Abstract
Many e-stores adopt personalized recommender systems to provide service for the customers nowadays, which they can rely on to predict customers’ preferences based on the detailed individual customer information. Customers got better services provided by the personalized recommender systems. However, customers also concerned that the websites may steal, misuse or sell their information to a third party. Such situation causes the “personalization-privacy paradox”. This study proposed a research model based on the privacy calculus theory to explore how the customers make decision between personalized service and privacy concern. An online survey was conducted to collect empirical data in order to test our research model. The results of PLS analysis indicate that personalized service is positively affects perceived benefit. Both information sensitivity and privacy concern positively affects perceived risk. However, when customers with low information sensitivity and low privacy concern, they are less likely to evaluate associated risks. Perceived value is influenced by perceived benefit and perceived risk and in term, affects customers’ willingness to provide personal information. The findings of this study provide implications for both researchers and practitioners of using personalized recommender systems.
You have full access to this open access chapter, Download conference paper PDF
Similar content being viewed by others
Keywords
1 Introduction
Recently, the internet and e-commerce booming provide convenient life for the customers. However, large information providing from numbers of website causes customers face the difficulties of information processing when conducting purchase decision. The information overloading also results in the customers hard to make a decision between various options. Personalized recommender systems have been pay attention to provide more relevant information for the customers to facilities their purchase decision. Such personalized systems provide the benefits for customers to search personalized products and services in efficient ways. The companies also benefit from such system which predict consumer behavior more accurately and bring in higher sales volume and revenue.
Previous studies claimed that the success of personalized recommender systems rely on the companies’ data collecting and processing capabilities. In addition, customers’ willingness to share their personal information and obtain personalized services is vital to the development of personalized recommender systems [6]. In fact, companies need many information to be collected from the customers to provide personalized services for customers, however, the customers often have low willingness to share such personalized information with companies [2, 26]. The privacy issue is the largest concern of the customers to share the personalized information because a lot of personalized individuals information are collected, obtained, and sold inappropriately to the third parties. The privacy concern and the worry about the personal data are used in illegal ways cause the customers prefer to protect their personal data.
Such issue is largely drawn the attentions from previous studies [11, 28, 32]. Personalization-privacy paradox was incurred because the customers tend to get benefits from the personalized services which provided by the websites but are also afraid of their personal privacy are invaded [27]. Since personalization-privacy paradox is a very interesting phenomenon and an important research issue in personalized service, many previous studies have investigated this issue and provided insightful implications [27, 42]. However, prior studies are more focus on the causal effects among the related variables, but less studies focused on examining the influence of sensitivity in different level of personalized service and privacy concern on the willingness of sharing personal information. Hence, the purpose of this study is to explore the sensitive relationships between personalized service and privacy concern while an individual is using a personalized recommender system. The following sections are organized as follows. The theoretical foundations and hypotheses are proposed in the Sect. 2. In the Sect. 3, the experimental scenario design and empirical survey are described. Then, the results of statistical analysis are discussed in the Sect. 4. Finally, we make a brief conclusion in the Sect. 5.
2 Theoretical Foundations and Hypotheses
Privacy calculus theory claimed that the individuals’ intention of disclosing personal information will depend on comparing perceived risks and anticipated benefits [12, 16]. Personal information is viewed as a tradable good between the customers and the websites. The consumer evaluate the benefit of providing information such as the privacy information and personal data and the risk of the concern of privacy invasion. Individuals will conduct a privacy calculation, which the customers compare the benefits and the risks of providing personal information to the websites and then decide the scope and the level of the privacy information they wants to expose.
Privacy calculation also is viewed as a cognitive process which indicates that the individuals make a decision based on measuring the potential cost and benefit of disclosing private information to others [5, 19]. If the foreseeable benefit is larger than (or equal to) the potential risk, the individuals will incline to disclose their personal information to exchange their desired products/services, such as financial incentives and personalized services [9, 13].
