefits of information disclosure will be stronger in push- versus pull-based services.

H8b: The effects of firm reputation on the risks of information disclosure will be stronger in push- versus pull-based services.

H8c: The effects of firm satisfaction on the benefits of information disclosure will be stronger in push- versus pull-based services.

H8d: The effects of firm satisfaction on the risks of information disclosure will be stronger in push- versus pull-based services.

H8e: The effects of firm loyalty on the benefits of information disclosure will be stronger in push- versus pull-based services.

H8g: The effects of purchase frequency on the benefits of information disclosure will be stronger in push- versus pull-based services.

H8h Purchase frequency on the risks of information disclosure will be stronger in push- versus pull-based services.

2.5 Information control

Another construct that is often mentioned in privacy disclosure literature is information control. In the article by Taylor et al. (2009) is stated that the control of information is critical to the level of consumers’ privacy concern. Perceived information control is the extent to which consumers feel they have control over their personal information and how it is used (Phelps, Nowak and Ferrell 2000), and the power of consumers to decide what is learned about them (Spiekermann 2007). In a firm specific context, Taylor et al (2009) define information control as the extent to which a consumer believes that she or he can influence if and how the firm uses their personal information for marketing purposes. Literature mentions two ways by which firms disclose personal data, explicit and implicit. Explicitly disclosed data comes from consumers actively providing information to the firm, and are therefore explicitly aware of the information exchange. Implicitly disclosed data is information that is exchanged beyond the consumer’s awareness (Taylor et al. 2009).

When consumers perceive a loss of information control in using LBA, it is expected that they will feel vulnerable and therefore will depend on trust in the LBA provider in reducing the privacy risks. It is likely that trust in the LBA provider will be less influential on privacy risks when the perceived information control will be high. Therefore it is expected that information control will moderate both the relationship of firm reputation and previous experience with the privacy calculus.

H9a: The effect of firm reputation on the benefits of information disclosure will be stronger for low information control than for high information control.

H9b: The effect of firm reputation on the risks of information disclosure will be stronger for low information control than for high information control.

H9c The effect of firm satisfaction on the benefits of information disclosure will be stronger for low information control than for high information control.

H9d The effect of firm satisfaction on the risks of information disclosure will be stronger for low information control than for high information control.

H9f The effect of firm loyalty on the risks of information disclosure will be stronger for low information control than for high information control.

H9g The effect of purchase frequency on the benefits of information disclosure will be stronger for low information control than for high information control.

H9h The effect of purchase frequency on the risks of information disclosure will be stronger for low information control than for high information control.

2.6 Control Variables

construct and DOI’s relative advantage and is therefore expected to be similar. Based on this analysis, effort expectancy, performance expectancy and compatibility will be taken into account as characteristics of the LBA.

According to the study of Carter & Belanger (2015), the intention to use e-government services will increase when citizens perceive the service as easy to use. Xu & Gupta (2009) found support for the direct relationship between effort expectancy and intention to use LBS. Therefore the hypothesis studied here will be:

H10: Effort Expectancy of the LBA is negatively related towards intention to use LBA.

As already mentioned before, Xu and Gupta (2009) argue that performance expectancy towards LBS is the degree to which consumers expect that the accessibility and mobility of the LBS reduce the time and effort required to search or access the needed information or service. Adapting this statement to a LBA setting, the higher the expectancy will be, the higher intention to use. Therefore it is hypothesized that:

H11: Performance expectancy of the LBA is positively related towards intention to use LBA.

Rogers (2003) mentions that compatibility is positively related to any innovation adoption decision. When any innovation fits with the lifestyles of the potential adopters, staying in line with their preferences, matching with the similar technologies that these potential adopters may have adopted in the recent past, it becomes seemingly more appealing to them (Kapoor et al. 2014b). Therefore in this study, it is expected that the higher the compatibility of the LBA, the higher the intention to use the LBA.

2.7 Conceptual Framework

Based on the theories of DOI, UTUAT and TAM, the privacy calculus and trustworthiness theory, and the hypotheses described above, the following research model is supposed.

Figure 1. Theoretical framework

3. Methodology

3.1 Experiment design

department stores makes it more likely to introduce in-store LBA than smaller retailers. Because this study is conducted in the Netherlands and Germany, the main Dutch and German department stores were chosen as it is expected to be familiar to every participant of this survey.

