Documente Academic
Documente Profesional
Documente Cultură
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 12 May 2015
Received in revised form
8 April 2016
Accepted 13 April 2016
Available online 22 April 2016
The expanding use of social networking sites (SNSs) and their exible functionalities allow individuals to
use these sites in unexpected and unethical ways. SNS users face ethical dilemmas each day and they
need to decide how to behave in these cases. The aim of this study was to identify the inuential factors
that have an impact on individuals' ethical decision-making and propose a model of the factors that are
signicant in the ethical decision-making process in the SNS context. This study employed the Theory of
Planned Behavior (TPB) and included personal normative beliefs, moral intensity, perceived threat of
legal punishment to the main constructs of TPB namely attitude, subjective norms, perceived behavioral
control to predict the behavioral intention using a scenario based questionnaire. The moderating effects
of several factors, including age, gender, level of income, ego strength, locus of control and religion, were
investigated for the probable effect on the proposed model. