See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/317243869

Websites, Data Types and Information Privacy Concerns: A Contingency

Model

Article  in  Telematics and Informatics · May 2017

DOI: 10.1016/j.tele.2017.05.012

CITATION READS

1 219

2 authors, including:

Jih-Hsin Tang
National Taipei University of Business
23 PUBLICATIONS   566 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

mobile mental health View project

All content following this page was uploaded by Jih-Hsin Tang on 17 October 2017.

The user has requested enhancement of the downloaded file.

 Telematics and Informatics 34 (2017) 1274–1284

Contents lists available at ScienceDirect

Telematics and Informatics

journal homepage: www.elsevier.com/locate/tele

Websites, data types and information privacy concerns: A

contingency model
Jih-Hsin Tang ⇑, Yi-Jen Lin
National Taipei University of Business, No. 321, Sec.1, Chi-Nan Rd., Taipei, Taiwan, Republic of China

a r t i c l e i n f o a b s t r a c t

Article history:
Received 28 October 2016 Ó 2017 Elsevier Ltd. All rights reserved.
Accepted 25 May 2017
Available online 27 May 2017

1. Introduction

Although the growth of e-business over the past years, the online users’ privacy concerns is a critical inhibitor of
e-business. The owners of e-businesses have to collect consumers’ information in order to provide ‘‘personalized” service;
however, some consumers feel their privacy is threatened (Culnan and Armstrong, 1999; Dinev and Hart, 2005). Consumers
show even higher privacy concerns on the big data environment (Shin and Choi, 2015) and on the location-based service of
e-business (Aloudat et al., 2013).
Users are motivated to conduct online activities such as to collect information, to make friends, to purchase goods/ser-
vices or to play online games. Some online activities require users to register in order to provide services. For example, users
are required to provide customers’ basic information in order to complete a transaction; sometimes, users are requested to
fill in more personal information such as personal interests and hobbies in order to acquire personalized services. Previous
scholars have contributed in designing a scale for information privacy concerns (Smith et al., 1996; Stewart and Segars,
2002), or in the relationship between privacy and online transactions (Dinev and Hart, 2005), or extend calculus privacy
model (Dinev and Hart, 2006).
Recently, the popularity of social-networking sites such as Myspace or Facebook makes online users’ privacy concerns a
hot topic (Dwyer et al., 2007; Hoadley et al., 2009). But our questions are: do the users have the same privacy concerns over
the different Websites? And, whether the individuals show different privacy concerns on required data types and optional
data types? This study attempts to investigate the relationship between Website types, data types and online consumers’
privacy concerns, and develop and validate a contingency model accordingly. The results may be of theoretical and practical
importance because the literature is still scant in describing the relationship between Websites, data type and information
privacy concerns.

2. Literature review

Information privacy has long been considered a critical ethical issue in the information age (Mason, 1986). Anecdotes of
personal information leaks have covered the daily headline news and frightened some online users. Zuckerberg, the founder
of Facebook, said in an award ‘‘privacy was no longer a social norm” (Johnson, 2010). The young CEO’s remarks make infor-

⇑ Corresponding author.
E-mail addresses: jefftang@ntub.edu.tw (J.-H. Tang), jen_lin074@hotmail.com (Y.-J. Lin).

http://dx.doi.org/10.1016/j.tele.2017.05.012
0736-5853/Ó 2017 Elsevier Ltd. All rights reserved.
 J.-H. Tang, Y.-J. Lin / Telematics and Informatics 34 (2017) 1274–1284 1275

mation privacy another hot-debated topic. Trust and privacy were compared in different social-networking sites (Dwyer
et al., 2007), and some authors considered that information privacy is just an illusion(Hoadley et al., 2009).
Information privacy is usually defined an ‘‘individual’s control of personal information and use”(Westin, 1967; Stone and
Stone, 1990). With the advance of internet technology, an individual may find out the control and use of personal informa-
tion over the Internet challenged.

2.1. Concerns for information privacy

Concern for Information privacy is a complex construct and many researchers attempted to clarify this concept in the IS
literature (Culnan and Armstrong, 1999; Hann et al., 2007; Smith et al., 1996; Stewart and Segars, 2002). Smith et al. (1996)
developed and validated an instrument, concern for information privacy (CFIP) that identifies and measures the major
dimensions of individuals’ concern about organizational information privacy practices. And they suggest that individuals
with a high concern for information privacy perceive that: (1) too much data are collected, (2) much of the data is inaccurate,
(3) corporations use personal information for undisclosed purposes, and (4) corporations fail to protect access to personal
information. CFIP has been examined (Stewart and Segars, 2002) and applied to ease individual’s privacy concerns (Hann
et al., 2007), and how it may affect individual’s intention of transaction online (Van Slyke et al., 2006).
The efforts of developing a valid scale for information privacy have been relentless. For instance, Culnan and Armstrong
(1999) suggested a concept of ‘‘privacy calculus” and argued that individuals will disclose personal information if they per-
ceive that the overall benefits of disclosure are at least balanced by, if not greater than, the assessed risk of disclosure, thus
equaling a cost-benefit analysis with privacy calculus(Culnan and Bies, 2003). Dinev and Hart (2006) extended and validated
the concept of privacy calculus to e-commerce settings and claimed that the cumulative influence of trust and personal
Internet interest are important factors that can outweigh privacy risk perceptions in the decision to disclose personal infor-
mation online. Malhotra et al. (2004) developed an Internet version of information privacy scale consisting of collection, con-
trol, and awareness of privacy practices. Although the dimensions of information privacy concerns are still under debate, this
study adopted CFIP for it is widely-used and clarity. The major dimensions of CFIP are described below

(1) Collection: amounts of identifiable information to be collected and stored in the database.
(2) Errors: the improper protection and violations of individual’s personal information.
(3) Unauthorized secondary use: the data collected from an institution but used in a different institution without con-
sumers’ consent.
(4) Improper access: individual’s personal data is without appropriate protection, and some ‘‘illegal parties or users” may
gain access to it.

2.2. Website types and information privacy

But, online users feel that ‘‘privacy is no longer a social norm”? Or do they have different levels of information privacy?
We would like to investigate this interesting research question by comparing users’ information-privacy concerns on two
Websites: one is relation-based (Facebook.com), and the other is transaction-based (Ruten.com, Taiwan’s version of eBay).

2.3. Website types and information privacy concerns

While individuals may engage in different online activities, they may register for membership. For instance, if an individ-
ual would like to engage in an online community, he or she has to provide personal information in order to become a mem-
ber. Similarly, if an individual would like to complete a transaction online, he or she has to provide ‘‘necessary personal
information” such as delivered address, e-mail address, and phone numbers in order to make the transaction complete. Users
are motivated to join a membership, thus they might have various information privacy concerns. Based on the motivation to
join an online community, Armstrong and John (1996) classified online communities into four types: transaction, interest,
fantasy, and relationship. Based on the business model, Rappa (2001) classified Websites into two groups: one is
transaction-based and the other relation-based. The main purpose of a transaction-based Website is to complete a business
transaction, while the relation-based Website is to maintain or extend personal social relations. Since Websites serve differ-
ent purposes, and possibly ask for different personal information for corporate use or online self-presentation, we derive that
individuals show the different level of information privacy concerns as Hypothesis 1.

H1. While engaging in different Websites (relational vs. transactional), individuals show different level of information
privacy concerns (collection, improper access, errors and unauthorized secondary use).

2.4. Data types and information privacy concerns

Several surveys confirm that the Internet users generally feel differently about the disclosure of different types of infor-
mation (Kobsa, 2007a,b). They are usually willing to disclose basic demographic and lifestyle information as well as personal
1276 J.-H. Tang, Y.-J. Lin / Telematics and Informatics 34 (2017) 1274–1284

tastes and hobbies. They are slightly less willing to disclose details about their online behavior, followed by the more
extended demographic information. Financial information, contact information, and specifically credit card and Social Secu-
rity numbers raise the highest privacy concerns.
Although the prior studies have contributed to the understanding of the sensitivity of personal information; users have to
decide to fill in the ‘‘required” and/or ‘‘optional” data in order to register for a Website. The purposes of required data fields
and optional data fields might be different. Usually the required data fields are declared to be used within a company;
however, the optional data might be served as online self-presentation on a social-networking site, or for corporate customer
profile analysis. According to a recent study: the top three personal information that Facebook users are unwilling to disclose
are: indoor telephone numbers, politics orientation and personal address (Hoadley et al., 2009). These three data fields are
optional. Therefore, it seems reasonable that an individual treats required data fields and optional data fields differently and
show the different level of privacy concerns.

H2. While registered for a Website, individuals show different level of information privacy concerns for required data
(corporate use) and optional data (online presence).
Prior research has demonstrated that some demographic variables might have an impact on the privacy concerns. Several
surveys since the early 1980s were able to cluster respondents into roughly three groups: privacy fundamentalists generally
express extreme concern about any use of their data and an unwillingness to disclose information, even when privacy pro-
tection mechanisms would be in place. The privacy unconcerned tend to express mild concern for privacy, and mild anxiety
about how other people and organizations use information about them. Privacy pragmatists are generally concerned about
their privacy as well, but less than the fundamentalists. They are more willing to disclose personal information when they
understand the reasons for its use, see benefits for doing so, or see privacy protections in place(Kobsa, 2007a,b). For instance,
females seem to show higher levels of concerns on information collection (Sheehan, 2000). Being cheated experiences online
was also demonstrated to be associated with an individual’s decision to register or not (Hui et al., 2007). On a large scale
survey, Lenhart (2005) reported that 79 percent of online teens in the United States declared ‘‘teens are not careful enough
about giving out their personal information online.” Moreover, parents of younger teens are more apt to be concerned about
the disclosure of personal data. Teenagers seem to provide their personal information freely to join social networks on the
Internet (Barnes, 2006). Younger users might have lower levels of risk, thus might fewer concerns about information privacy.
Similarly, if an individual does not use a social networking site, he or she may not understand the privacy risks exactly, thus
concerns less about information privacy. Therefore, we consider an individual’s level of information privacy concerns toward
a Website or data type may be moderated by gender, age, use and experiences of being cheated online.

H3a. Individual’s level of information privacy concerns toward Website types (relational vs. transactional) is moderated by
gender, age, use and experiences of being cheated online.

H3b. Individual’s level of information privacy concerns toward data types (required vs. optional) is moderated by gender,
age, use and experiences of being cheated online.

3. Reseach methodology and results

3.1. Representativeness of Website type

Initially we would like to choose the largest transactional Website (eBay) and the largest social-networking site (Face-
book) for comparison. However, eBay.com has stopped its business in Taiwan since 2004; we chose one of the largest
C2C transactional Websites Ruten.com instead. Ideally it is better to choose several Websites (both transactional and
social-networking) in order to make the comparison of Website type more sound and valid; we chose just one representative
Website each type.

3.2. Scale development and survey administration

The research model was empirically tested using data collected with two online surveys that included items for the con-
structs specified in the model. We constructed the initial set of items by analyzing the literature and translated Smith et al.’s
work (1996) into Chinese and adapted into two versions, one is for Facebook.com and the other for Ruten.com. As for the
data type, all the required data fields and optional data fields of these two Websites were provided for CFIP measurement
each. That is, respondents were required to respond to their privacy concerns on all data fields individually (14 data fields
for Facebook, and 9 for Ruten.com). These two forms of the survey are identical except in some background descriptions. All
the items were measured by Likert five scales, where 1 represents disagreement and 5 agreement. These initial surveys were
validated by 2 domain experts and pilot tested by 43 respondents. All respondents were asked whether they ever registered
for Facebook or Ruten.com. Pretest results demonstrated fair reliability for all CFIP dimensions (Cronbach a ranged from 0.74
to 0.96), and confirmatory factor analysis results demonstrated convergent and divergent validity, as shown in Table 1. The
 J.-H. Tang, Y.-J. Lin / Telematics and Informatics 34 (2017) 1274–1284 1277

Table 1
Confirmatory Factor Analysis Results of CFIP.

Dimension item Errors Unauthorized secondary use Improper access Collection

Errors E_1 0.96 0.06 0.12 0.11
E_2 0.94 0.15 0.03 0.07
E_3 0.92 0.07 0.18 0.03
E_4 0.93 0.01 0.15 0.08
Unauthorized secondary use SU_1 0.08 0.87 0.25 0.07
SU_2 0.05 0.93 0.21 0.01
SU_3 0.06 0.77 0.12 0.19
SU_4 0.11 0.83 0.24 0.21
Improper access IA_1 0.08 0.19 0.79 0.04
IA_2 0.02 0.26 0.91 0.01
IA_3 0.05 0.30 0.84 0.18
Collection C_1 0.10 0.25 0.24 0.81
C_2 0.04 0.21 0.08 0.86
C_3 0.33 0.08 0.48 0.59
C_4 0.12 0.02 0.12 0.89
Construct Reliability 0.96 0.95 0.95 0.74
Eigen Value 5.26 3.21 2.29 1.55
Variance Explained% 35.1% 21.4% 15.3% 10.3%
Accumulated Variance Explained% 35.1% 56.5% 71.7% 82.1%
Average Variance Extracted (AVE) 0.88 0.73 0.72 0.63

formal survey was administered online. To encourage volunteers to participate this study, we provided 20 prizes for lot-
drawing at the end of the study on each form. About 3 weeks, a total 643 responses from Facebook and a total 570 from
Ruten.com were obtained.
After checking for e-mail duplicates, IP duplicates, and incomplete responses, we obtained 356 valid responses from Face-
book and 330 from Ruten.com. The demographics of these participants were described in Table 2.
As shown in Table 2, participants of both Facebook and Ruten.com were heterogeneous. Approximately the same portion
of male and female participated Facebook, female consisted of two-thirds of the sample on Ruten.com. The users of Ruten.-
com seem younger, with the majority of 18–24 years old. Surprisingly, a considerable portion of the sample was declared
being cheated online, with approximately 20% of Facebook and 39% of Ruten.com.
Because all these samples were volunteers for incentive-based online survey, we followed Armstrong and Overton’s sug-
gestion (1977) for non-response bias analysis. We compared the earliest 25% and latest 25% respondents on both samples in
all demographic variables and the four CFIP dimensions, and all except collection dimension of Facebook showed the signif-
icant difference. That is, the non-response bias was minimal in this study (See Table 3).

3.3. Website type on CFIP

We conducted a One-Way MANOVA analysis; Website type as the independent variable and CFIP subscales (collect, error,
secondary use and improper access) as dependent variables. As shown in Table 2, the overall CFIP differs significantly in
these two Websites (Wilk’s K = 0.78, p < 0.001), implying that information privacy concerns might contingent on Website
types. Our findings showed that users might express the higher level of information privacy concerns (errors, secondary
use and improper access) on transaction-based Websites than on relation-based Websites. Hypothesis I was supported.
On average, individuals are most concerned about the unauthorized secondary use of their personal information, followed
by improper access, errors, and collection the least, similar to Milberg et al.’s results (1995). Regardless the Websites, users
consider the unauthorized secondary use the most serious problem, yet the collection the least serious.

3.4. Website type and data type on CFIP

The basic statistics of Website type, data type on CFIP were presented on Table 4. When it comes to collection dimension,
users seem to show the higher level of concerns on optional data than required data. However, when it comes to unautho-
rized secondary use dimension, user concern more about the required data than the optional data. It seems no distinguish-
able difference on the improper access and errors dimension. Furthermore, as the interaction effect of Website Type and Data
Type on CFIP was significant (F = 3.68, p = 0.006 < 0.05) , simple main effects were examined first. As shown in Table 5, the
effects of data type on CFIP were both found significantly with Wilk’s K equals to 0.90 and 0.96 respectively. Similarly, the
main effects of Website type on CFIP were also found significant with Wilk’s K equals to 0.78 and 0.79 respectively. While we
examined the effect of data type on CFIP at Facebook.com in Table 5, interesting results were found. On the collection and
improper access dimensions, privacy concerns about the optional data were significantly higher than the required data
(F = 50.53, F = 12.14, both p < 0.01 respectively). However, on the unauthorized secondary use dimension, privacy concerns
about the optional data were significantly lower than the required data (F = 6.62, p < 0.05). No significant difference was
found on the errors dimension. The findings imply that users of social networking sites show the higher level of concerns
1278 J.-H. Tang, Y.-J. Lin / Telematics and Informatics 34 (2017) 1274–1284

Table 2
Descriptive Statistics of Survey Respondents.

Variables Groups count percentage count percentage

Facebook (N = 356) Ruten.com (N = 330)
Gender Male 172 48.3% 137 41.5%
Female 184 51.7% 193 58.5%
Age Below 17 12 3.4% 5 1.5%
18–24 145 40.7% 209 63.3%
25–34 162 45.5% 109 33.0%
35–44 25 7.0% 4 1.2%
45 above 12 3.4% 3 0.9%
Education Junior high school and below 10 2.8% 2 0.6%
Senior high school 49 13.8% 17 5.2%
College 240 67.4% 223 67.6%
Graduate school 57 16.0% 88 26.7%
Profession Students 145 40.7% 212 64.2%
Medical 5 1.4% 6 1.8%
IT 27 7.6% 16 4.8%
Real estate 7 2.0% 1 0.3%
Accounting 7 2.0% 1 0.3%
Banking/insurance 14 4.0% 1 0.3%
Customer service 29 8.1% 8 2.4%
Marketing 4 1.1% 1 0.3%
Design 8 2.2% 4 1.2%
Education 12 3.4% 16 4.8%
Sales 10 2.8% 4 1.2%
Logistics 3 0.8% 2 0.6%
Manufacturing 27 7.6% 12 3.6%
Retailing 10 2.8% 2 0.6%
Media 1 0.3% 1 0.3%
Housewives 11 3.1% 8 2.4%
Unemployed 25 7.0% 20 6.1%
Retired & others 11 3.1% 15 4.5%
Years used Half a year 96 27.0% 74 22.4%
A year 195 54.8% 54 16.4%
1–2 years 58 16.3% 79 23.9%
Over 2 years 7 2.0% 123 37.3%
Hours spent daily Within an hour 155 43.5% 238 72.1%
1–2 h 110 30.9% 69 20.9%
2–3 h 33 9.3% 11 3.3%
3–4 h 22 6.2% 4 1.2%
Above 4 h 36 10.1% 8 2.4%
Experiences Used this site? (Yes) 339 95.2% 320 97.0%
Being cheated on this site? (Yes) 70 19.7% 127 38.5%
Heard cheating incidents of this site? (Yes) 270 75.8% 309 93.6%

Table 3
One-Way MANOVA Summary for Website Type on CFIP.

Websites Facebook N = 356 Ruten.com N = 330 Wilk’s K F value

CFIP dimensions Mean (S.D.) Mean (S.D.) 0.78***
Collection 3.40 (0.73) 3.67 (0.73) 22.62***
a<b
Improper Access 3.81 (0.73) 4.42 (0.68) 124.03***
a<b
Errors 3.59 (0.74) 4.16 (0.67) 114.04***
a<b
Unauthorized Secondary Use 4.19 (0.72) 4.73 (0.54) 124.58***
a<b

Note: a represents Facebook and b Ruten.com.

***
p < 0.001.

about the collection and improper access of their optional data such as home addresses, home phone numbers than the
required data fields. Yet, on the unauthorized secondary, the required data seems worse than the optional data.
While we examined the simple effect of data type on CFIP at Ruten.com in Table 4, similarly results were found. On the
collection dimension, privacy concerns about the optional data were significantly higher than the required data (F = 14.11,
 J.-H. Tang, Y.-J. Lin / Telematics and Informatics 34 (2017) 1274–1284 1279

Table 4
Basic Statistics of Website Type and Data Type on CFIP.

Dependent Variables Website Type Data Type

Required Mean (SD) Optional Mean (SD)
Collection F 3.40 (0.04) 3.76 (0.04)
R 3.66 (0.04) 3.87 (0.04)
Improper Access F 3.81 (0.04) 4.00 (0.04)
R 4.42 (0.04) 4.39 (0.04)
Errors F 3.59 (0.43) 3.58 (0.43)
R 4.16 (0.04) 4.16 (0.04)
Unauthorized Secondary Use F 4.19 (0.04) 4.05 (0.04)
R 4.73 (0.03) 4.62 (0.03)

Note: F represents Facebook, and Ruten.com.

Table 5
Two-way MANOVA Summary for Websites, Data Type on CFIP.

Source df Wilk’s K F value

Collection Improper Access Errors Unauthorized Secondary Use
Websites
required 1 0.78*** 22.62*** 124.03*** 114.04*** 124.58***
optional 1 0.79*** 4.51* 56.19*** 85.50*** 128.73***
Data Type
Facebook 1 0.90*** 50.53*** 12.14** 0.00 6.62*
Ruten.com 1 0.96*** 14.11*** 0.27 0.01 6.65*
***
p < 0.001.
**
p < 0.01.
*
p < 0.05.

p < 0.001). On the unauthorized secondary use dimension, privacy concerns about the optional data were significantly lower
than the required data (F = 6.65, p < 0.05). No significant differences were found on the improper access and the errors
dimensions. Again, users of a transactional Website concern more about the collection of optional data; however, they con-
cern more about the unauthorized secondary use of required data.
On the simple effect of required data type, all CFIP dimensions of Ruten.com were found significantly higher than those on
Facebook.com (F = 22.62; F = 124.03; F = 114.04; F = 124.58, all p < 0.01), implying that online users show the higher privacy
concerns of the required data on a transactional Website. Similarly, on the simple effect of optional data type, all the CFIP
dimensions of Ruten.com were found significantly higher than those on Facebook.com (F = 4.51, p < 0.05; F = 56.19;
F = 85.50; F = 128.73, all p < 0.001), implying that online users show the higher privacy concerns of the optional data on a
transactional Website.
While comparing users’ privacy concerns on the identical data fields of two Websites, users tend to show more concerns
on collection of their names, e-mails and cell phone numbers on Facebook.com rather than on Ruten.com (with t = 3.35, 5.26
and 2.07 respectively). However, on the dimensions of errors and unauthorized secondary use the situations are totally the
opposite. Users tend to show the higher concerns on their names, e-mails, home phone numbers and cell phone numbers on
Ruten.com rather than Facebook. Furthermore, users show higher concerns on the improper access of their names, indoor
phone numbers, and cell phone numbers on Ruten.com than Facebook. No significant differences were found on improper
access of their e-mails and addresses.
The findings imply that users of a social networking site concerns more on their personal information such as names, e-
mails and cell phone numbers to be collected; however, users of a transactional site concerns even more on the improper
access, errors and unauthorized secondary use of their personal information.
While comparing the CFIP of un-identical data fields of two Websites, no suitable statistics could be applied and only
means and standard deviations were presented. As shown in Table 6, users seem show the higher concerns on the improper
access and unauthorized secondary use of the data fields regardless of the Websites (from mean = 4.08–4.78). The highest
concern is the unauthorized secondary use of their ID number and cell phone number at Ruten.com. They also show high
level of concerns about the errors of their ID and cell phone numbers (See Tables 7 and 8).
However, they seem to show a different pattern on Facebook. For instance, they concern more on their data to be col-
lected on Facebook (means from 3.8 to 4.22) than Ruten.com. The only exception is that Ruten.com users concern about their
ID No. to be collected very much (mean = 4.33). The findings give us an insight about how users view their personal infor-
mation to be collected, accessed improperly, and secondary used without notification. They seems to concern more on a
transactional site rather than a social networking site; yet they also show a high level concern on their personal information
to be collected at a social networking site. Their privacy concerns seem to be contingent on the type of Websites, the data
type and the exact data fields individually.
1280 J.-H. Tang, Y.-J. Lin / Telematics and Informatics 34 (2017) 1274–1284