3.2 Scale development

All constructs in the proposed model come from measurement scales used in prior studies. Some of them are adapted to fit in the LBA context. The intention to use LBA (INT) comes form the study of Venkatesh et al. (2003) and was measured with questions on whether the respondents were likely to use the LBA in six months. Performance expectancy (PERF) comes from that same study too and was measured with four questions to capture the extent to which an individual would believe that using LBA would reduce his or her time and effort required to search or access the needed information or service. Effort expectancy (EFF) was measured with questions on whether using LBA would be clear, understandable, and easy to use (Venkatesh et al. 2003). Compatibility (COMP) was measured with the same items as Carter & Belanger (2015) did. The privacy calculus was measured as the perceived value of weighing the privacy disclosure benefits against the privacy disclosure risks (Xu & Teo, 2010; Xu et al. 2011). Perceived risk of information disclosure (RISK) was measured with three items from the study of Xu & Teo (2010). Measuring both Personalization benefits (PERS) and Locatability Benefits (LOC) measured perceived Benefits of information disclosure (Xu & Teo 2010). Perceived value of information disclosure (VAL) as outcome of the calculus was measured with three items from the study of Xu et al. (2011). The tree items that measured Reputation of the firm (REP) come from the study by Jarvenpaa et al. (1999). Firm satisfaction (SAT) was measured with items delivered by the study of Jin et al. (2007). Loyalty (LOY) was measured with three items delivered from the study of Salanova et al. (2005). Purchase frequency was measured with the item delivered from Sen & Block (2009).

3.3 Manipulations

Both information delivery mechanism and information control were operationalized by a scenario-based method. This method is chosen since most of the LBS- or LBA-related studies adopted this because of the technological novelty of the construct (Xu & Teo 2010). Sheng et al. (2008) mention that the use of scenarios makes it possible for researchers to study the emerging LBS phenomenon without being constrained by the timing of the study or the state-of-the-art of technology.

Based on the study of Mothersbaugh et al (2012), information control is manipulated using statements appearing as a disclaimer on the introduction page with the information about the V&D in-store App. These statements are related to the ownership of the information, the need for consent for future marketing or the rights to sell or share the information with other companies. This approach is consistent with consumers’ explicitly disclosed information and their control of the marketing usage of their personal information by a firm.

The manipulation of the delivery mechanisms (MECH) is checked with two items delivered by Xu & Teo (2010) and the information control (CONT) manipulation is checked with three items mentioned by Mothersbaugh et al. (2010).

3.4 Procedure

All participants began the survey with some questions about their personal information and characteristics. Next, a cover story about the introduction of a new service by V&D was provided. The story is about a new location-based application that would soon be introduced and the participants were asked to give feedback on this service. After this cover story, all participants were randomly assigned to one of the four different conditions. After the assignment they had to fill in the rest of the survey.

4. Results

4.1 Sample

Variable Category Frequency (percentage)

Gender Male 85 (49,4%)

Female 87 (50,6%)

Age 19 and below 19 (11%)

20-24 96 (55,8%) 25-29 18 (10,5%) 30-34 10 (5,8%) 35-39 2 (1,2%) 40-49 10 (5,9%) 50 and over 17 (9,9%)

Education level Studying in secondary school 5 (2,9%) Secondary school diploma 14 (8,1%)

Bachelor’s Degree 83 (48,3%)

Master’s Degree 67 (39,0%)

Ph.D. 4 (1,7%)

Mobile phone ownership Less than 12 months 2 (1,2%) 12 months to 24 months 6 (3,5%) 25 months to 36 months 7 (4,1%) More than 36 months 157 (91,3%) Mobile internet usage Several times each day 158 (91,9%)

Once per day 3 (1,7%)

Several times each week 3 (1,7%)

Once per week 3 (1,7%)

Less than once per month 1 (0,6%)

Never 3 (2,3%)

Department Store V&D 26 (26,2%)

HEMA 98 (57%) Bijenkorf 20 (11,6) Flying Tiger 5 (2,9%) Galeria Kaufhof 3 (1,7%) Table 1: Sample

4.2 Factor Analysis

the KMO is .829. Also the Bartlett’s test of sphericity was used in the analysis. This test can be used to test the null hypothesis that the variables are uncorrelated in the population. The significance level of the Bartlett’s test of sphericity in the analysis had a value of .000 with 666 degrees of freedom.

Kaiser-Meyer-Olkin Measure of Sampling adequacy

Bartlett’s Test of Sphericity df. Sig

0.829

666

.000 Table 2: KMO and Bartlett’s test of sphericity

This means it is fully significant and that the correlation of the identity matrix differs from the correlations of variables. Communalities measure the percentage of variance in a given variable explained by all the extracted factors. The common rule in checking the communalities is that the extracted values of the communalities should be above .40. Communality is the amount of variance a variable shares with all the other factors. The extracted values of the communalities in this analysis were all above .50, so each variable explains over 50% of the variance and therefore no item has to be excluded (see Appendix A). Based on these three tests, factor analysis in appropriate.

on. The combination of initially separated constructs had some consequences for the conceptual model and the hypotheses as well: the hypotheses H6a, H6b, H8e, H8f, H9e, H9f and H12 were removed and will not be tested anymore. The revised model is visible below (figure 2).

Figure 2: Revised conceptual framework

discriminant validity. Based on both convergent and discriminant validity, one can conclude that the constructs are valid.

The descriptives of each construct are visible in table 3. In this table the means and standard deviations of both the whole sample and the means and standard deviations per condition.

Table 3: Descriptives

4.3 Main effects

To test the main effects of all constructs, the conceptual framework was split into four separate sub-models. The first sub-model is the privacy calculus model measuring the benefits and risks of information disclosure on the perceived value of disclosure (H1 & H2), the second sub-model is the model where the relationships regarding the intention to use were tested (H3, H10 & H11), and the third sub-model tests the influence of Reputation, Positive attitude and Purchase Frequency on Perceived benefits of information disclosure (H4a, H5a & H7a), where the fourth sub-model tests the influence of Reputation, Positive attitude and Purchase Frequency on Perceived risks of information disclosure (H4b, H5b & H7a).

significant negative relationship between the risk and the value of information disclosure (Beta of -.243 with p-value = .007) and therefore the hypothesis can be supported. The model shows a significant positive relationship between the benefits and the value of information disclosure (Beta of .248 with p-value= .035) and therefore the hypothesis H1 and H2 can be supported. Multicollinearity is no issue with both VIF-scores of 1.029.

Dependent Variable: Perceived value of information disclosure

Construct B p VIF

Risk -.243 .007 1.029

Benefits .248 .035 1.029

R2 _{= .077, R}2 _{adjusted = .067}

Table 4

The second sub-model that has been tested is the adjusted TAM model with Perceived value of information disclosure, Perceived performance and Perceived effort as independent variables and Intention to use LBA as dependent variable (see table 5). The model was overall significant (F=21.954 and p-value = .000). The adjusted R Square was .269 meaning that 26.9% of all variance of the intention to use LBA was caused by the three independent variables. The model showed significance for all three relationships. There is a significant positive relationship between the Value of information disclosure and intention to use LBA (Beta of .263 with p-value= .000), a significant positive relationship between Perceived performance and intention to use LBA (standardized beta of .271 with p-value= .002) and a significant positive relationship between Perceived effort and intention to use LBA (Beta of -.359 with p-value= .001). Again, multicollinearity is no issue with VIF-scores of respectively 1.069, 1.263 and 1.205. Therefore all three hypotheses H3, H10 and H11 are supported.

Dependent variable: Intention to use

Construct B p VIF Perceived Value .263 .000 1.069 PerceivedPerformance Perceived Effort .271 -.359 .002 .001 1.263 1.205 R2 _{= .282, R}2 _{adjusted = .269} Table 5

benefits of information disclosure is due to the perceived reputation of the firm and the Positive attitude with the firm. The model showed a positive significant relationship between Positive attitude with the firm and benefits to disclose information (Beta of .258 with p-value= .005). The VIF-score is 1.171, showing no multicollinearity issues. Therefore hypothesis H5a is supported. The model showed no significant link between the perceived reputation of the firm and the benefits of information disclosure (p-value= .146). Therefore this hypothesis H4a is not supported. Purchase frequency neither has a significant relationship with benefits of information disclosure (p-value= .482) and therefore hypothesis H7a is not supported.

Dependent variable: Benefits of information disclosure

Construct B p VIF Positive attitude .258 .005 1.520 Reputation Purchase frequency .116 -.060 .146 .482 1.172 1.338 R2 _{= .088, R}2 _{adjusted = .072} Table 6

The fourth sub-model, testing the same model with risks of the information disclosure instead of the benefits of information disclosure as dependent variable showed no overall significance (F= 1.145 and p-value= .333). Even if the model was significant, output showed there were no significant relations with perceived risk (see table 7). Therefore, none of the hypotheses H4b, H5b and H7b regarding the risk of information disclosure can be supported.