According to the prior research, the privacy calculation has been viewed as one kind of functional exchange between the customers and the companies. The customers will calculate how much higher quality services they can earn from the website and decide the level of the personal information that they are willing to offer [9]. The individual will evaluate the expected benefits and privacy risks of offering such private information to other people [22]. Once the individual perceived the benefits are higher than the risks of providing the information, they will have the higher willingness of sharing their personal information and preferences to others [8]. Hence, we propose hypothesis H1 as follows,
-
H1: The willingness of providing personal information will be positively affected by perceived value of personalized service
Previous studies found that when the consumer realized the benefit can generate from disclosing private information, they tend to behave lower concern of privacy issues. At this situation, the consumers prefer to release their individual information to exchange the potential benefits [12, 43]. Studies also found the evidences that the consumers have higher willingness to disclose their individual information if they can gain offerings or discounts from providing personal information [35]. In this study, perceived benefit is defined as the benefits of personalized service that the individuals can earn through disclosing their personal information. Thus, the hypothesis H2 is proposed here:
-
H2: Perceived value of personalized service will be positively affected by perceived benefit of information disclosure.
Privacy calculation theory also proposed that privacy risk will affect the likelihood of consumers’ privacy intention and consumer behavior [21, 32]. For example, the uncertainty of using the Internet will force the consumers feel hesitate to disclose their personal information. The tendency of disclosing information will rely on the evaluation of risk and benefit. The perceived uncertainty results in the customers perceived risks of disclosing private information in inappropriately and unauthorized ways [24], such as internal personnel’s use, intentionally access and selling to the third parties [15]. Such unauthorized information disclosing let the consumers suffer from possible risks because of exposing personal privacy under public without noticed. Therefore, we hypothesize that
-
H3: Perceived value of personalized service will be negatively affected by perceived risk of information disclosure.
Personalization provides the customized service to the customer, which let the customers can enjoy the services based on their personal preference and needs. Personalization also lets the consumers experience as serving by a customized salesperson during the purchase process [25]. Customer loyalty also is developed through satisfying the customers’ personalized needs [18]. Recently, recommender system provides the personalized recommendations and information to the customers in helping their purchase decision in e-commerce websites [30]. Recommender systems can provide the relevant information and suggestions to the customers based on their preference and customer behaviors [1]. According to the prior purchasing history, the system recommends the products/services that may be interested by the customers. According, the customers can get accurate information and shorten the searching time in purchasing process. The personalization bring the benefits to the customers in reducing transaction cost and searching cost through providing useful information to the customers. The system are thereby can earn the loyalty and recognition from the customers. Therefore, we propose hypothesis H4 as follows,
-
H4: Perceived benefit will be positively affected by personalization.
The risk of disclosing information depends on the types of information. The higher sensitive information represents the higher level of closeness of information to the individuals. Prior studies pointed out that higher closeness of information disclosing will cause the higher risk of personal losing [23]. Consumers prefer to disclose their demographic information rather than the sensitive information such as financial or identification information [29]. Moreover, consumer will perceive higher risk if they are asked to provide more sensitive information [29, 31, 37]. Such perceived risk will cause the consumers have more negative attitude and intention toward disclosing their personal information [24]. Therefore, the hypothesis H5 is proposed here:
-
H5: Perceived risk of information disclosure is positively affected by the sensitivity of information.
Warren and Brandeis [38] claimed that privacy is related to the individuals’ rights and capabilities to control their owned information, to occupy their personal space [4, 39], to have their personal life [3], and to control the personal message [39]. Flushing internet draws scholars’ attention from avoiding personal privacy invasion to learning the ways to protect personal information. Scholars suggest that information privacy should address the importance of how the private information to be used and to be transferred to users [40].
Consumers concern about the possible loss and risks after disclosing their private information [32]. In particular, the development of technology in monitoring and in searching information causes the consumers have the higher awareness that the personal privacy is under-protected and is invaded [7, 36]. People concern about their personal information is being misused [44]. In addition, unauthorized information assessed by the other organizations is also a vital concern to the customers [10]. Thus, the consumers’ fell the difficulties of controlling the information after disclosing private information [11].
Studies also found that the privacy invaded causes consumers perceived risks to provide personal information in the future purchase. The negative perception and experience will be incurred to the customers, which causes the customers incline to provide the information or correct information to the website. Even more, some customers will claim to the organization or the official authorities [34]. Prior studies also found evidence that the perceived risk of information disclosure is positively associated with privacy concern [12]. The higher risks perceived by the consumers will lead the customers have higher concern of information disclosure [17]. Hence, we hypothesize that:
-
H6: Perceived risk of information disclosure will be positively affected by privacy concern.
3 Research Method
A survey questionnaire was developed to test the proposed hypotheses. The conceptual definition and the source of the measurement items for each construct are listed in Table 1. All constructs were measured using items equivalent to those used by previous studies. This study adopted the Likert Scales, allowing participants to choose one of seven levels of agreement with anchors ranging from 1 (strongly disagree) to 7 (strongly agree) with the exception of the personalization construct which was measured by the perception of personalized degree.
In order to explore the sensitive reaction to varied degrees of personalized service and personal information disclosure, three scenarios were designed as Table 2. Then three type questionnaires were designed to collect empirical data. Each questionnaire includes three parts. In part 1, the purpose of survey and one of three scenario, i.e., S1, S2, or S3, was described. In part 2, all items of the research constructs were designed by seven-Likert scales based on previous studies. In part 3, some questions related to respondents’ profile and shopping experience were designed.
After questionnaire pretesting and revision, we conducted an empirical survey in May 2017 and gave 2 US dollars e-coupon to respondents who had filled the questionnaire successfully as incentive. Finally, 475 questionnaires were collected but 41 questionnaires were invalid. The valid amounts of three scenarios are S1:148, S2:142, S3:144, respectively. There were more females (67.7%) than males (32.3%). The majority are students (68.7%). Most of them averagely did online shopping once per month (46.1%) while some of them averagely purchased 2 or 3 times per month (35.9%). Furthermore, Most of them averagely spent no more than 35 USD (48.8%).
4 Preliminary Findings and Further Analysis
Partial Least Squares (PLS) was used to test the research model and six hypotheses because PLS is more appropriate to measure research models which are in the early development stages and have not yet been extensively tested. This study utilized the SmartPLSFootnote 1 software to conduct PLS analysis. Reliability and discriminant validity were tested before the research model was tested.
As Table 3 shows, the Cronbach’s Alpha of each construct was higher than 0.74. The composite reliability of each construct is higher than 0.85. Notably, each square root of AVE is higher than 0.817 and also higher than the inter-construct correlation coefficients. Hence, the indicators of reliability and validity for the measurement model (i.e., item reliability, convergent validity, and discriminant validity) are all acceptable. Then PLS was used to assess the structural model. All path coefficients and explained variances for the model (all sample) are shown in Fig. 1.
The preliminary results of PLS analysis with all sample show that the explanatory power (R2) of willingness of providing personal information is 25.6%. The path coefficient from perceived value to willingness of providing personal information is 0.506 (p < 0.001), which means that H1 is significantly supported. In addition, there is a significantly positive association between perceived benefit and perceived value (b = 0.515, p < 0.001), while the association between perceived risk and perceived value is significantly negative (b = −0.164, p < 0.001). Hence, both H2 and H3 are significantly supported. The explanatory power (R2) of perceived value is 30.4%. There is also a significantly positive association between personalization and perceived benefit. Hence, H4 is significantly supported and 42.8% of the variance of perceived benefit can be explained by providing personalization. Finally, we found that perceived risk was significantly affected by sensitivity of information and privacy concern, R2 = 68.7% means that the explanatory power is high. The test results of H1–H6 are listed in Table 4.
In addition, we found that when customers are asked for data with low information sensitivity and low privacy concern, they are less likely to evaluate associated risks by performing a cost-benefit analysis. Since three scenarios were designed in this study, further analysis is required to explore the sensitive relationships between perceived benefit and perceived risk.
5 Conclusion
This study proposes a research model based on the privacy calculus theory to explore the sensitive relationships between personalized service and privacy concern. i.e., how the users react when they perceived the decision dilemma between exposing the personal privacy (perceived risk) and earning the benefit of personalization (perceived benefit), and how the customer perceived value of personalization contribute to their willingness to provide personal information. Accordingly, this study aim to investigate the impact of personalization on perceived benefit of personalized service with information disclosure. In addition, the influence of sensitivity of information and privacy concern on perceived risk of information disclosure. Moreover, we further examined the influence of perceived benefit of personalized service with information disclosure and perceived risk of information disclosure on perceived value of information disclosure. Finally, we tested the relationship between perceived value of information disclosure and the willingness of providing personal information.
We collect the empirical data through an online survey to test our proposed research model. The results of PLS analysis indicate that personalized service positively affects perceived benefit. In addition, both of information sensitivity and privacy concern positively affect perceived risk. The results also indicate perceived value positively affected by perceived benefit but negatively affected by perceived risk. Moreover, perceived value will increase the customers’ willingness to provide personal information. The findings of this study provide implications for both researchers and practitioners of personalized recommender systems.
Notes
- 1.
C.M. Ringle, S. Wende and A. Will, SmartPLS 2.0 (M3) Beta, Hamburg 2005.
References
Adomavicius, G., Tuzhilin, A.: Personalization technologies: a process-oriented perspective. Commun. ACM 48, 83–90 (2005)
Adomavicius, G., Tuzhilin, A.: Toward the next generation of recommender systems: a survey of the state-of-the-art and possible extensions. IEEE Trans. Knowl. Data Eng. 17, 734–749 (2005)
Allen, A.L.: Uneasy Access: Privacy for Women in a Free Society. Rowman and Littlefield, Totowa (1988)
Altman, I.: The Environment and Social Behavior: Privacy, Personal Space, Territory, and Crowding. Brooks/Cole, Monterey (1975)
Angst, C.M., Agarwal, R.: Adoption of electronic health records in the presence of privacy concerns: the elaboration likelihood model and individual persuasion. MIS Q. 33, 339–370 (2009)
Chellappa, R.K., Sin, R.G.: Personalization versus privacy: an empirical examination of the online consumer’s dilemma. Inf. Technol. Manag. 6, 181–202 (2005)
Culnan, M.J.: “How did they get my name?” An exploratory investigation of consumer attitudes toward secondary information use. MIS Q. 17, 341–363 (1993)
Culnan, M.J., Armstrong, P.K.: Information privacy concerns, procedural fairness, and impersonal trust: an empirical investigation. Organ. Sci. 10, 104–115 (1999)
Culnan, M.J., Bies, R.J.: Consumer privacy: balancing economic and justice considerations. J. Soc. Issues 59, 323–342 (2003)
Culnan, M.J., Williams, C.C.: How ethics can enhance organizational privacy: lessons from the choicepoint and TJX data breaches. MIS Q. 33, 673–687 (2009)
Dinev, T., Hart, P.: Internet privacy concerns and their antecedents- measurement validity and a regression model. Behav. Inf. Technol. 23, 413–422 (2004)
Dinev, T., Hart, P.: An extended privacy calculus model for e-commerce transactions. Inf. Syst. Res. 17, 61–80 (2006)
Dinev, T., Hart, P., Mullen, M.R.: Internet privacy concerns and beliefs about government surveillance–an empirical investigation. J. Strat. Inf. Syst. 17, 214–233 (2008)
Dinev, T., Xu, H., Smith, J.H., Hart, P.: Information privacy and correlates: an empirical attempt to bridge and distinguish privacy-related concepts. Eur. J. Inf. Syst. 22, 295–316 (2013)
Featherman, M.S., Pavlou, P.A.: Predicting e-services adoption: a perceived risk facets perspective. Int. J. Hum.-Comput. Stud. 59, 451–474 (2003)
Hui, K.L., Teo, H.H., Lee, S.Y.T.: The value of privacy assurance: an exploratory field experiment. MIS Q. 31, 19–33 (2007)
Laufer, R.S., Wolfe, M.: Privacy as a concept and a social issue: a multidimensional developmental theory. J. Soc. Issues 33, 22–42 (1977)
Leppaniemi, M., Karjaluoto, H.: Factors influencing consumers’ willingness to accept mobile advertising: a conceptual model. Int. J. Mob. Commun. 3, 197–213 (2005)
Li, H., Sarathy, R., Xu, H.: Understanding situational online information disclosure as a privacy calculus. J. Comput. Inf. Syst. 51, 62–71 (2010)
Li, H., Sarathy, R., Xu, H.: The role of affect and cognition on online consumer’s decision to disclose personal information to unfamiliar online vendors. Decis. Support Syst. 51, 434–445 (2011)
Li, Y.: Empirical studies on online information privacy concerns: literature review and an integrative framework. Commun. Assoc. Inf. Syst. 28, 453–496 (2011)
Li, Y.: Theories in online information privacy research: a critical review and an integrated framework. Decis. Support Syst. 54, 471–481 (2012)
Lwin, M., Wirtz, J., Williams, J.D.: Consumer online privacy concerns and responses: a power–responsibility equilibrium perspective. J. Acad. Mark. Sci. 35, 572–585 (2007)
Malhotra, N.K., Kim, S.S., Agarwal, J.: Internet users’ information privacy concerns (IUIPC): the construct, the scale, and a causal model. Inf. Syst. Res. 15, 336–355 (2004)