Table 6
Summary of CFIP results on the identical data fields of two Websites.

Websites Collection Improper Access Errors Unauthorized

Secondary Use
Mean (S.D) t value Mean (S.D) t value Mean (S.D) t value Mean (S.D) t value
** *** ***
Required Name F 3.72 (0.89) 3.35 4.15 (0.79) 5.24 3.77 (0.85) 7.06 4.43 (0.71) 6.56***
R 3.49 (0.92) 4.45 (0.72) 4.20 (0.75) 4.75 (0.53)
Email F 3.23 (0.88) 5.26*** 3.93 (0.80) 0.41 3.72 (0.81) 2.12* 4.33 (0.78) 4.18***
R 2.88 (0.88) 3.90 (0.90) 3.85 (0.76) 4.57 (0.72)
Optional Home address F 4.22 (0.83) 1.29 4.46 (0.74) 1.41 4.00 (0.94) 5.26*** 4.46 (0.85) 4.67***
R 4.14 (0.80) 4.53 (0.66) 4.34 (0.75) 4.72 (0.54)
Home phone No. F 4.20 (0.78) 0.36 4.37 (0.77) 2.77** 3.96 (0.94) 5.05*** 4.48 (0.71) 4.93***
R 4.22 (0.78) 4.53 (0.68) 4.29 (0.77) 4.72 (0.54)
Cell phone No. F 4.11 (0.80) 2.07* 4.34 (0.76) 2.60* 3.90 (0.92) 6.43*** 4.40 (0.74) 5.34***
R 3.98 (0.86) 4.48 (0.66) 4.31 (0.74) 4.68 (0.60)

F stands for Facebook (N=356) and R Ruten.com (N=330).

***
p < 0.001;
**
p < 0.01.
*
p < 0.05.

Table 7
Summary of CFIP results on the un-identical data fields of two Websites.

Information Privacy Concerns Collection Improper Access Errors Unauthorized Secondary Use
Data Type Website Data field Mean (SD) Mean (SD) Mean (SD) Mean (SD)
required Facebook Gender 3.80 (0.99) 4.14 (0.89) 3.73 (0.89) 4.40 (0.82)
Birthday 3.82 (0.91) 4.12 (0.81) 3.72 (0.99) 4.42 (0.78)
Ruten.com Identification No. 4.33 (0.79) 4.62 (0.69) 4.41 (0.74) 4.78 (0.52)
Cell Phone No. 3.67 (0.88) 4.43 (0.75) 4.22 (0.73) 4.73 (0.57)
Optional Facebook Relationship Status 4.22 (0.83) 4.36 (0.84) 3.89 (1.01) 4.42 (0.80)
Sexual Orientation 4.20 (0.78) 4.37 (0.77) 3.80 (1.07) 4.43 (0.77)
Politics 4.05 (0.86) 4.34 (0.77) 3.96 (0.94) 4.48 (0.71)
Religion 4.08 (0.88) 4.33 (0.83) 3.85 (1.02) 4.36 (0.81)
City 4.17 (0.86) 4.33 (0.83) 3.79 (1.05) 4.31 (0.87)
School/Company 4.11 (0.80) 4.34 (0.76) 3.83 (1.00) 4.39 (0.82)
Ruten.com e-mail address 3.23 (0.79) 4.09 (0.79) 4.03 (0.77) 4.48 (0.71)
MSN account 3.36 (0.83) 4.08 (0.82) 3.91 (0.79) 4.47 (0.74)

Table 8
Summary of Hypothesis Testing Results.