Dependent variable: Perceived risk of information disclosure

Construct B p VIF Positive attitude -.131 .166 1.520 Reputation Purchase frequency .024 .092 .768 .299 1.172 1.338 R2 _{= .016, R}2 _{adjusted = .002} Table 7

Hypotheses Supported Coefficient

H1: Risks à Perceived Value (-) Yes -.243

H2: Benefits à Perceived Value (+) Yes .248

H3: Perceived Value à Intention to use (+) Yes .263

H4a: Reputation à Benefits (+) No

H4b: Reputation à Risk (-) No

H5a: Positive attitude à Benefits (+) Yes .261

H5b: Positive attitude à Risk (-) No

H7a: Purchase Frequency à Benefits (+) No H7b: Purchase Frequency à Risk (-) No

H10: Perceived Effort à Intention to use (-) Yes -.359 H11: Perceived Performance à Intention to use (+) Yes .271 Table 8: Results of testing main effects

4.4 Moderating effects

To test whether the type of delivery mechanism and the amount of information control moderate the relationships between the trustworthiness factors and both benefits and risks of information disclosure, moderation analyses were conducted. Both moderators were tested separately. In the first analysis, type of delivery mechanism is tested (H8a, b, c, d, g, h). In the second analysis, information control is tested (H9a, b, c, d, g, h). To test interaction effects, two dummies were computed. One dummy was computed for Delivery mechanism (DummyDel, where 1 is push-based and 0 is pull-based) and one for Information control (DummyInfo, where 1 is high information control and 0 is low information control). Since Purchase Frequency is measured on another scale (5-point scale) than the other constructs (7-point scale) all constructs were standardized. In the tables this is visible with the z in front of the original variable (e.g. zPositive attitude instead of Positive attitude). In the description of the analyses, they will be mentioned as the original constructs.

Regarding delivery mechanism in the firm trustworthiness à Benefits of information disclosure relationship, a two-step regression analysis showed overall significance for both the model with only main effects (F= 4.140 and p-value= .003) and the model with the interaction effects included (F= 2.516 and p-value= .018). However, besides the significant main effect of Positive attitude on Benefits of information disclosure, no significant effects were found (see table 9). Therefore, Delivery mechanism has no moderating effect on the relationship between the trustworthiness factors and the benefits of information disclosure, so H8a, c and g cannot be supported.

Dependent variable: Standardized score of Benefits of information disclosure Construct B p VIF Main effects ZPositive attitude .193 .131 2.946 ZReputation .200 .069 2.184 ZPurchase frequency -.005 .971 3.180 DummyDel -.189 .203 1.005 Interaction effects DummyDel*ZPositive attitude .122 .510 2.667 DummyDel*ZReputation -.165 .309 2.042 DummyDel*ZPurchase Frequency .101 .561 3.108 R2 _{= .105, R}2 _{adjusted = .067} Table 9

Regarding delivery mechanism in the firm trustworthiness à Risk of information disclosure relationship, a two-step regression analysis showed no overall significance for both the model with only main effects (F= .855 and p-value= .492) and the model with the interaction effects included (F= .753 and p-value= .628). The model self, as visible in table 10, shows no significant effects either. Therefore, H8b, d and h cannot be supported.

Dependent variable: Standardized score of Risks of information disclosure

Based on the two analyses above, one could conclude that the type of delivery mechanism has no significant moderating effect in the relation between trustworthiness of the firm and both risk and benefits of information disclosure.

Regarding Information Control in the firm trustworthiness à Benefits of information disclosure relationship, a two-step regression analysis showed overall significance for both the model with only main effects (F= 4.135 and p-value= .003) and the model with the interaction effects included (F= 2.452 and p-value= .020). However, besides the significant main effect of Positive attitude on Benefits of information disclosure, no significant effects were found (see table 11), so no support for H9a, c and g.

Dependent variable: Standardized score of Benefits of information disclosure

Construct B p VIF Main effects ZPositive attitude .336 .013 3.232 ZReputation .89 .485 2.924 ZPurchase frequency .130 .304 2.892 DummyInfo .029 .846 1.019 Interaction effects DummyInfo*ZPositive attitude -.170 .373 2.923 DummyInfo*ZReputation .013 .938 2.500 DummyInfo*ZPurchase Frequency -.147 .406 3.110 R2 _{= .094, R}2 _{adjusted = .056} Table 11

Dependent variable: Standardized score of Risks of information disclosure Construct B p VIF Main effects ZPositive attitude -.152 .263 3.232 ZReputation .138 .285 2.924 ZPurchase frequency -.099 .443 2.892 DummyInfo -.358 .019 1.019 Interaction effects DummyInfo*ZPositive attitude -.078 .686 2.923 DummyInfo*ZReputation -.279 .103 2.500 DummyInfo*ZPurchase Frequency .078 .662 3.110 R2 _{= .068, R}2 _{adjusted = .028} Table 12

ANOVA analyses and T-tests showed no significant demographic mean differences on the intention to use LBA (see table 13). LSD Post-Hoc tests showed that regarding education level, respondents with a secondary school diploma and respondents with a Master’s degree were scoring significantly higher than respondents with a Ph.D. title. Regarding the perceived reputation of the department stores there were a few significant differences too. V&D had a significant lower reputation than both HEMA and Bijenkorf. Construct F Sig. Department Store 1.076 .371 Gender .891 .650 Age 1.036 .427 Education level .829 .742

Mobile phone ownership .934 .582

Mobile internet usage 1.035 .429

Table 13: ANOVA analyses on intention to use

5. Discussion

5.1 Theoretical implications

Figure 3: conceptual framework including coefficients.