Mittal, B., Lassar, W.M.: The role of personalization in service encounters. J. Retail. 72, 95–109 (1996)
Murthi, B.P.S., Sarkar, S.: The role of the management sciences in research on personalization. Manag. Sci. 49, 1344–1362 (2003)
Norberg, P.A., Horne, D.R., Horne, D.A.: The privacy paradox: personal information disclosure intentions versus behaviors. J. Consum. Aff. 41, 100–126 (2007)
Pavlou, P.A.: State of the information privacy literature: where are we now and where should we go? MIS Q. 35, 977–988 (2011)
Phelps, J., Nowak, G., Ferrell, E.: Privacy concerns and consumer willingness to provide personal information. J. Public Policy Mark. 19, 27–41 (2000)
Resnick, P., Varian, H.R.: Recommender systems. Commun. ACM 40, 56–58 (1997)
Sheehan, K.B., Hoy, M.G.: Dimensions of privacy concern among online consumers. J. Public Policy Mark. 19, 62–73 (2000)
Smith, H.J., Dinev, T., Xu, H.: Information privacy research: an interdisciplinary review. MIS Q. 35, 989–1016 (2011)
Smith, H.J., Milberg, S.J., Burke, S.J.: Information privacy: measuring individuals’ concerns about organizational practices. MIS Q. 20, 167–196 (1996)
Son, J.Y., Kim, S.S.: Internet users’ information privacy-protective responses: a taxonomy and a nomological model. MIS Q. 32, 503–529 (2008)
Sultan, F., Rohm, A.J., Gao, T.T.: Factors influencing consumer acceptance of mobile marketing: a two-country study of youth markets. J. Interact. Mark. 23, 308–320 (2009)
Wang, H., Lee, M.K., Wang, C.: Consumer privacy concerns about internet marketing. Commun. ACM 41, 63–70 (1998)
Wang, P., Petrison, L.A.: Direct marketing activities and personal privacy: a consumer survey. J. Direct Mark. 7, 7–19 (1993)
Warren, S.D., Brandeis, L.D.: The right to privacy. Harv. Law Rev. 4, 193–220 (1890)
Westin, A.F.: Privacy and Freedom. Athenaum, New York (1967)
Westin, A.F.: Social and political dimensions of privacy. J. Soc. Issues 59, 431–453 (2003)
Wilson, D.W., Valacich, J.S.: Unpacking the privacy paradox: irrational decision-making within the privacy calculus, IS security and privacy. In: Proceedings of the 33rd International Conference on Information Systems, Orlando, FL (2012)
Xu, H., Luo, X.R., Carroll, J.M., Rosson, M.B.: The personalization privacy paradox: an exploratory study of decision making process for location-aware marketing. Decis. Support Syst. 51, 42–52 (2011)
Xu, H., Teo, H.H., Tan, B.C., Agarwal, R.: The role of push-pull technology in privacy calculus: the case of location-based services. J. Manag. Inf. Syst. 26, 135–174 (2009)
Zeng, S.Y., Wu, L.L., Chen, H.G.: Sharing private information online: the mediator effect of social exchange, service innovations for E-commerce. In: Proceedings of the 11th International Conference on Electronic Commerce, Taipei, Taiwan (2009)
Acknowledgment
This work was supported in part by the Ministry of Science and Technology of the Republic of China under the grant MOST 103-2410-H-030-087-MY3.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer International Publishing AG, part of Springer Nature
About this paper
Cite this paper
Ku, YC., Li, PY., Lee, YL. (2018). Are You Worried About Personalized Service? An Empirical Study of the Personalization-Privacy Paradox. In: Nah, FH., Xiao, B. (eds) HCI in Business, Government, and Organizations. HCIBGO 2018. Lecture Notes in Computer Science(), vol 10923. Springer, Cham. https://doi.org/10.1007/978-3-319-91716-0_27
Download citation
DOI: https://doi.org/10.1007/978-3-319-91716-0_27
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-91715-3
Online ISBN: 978-3-319-91716-0
eBook Packages: Computer ScienceComputer Science (R0)