Hypothesis Results
H1 While engaging in different Websites (relational vs. transactional), individuals show different level of information Supported
privacy concerns
H2 While registered for a Website, individuals show different level of information privacy concerns for required data Supported
(corporate use) and optional data (online presence).
H3a Individual’s level of information privacy concerns toward Website types (relational vs. transactional) is moderated by Partially
gender, age, use and experiences of being cheated online. Supported
H3b Individual’s level of information privacy concerns toward data types (required vs. optional) is moderated by gender, Partially
age, use and experiences of being cheated online. Supported

3.5. Moderating effects

To examine the moderating effects of gender, age, use and being cheated experiences online, we used one variant statistic
in two conditions.

3.5.1. CFIP towards Website types

The interaction effect between Website type and gender was found significant at unauthorized secondary use (F = 4.99,
p < 0.05). As shown in Fig. 2, female showed more concerns on unauthorized secondary use at a transactional Website. Yet,
no significant gender difference was found on the same Website.
Besides, The interaction effect between Website type and the experiences of being cheated online was also found signif-
icant at improper access (F = 4.56, p < 0.05). As shown in Fig. 3, those who were being cheated showed the higher concerns
 J.-H. Tang, Y.-J. Lin / Telematics and Informatics 34 (2017) 1274–1284 1281

Websites
Relational H1
Transactional
CFIP
Collection
Improper Access
H2 Errors
Data Type
Unauthorized Secondary Use
Required
Optional H3a
H3b

Moderators
Gender
Age
Use
Being cheated experiences

Fig. 1. Hypothesized Relationships of the Contingency Model of Information Privacy Concerns.

Fig. 2. The Moderating Effect of Gender on Website Type and Unauthorized Secondary Use.

Fig. 3. The Moderating Role of Being Cheated Experiences on Website Type and Improper Access.
1282 J.-H. Tang, Y.-J. Lin / Telematics and Informatics 34 (2017) 1274–1284

Fig. 4. The Moderating Role of Age on Data Type and Collection.

Fig. 5. The Moderating Role of Website Use on Data Type and Collection.

on improper access on a transactional Website than on a social networking site. All other moderating effects (age and usage)
were not found.

3.5.2. CFIP towards data types

The interaction effect between data type and age was found significant at collection (F = 2.79, p < 0.05). As shown in Fig. 4,
younger users showed more concerns on optional data collection and older user (>35) view the equal importance both of the
optional and the required data.
Besides, The interaction effect between data type and the experiences of use was also found significant at collection
(F = 4.45, p < 0.05). As shown in Fig. 5, those who did not register for a Website show no noticeable difference on their con-
cern towards the required data and the optional data. However, once they registered for a Website and used for a while, they
showed the higher concerns on their optional data to be collected. All other moderating effects (gender and being cheated)
were not found.

4. Conclusion

As predicted, online users seems to show higher level of privacy concerns on a transactional Website rather than a social-
networking site. It is reasonable since a poor management of personal information on an e-commerce site might cause huge
monetary loss, yet the same errors on a social networking site seem to be less vulnerable. Generally, online users concerns
about the unauthorized secondary use the most, reflecting their doubt about the privacy assurance provided by Website
administrators. Website type may play an important role in information privacy concerns.
Besides, online users seem to treat their personal information (required data and optional data) differently. They seem to
show the higher level concerns on the collection and unauthorized secondary use of required data; however, the users of a
 J.-H. Tang, Y.-J. Lin / Telematics and Informatics 34 (2017) 1274–1284 1283

social-networking site than those of a transactional site concerns more on their contact information (names, e-mail address
and cell phone numbers) to be collected online. Interestingly, users’ perception of privacy seems naïve: while they do not
register for a Website, they consider the required data fields on the registration form more private and personal. Once upon
they registered for the Website, they show the higher level of privacy concerns on the optional data fields. The younger users
concern more on their optional data to be collected; the older ones see no difference between required data and optional
data. Information privacy concerns were analyzed in the unit of data field in this study.
Although our samples were from Taiwan alone, our results demonstrated that online users are still concerned about infor-
mation privacy regardless of Websites. That is to say, our respondents consider ‘‘privacy is still a social norm”, counter to
Zuckerberg’s remarks. Besides, users treat data fields individually. That is, users view the sensitivity of each data field differ-
ently. Future work should expand Website types and data fields and gain a deeper understanding how this factors may con-
tribute to individuals’ information privacy concerns.

Limitations

The major limitation of this study is that subjects were self-selected and from Taiwan only; therefore, the findings could
not be generalized for all Website users. The self-reported survey may also be subject to honesty or social desirability issues,
and culture may also be a potential variable for self-disclosure.