The results suggested that the privacy calculus model worked as expected. Whereas the benefits had a positive influence on the perceived value of disclosing personal information, risk has a negative influence. Interesting is the fact that both benefits and risk explain a low level of variance in the perceived value. This implicates that other factors play an important role too in explaining the perceived value of information disclosure. Other studies mention the general privacy concern influencing the intention to use new technology (e.g. Xu & Gupta 2009). This general privacy concern based on the concern for informational privacy (CFIP) by Smith et al. (1996) is not explicitly taken into account in measuring the perceived value of information disclosure, but is measured implicitly in the risk of disclosure items. This fact could explain the low level of variance in the privacy calculus model.

Regarding the intention to use LBA, all results were as expected. Perceived value, perceived performance and perceived effort have a significant influence on the intention to use. These findings confirm that besides the often-mentioned perceived effort and perceived performance, privacy calculus plays a significant role in consumers’ usage intention and should be taken into consideration when measuring the intention to use LBA.

Positive attitude, it is interesting to see that Positive attitude has no significant relationship with the risks of information disclosure. So, whether a consumer is satisfied and loyal with a firm does not influence the perceived risk of information disclosure. One possible explanation for this could be the fact that the benefits of information disclosure can be directly related towards the service provider, whereas the risks of information disclosure cannot. External factors like unauthorized access or errors (Smith et al. 1996) can increase perceived risk regardless of Positive attitude with the firm. Another finding is that a firm’s perceived reputation has no significant influence on both benefits and risks of information disclosure, although previous literature states otherwise (e.g. Beldad et al. 2011). This is possibly due to the fact that that literature stems from the area of governmental e-services acceptance. Carter & Belanger (2015) state that although there are many similarities between government and e-commerce, there is a clear difference between them in terms of accountability. Unlike e-government services, e-commerce does not have the accountability to act in the best interest of the public. Therefore it is possible that although the reputation of a firm is good, there is no influence on the benefits and the risks of information disclosure. Purchase frequency has no significant effect either.

The results regarding the moderators Delivery mechanism and Information control do not confirm the findings from previous studies. Both moderators have no significant effects on the relationship between the trustworthiness factors and the privacy calculus. Regarding the delivery mechanism, there is no significant difference in benefits and risks between push- and pull-based LBA. Although Xu et al (2011), Barkhuus & Dey (2003) and Xu & Teo (2010) found a difference; this is not the case in this study. A possible explanation is the novelty of the LBA concept among consumers. It could be that since most respondents read about LBA in this survey for the first time, without knowing that there are several types of LBA. Therefore the focus of the respondent might be on the LBA concept as a whole instead of the aspects of the push-versus pull-type. That could explain the non-significant difference. Another possible explanation is the fact that the previously mentioned studies that found significant differences used LBA based on a SMS- or Dial-service. They did not use mobile applications as used in this study but send their advertisements via SMS. It is possible that consumers perceive the difference between dialling a number for advertisements in the neighbourhood (pull-based LBA) and automatically receiving offers via SMS (push-based LBA) as bigger than the difference between pushing a button in a mobile application (pull-based LBA) and automatically receive advertisements in the application (push-based LBA). This difference could be explained by the fact that nowadays many consumers have several location-tracking applications on their mobile phones already (e.g. Facebook, weather-forecast applications, dating applications and running-applications), so the difference between the push- and pull-type of delivery mechanism would be perceived as smaller.

findings of Taylor et al. (2009), who stated that information control had a significant moderating effect on the relation between personalization and privacy concerns. Compared to that study, this study focused on the perceived trustworthiness of the firm instead of personalization of the advertisement. That could explain the non-significant moderating effect. Legislation could be an explanation for this. Since service providers are not allowed by law to share data without the consent of the consumer, consumers could perceive information control as irrelevant since they know that the law protects their personal information.

There was a significant main effect on the risks of information disclosure though. The more information control, the lower the risks. This seems logical since the more control a consumer has about his own information, the lower the chance that his information will be used for purposes he does not know or has to say anything about. Extra analysis showed a significant main effect of information control on the intention to use LBA. This is interesting since it shows that there is an effect, but not at the expected place. Even though information control does not serve as a moderator, it is a factor explaining the intention to use LBA.

The manipulations that were used in this study could explain the insignificant moderating effects of both delivery mechanism and information control. First, this could be due to the way both moderators were manipulated in the study. Although both manipulations were adopted from other studies, these manipulations were adjusted to meet with the latest technologies like mobile Internet (e.g. using an application instead of dialling a phone-number). These adjustments could have changed the manipulations in such a way that the new manipulations caused the non-existence of the moderating effects. Finally, there was a forced response in the survey on the manipulation checks. Regarding the checks there was only one right answer. When respondents answered the manipulation check wrong, they received a statement that they had to read the story again and to answer the yes/no question again. This could have forced respondents to fill in the forced answer without really knowing why the answer was right. That could have resulted in people not recognizing in what condition they were, what consequently could have led to insignificant differences between the conditions.

5.2 Managerial Implications

satisfied customers. They perceive the highest value of information disclosure, leading to higher usage intention. Based on this study, information control has a direct effect on the risks of information disclosure in the privacy calculus. It is important for managers to know that information control is important to consumers. The higher the information control, the lower the risks, leading to higher perceived value of information disclosure and finally a higher intention to use LBA. Besides the effect on the risks of information disclosure, legal aspects of information control play a role too. According to the Dutch Centre of Data Protection (CBP), it is forbidden to use tracking-data without the consumer’s consent. Besides the legal issues, information control does have a direct effect on intention to use too and should therefore treated with caution.

5.3 Limitations

6. References

Adomavicius, Gediminas, and Alexander Tuzhilin (2005), "Personalization technologies: a process-oriented perspective." Communications of the ACM 48, 83-90.

Alhabeeb, Musaddak Jamael (2007), "On consumer trust and product loyalty." International Journal of Consumer Studies Vol. 31 (6), 609-612.

Barkhuus, Louise, and Anind K. Dey (2003), "Location-Based Services for Mobile Telephony: a Study of Users' Privacy Concerns." Interact. Vol. 3.

Barnes, Stuart J. (2003), “Known by the network: The emergence of location-based mobile commerce,” in Advances in mobile commerce technologies, Ee-Peng Lim and Keng Siau: Idea Group Publishing, 172-189

Beke, Frank (Forthcoming), Privacy calculus measurement scale

Beldad, Ardion, et al. (2012), "A cue or two and I'll trust you: Determinants of trust in government organizations in terms of their processing and usage of citizens' personal information disclosed online." Government Information Quarterly 29, 41-49.

Benamati, John S. et al. (2010), “Clarifying the integration of trust and TAM in e-commerce environments: implications for systems design and management”, IEEE Transactions on Engineering Management, Vol. 57.3, 380-93.

Berg Insight, (2015, August 31), available at: http://www.berginsight.com/News.aspx

Bruner, Gordon C., and Anand Kumar (2007), "Attitude toward location-based advertising." Journal of Interactive Advertising, Vol. 7.2, 3-15.

Carter, Lemuria, and France Bélanger (2005), "The utilization of e­government services: citizen trust, innovation and acceptance factors", Information systems journal, vol. 15.1, 5-25.

Clarke, Roger (2001), “Person location and person tracking: Technologies, risks and policy implications”. Information Technology & People, Vol. 14.2, 206-231.

Chen, Kai-Ying. & Meng-Lin Chang, (2013), User acceptance of ‘near field communication’ mobile phone service: an investigation based on the ‘unified theory of acceptance and use of technology’ model. The Service Industries Journal, Vol. 33(6), 609–623.

Culnan, Mary J., and Robert J. Bies (2003), "Consumer privacy: Balancing economic and justice considerations." Journal of social issues, Vol. 59.2, 323-342.

Davis, Fred D., Richard P. Bagozzi, and Paul R. Warshaw (1989), "User acceptance of computer technology: a comparison of two theoretical models." Management science, Vol. 35.8, 982-1003.

Dhar, Subhankar, and Upkar Varshney (2011), "Challenges and business models for mobile location-based services and advertising." Communications of the ACM, Vol. 54.5, 121-128.

Doney, Patricia M. and Joseph, P. Cannon (1997), “An examination of the nature of trust in buyer-sellerrelationships”, Journal of Marketing, Vol. 61.2, 33-51.