Appendix data fields of Facebook and Ruten.Com

Required Optional
Facebook Last Name Home address
First Name Home phone No.
Email address Cell Phone No.
Gender Relationship Status
Birth day Sexual orientation
Political affiliations
Religion
Country of residence
School or Company employed

Ruten.com Full Name Home address

ID No. Home phone No.
Cell Phone Cell Phone No.
E-mail address Instant Messenger account

References

Aloudat, A., Michael, K., Chen, X., Al-Debei, M.M., 2013. Social acceptance of location-based mobile government services for emergency management.
Telematics Inform. 31 (1), 153–171.
Armstrong, A., John, H.I., 1996. The real value of online communities. Harv. Bus. Rev. 74 (3), 134–140.
Barnes, S.B., 2006. A privacy paradox: Social networking in the United States. First Monday 11 (9).
Culnan, M.J., Armstrong, P.K., 1999. Information privacy concerns, procedural fairness, and impersonal trust: an empirical investigation. Organ. Sci. 10 (1),
104–115.
Culnan, M.J., Bies, R.J., 2003. Customer privacy: balancing economic and justice considerations. J. Social Issues 59 (2), 323–342.
Dinev, T., Hart, P., 2005. Internet privacy concerns and social awareness as determinants of intention to transact. Int. J. Electron. Commerce 10 (2), 7–29.
Dinev, T., Hart, P., 2006. An extended privacy calculus model for e-commerce transactions. Inf. Syst. Res. 17 (1), 61–80.
Dwyer, C., S. R. and Hiltz, et al. 2007. Trust and privacy concern within social networking sites: A comparison of Facebook and MySpace. Proceedings of the
Thirteenth Americas Conference on Information Systems.
Hann, I.-H., Hui, K.-L., et al, 2007. Overcoming online information privacy concerns: an information processing theory approach. J. Manage. Inf. Syst. 24 (2),
13–42.
Hoadley, C. M., H. and Xu, et al., 2009. Privacy as information access and illusory control: The case of the Facebook News Feed privacy outcry. Electronic
Commerce Research and Applications.
Hui, K.-L., TeoH, H., et al, 2007. The value of privacy assurance: an exploratory field experiment. MIS Q. 31 (1), 19–33.
Johnson, B., 2010. Privacy no longer a social norm, says Facebook founder Available at https://www.theguardian.com/technology/2010/jan/11/facebook-
privacy.
Kobsa, A., 2007a. Privacy-enhanced personalization. Commun. ACM 50 (8), 24–33.
Kobsa, A. (Ed.), 2007b. Privacy-enhanced Web personalization. The adaptive web: methods and strategies of web personalization. Springer Verlag, The
Adaptive Web.
Lenhart, A. (2005). Protecing teens online, Pew Internet & American Life Project. Available at http://www.pewinternet.org/2005/03/17/protecting-teens-
online/.
Malhotra, N.K., Kim, S., et al, 2004. Internet users’ information privacy concerns (IUIPC): the construct, the scale, and a causal model. Inf. Syst. Res. 15 (4),
336–355.
Mason, R.O., 1986. Four ethical issues of the information age. MIS Q. 10 (1), 4–12.
Milberg, S., Burke, S., et al, 1995. Values, personal information, privacy and regulatory approaches. Commun. ACM 38 (12), 65–74.
1284 J.-H. Tang, Y.-J. Lin / Telematics and Informatics 34 (2017) 1274–1284

Rappa, M., 2001. Managing the digital enterprise – business model on the Web Available at http://digitalenterprise.org/models/models.html.
Sheehan, K.B., 2000. An investigation of gender differences in online privacy concerns and resultant behaviors. J. Inter. Market. 13 (4), 24–38.
Van Slyke, C., Shim, J.T., Johnson, R., Jiang, J.J., 2006. Concern for information privacy and online consumer purchasing. J. Assoc. Inf. Syst. 7 (6), 415–444.
Shin, D.H., Choi, M.J., 2015. Ecological views of big data: perspectives and issues. Telematics Inform. 32, 311–320.
Smith, H.J., Milberg, S.J., Burke, S.J., 1996. Information privacy: measuring individuals’ concern about organizational practices. MIS Q. 20 (2), 167–196.
Stewart, K.A., Segars, A.H., 2002. An empirical examination of the concern for information privacy instrument. Inf. Syst. Res. 13 (1), 36–49.
Stone, E. F. and Stone D. L. 1990. Privacy in organizations: theoretical issues, research findings, and protection mechanisms. Research in Personnel and
Human Resources Management. K. M. Rowland and G. R. Ferris. Greenwich, CT, JAI Press. 8: 349-411.
Westin, A., 1967. Privacy and Freedom. Atheneum, New York.

View publication stats