Edwards, Steven M., Hairong Li, and Joo-Hyun Lee (2002), “Forced exposure and psychological reactance: antecedents and consequences of the perceived intrusiveness of pop-up ads”, Journal of Advertising, Vol. 31(3), 83-95.

eMarketer,(2014, December 11), available at

http://www.emarketer.com/newsroom/index.php/emarketer-2-billion-consumers-worldwide-smartphones-2016/

Estelami, Hooman, Dhruv Grewal, andAnne L. Roggeveen (2014), “The effect of retailer reputation and response on post purchase consumer reactions to price-matching guarantees”, MSI Reports, Working Paper No. 04-003, 27-47.

Garbarino, Ellen, and Mark S. Johnson (1999), "The different roles of satisfaction, trust, and commitment in customer relationships." the Journal of Marketing, 70-87.

Geyskens, Inge, et al. (1996), "The effects of trust and interdependence on relationship commitment: A trans-Atlantic study." International Journal of research in marketing, Vol. 13(4), 303-317.

Greengard, Samuel (2012), "Advertising gets personal." Communications of the ACM, Vol. 55(8), 18-20.

Hagel, John, and Jeffrey F. Rayport (1997) "The coming battle for customer information." McKinsey Quarterly, 64-77.

Huth Carly L. (2015), ”A privacy primer on beacon technology”, Journal Of Internet Law, Vol. 18(8), 21-25.

Jarvenpaa, Sirkka L., Noam Tractinsky, and Lauri Saarinen (1999), "Consumer trust in an internet store: a cross­cultural validation", Journal of Computer­Mediated Communication, Vol. 5(2), 0-0.

Kapoor, Kawaljeet Kaur, Yogesh K. Dwivedi, and Michael D. Williams (2014a) "Rogers’ innovation adoption attributes: a systematic review and synthesis of existing research." Information Systems Management, Vol. 31(1), 74-91.

Kapoor, Kawaljeet Kaur, Yogesh K. Dwivedi, and Michael D. Williams (2014b), "Examining consumer acceptance of green innovations using innovation characteristics: A conceptual approach." International Journal of Technology Management & Sustainable Development, Vol. 13(2), 135-160.

Kim, Hee-Woong, Hock Chuan Chan, and Sumeet Gupta (2007) "Value-based adoption of mobile internet: an empirical investigation." Decision Support Systems, Vol 43 (1), 111-126.

Kim, Dan J. (2008), “Self-perception-based versus transference-based trust determinants in computer-mediated transactions: a cross-cultural comparison study”, Journal of Management Information Systems, Vol. 24 (4), 13-45.

Lin, Trisha TC, Fernando Paragas, and John Robert Bautista (2015), “Determinants of mobile consumers’ perceived value of location-based mobile advertising and user responses”, International Journal of mobile communications. (Pre-publication version).

McMillan, Sally J., and Jang-Sun Hwang (2002) "Measures of perceived interactivity: An exploration of the role of direction of communication, user control, and time in shaping perceptions of interactivity." Journal of Advertising, Vol. 32(3), 29-42.

Mothersbaugh, David L., et al. "Disclosure antecedents in an online service context: the role of sensitivity of information." Journal of Service Research

Okazaki, Shintaro, et al. (2012), "Privacy concerns in quick response code mobile promotion: The role of social anxiety and situational involvement." International Journal of Electronic Commerce, Vol. 16 (4), 91-120.

Phelps, Joseph, Glen Nowak, and Elizabeth Ferrell. "Privacy concerns and consumer willingness to provide personal information." Journal of Public Policy & Marketing Vol. 19 (1), 27-41.

Rogers Everett, M. (1995), Diffusion of innovations, Fourth edition. New York: The Free Press.

Rogers, Everett, M. (2003). Diffusion of innovations. Fifth edition. New York: The Free Press.

Salanova, Marisa, Jose Maria Peiro, and Sonia Agut (2005), "Linking Organizational Resources and Work Engagement to Employee Performance and Customer Loyalty: The Mediation of Service Climate", Journal of Applied Psychology, vol. 90, 1217-1227.

Sen, Sankar, and Lauren G. Block (2009), "“Why My Mother Never Threw Anything Out”: The Effect of Product Freshness on Consumption", Journal of Consumer Research, Vol. 36, 47-55.

Sheng, Hong, Fiona Fui-Hoon Nah, and Keng Siau (2008), "An experimental study on ubiquitous commerce adoption: Impact of personalization and privacy concerns." Journal of the Association for Information Systems, Vol. 9 (6), 15

Slade, Emma L., et al (2015) "Modeling Consumers’ Adoption Intentions of Remote Mobile Payments in the United Kingdom: Extending UTAUT with Innovativeness, Risk, and Trust." Psychology & Marketing, Vol. 32 (8), 860-873.

Smith, H. Jeff, Tamara Dinev, and Heng Xu (2011), "Information privacy research: an interdisciplinary review." MIS quarterly, Vol. 35.4, 989-1016.

Spiekermann, Sarah (2007), Privacy enhancing technologies for RFID in retail-an empirical investigation, Springer Berlin Heidelberg.

Sundar, Shyam S. and Sampada S. Marathe (2006), “Is it tailoring or is it agency? Unpacking the psychological appeal of customized news”, the 89th Annual Convention of the Association for Education in Journalism and Mass Communication, San Francisco, CA.

Sztompka, Piotr (1999), Trust: A sociological theory: Cambridge University Press.

Taylor, David G., Donna F. Davis, and Ravi Jillapalli (2009), "Privacy concern and online personalization: The moderating effects of information control and compensation." Electronic Commerce Research, Vol. 9 (3), 203-223.

Tekawade, Abhijeet, et al. (2013), "Mobile Tracking Application for Locating Friends using LBS." International Journal of Innovative Research in Computer and Communication Engineering, Vol. 1(2).

Van Doorn, Jenny, and Janny C. Hoekstra (2013), "Customization of online advertising: The role of intrusiveness." Marketing Letters, Vol. 24(4), 339-351,

Venkatesh, Viswanath, et al. (2003), "User acceptance of information technology: Toward a unified view." MIS quarterly, 425-478.

Xu, Heng, et al. (2009) "The role of push-pull technology in privacy calculus: the case of location-based services." Journal of Management Information Systems, Vol. 26(3), 135-174.

Xu, Heng, et al. (2011), "The personalization privacy paradox: An exploratory study of decision-making process for location-aware marketing", Decision Support Systems Vol. 51(1), 42-52.

Xu, Heng, and Sumeet Gupta (2009) "The effects of privacy concerns and personal innovativeness on potential and experienced customers’ adoption of location-based services." Electronic Markets, Vol. 19(2-3), 137-149.

Zhou, Tao (2011), "The impact of privacy concern on user adoption of location-based services." Industrial Management & Data Systems, Vol. 111(2), 212-226.

7. Appendix

Appendix A: Communalities

LOY1 I will return to the department store in the future 1,000 .652 LOY2 I will recommend the department store to others 1,000 .690 SAT1 I am satisfied with the assortment of the department store 1,000 .690 SAT2 am overall satisfied with the department store 1,000 .748 SAT3 I am satisfied with the purchase experience at the department store 1,000 .628

REP2 The department store has a good reputation 1,000 .879

REP3 The department store has a bad reputation 1,000 .877

PERF1 the in-store app reduces my searching time to find the information that I need 1,000 .785 PERF2 the in-store app reduces my searching efforts to find the information that I need 1,000 .800 PERF3With the in-store app, I can quickly access the information that I need 1,000 .845 PERF4 With the in-store app, I can easily access the information that I need 1,000 .825 EFF1 I expect that my interaction with the in-store app would be clear and

understandable

1,000 .725

EFF2 I expect that I would find the in-store app easy to use 1,000 .720 EFF3 Learning to use the in-store app will be easy for me 1,000 .677 INT1 I intend to download the in-store app in the next six months 1,000 .906 INT2 I intend to use the in-store app in the next six months 1,000 .919 INT3 I predict I will use the -store app in the next six months 1,000 .929 INT4 I plan to use in-store app in the next six months 1,000 .900 COMP1 I think using the in-store app would fit well with the way that I like to gather

information from the department store

1,000 .686

COMP2 I think using the in-store app would fit well with the way that I like to interact with the department store

1,000 .731

COMP3 Using the in-store app to interact with the department store would fit into my lifestyle

1,000 .603

LOC1 With the application, I am able to get the up-to-date information/services whenever I need to.

1,000 .671

LOC2 With the application, I am able to access the relevant information/services at the right place

1,000 .646

LOC4 With the application, I am able to access the relevant information/services wherever I want to.

1,000 .580

PER1 The application can provide me with personalized services tailored to my activity context.

1,000 .631

PER2 The application can provide me with more relevant information tailored to my preferences or personal interests.

1,000 .706

PER3 application can provide me with the kind of information or service that I might like.

1,000 .617

RISK1 Providing the department store with my personal information would involve many unexpected problems.

1,000 .638

RISK2 It would be risky to disclose my personal information to the department store 1,000 .843 RISK3 There would be high potential for loss in disclosing my personal information to

the department store

1,000 .756

VAL1 I think my benefits gained from the use of the application canoffset the risks of my information disclosure.

1,000 .699

VAL3 I think the risk of my information will be greater than the benefits gained from the use of the application.

1,000 .557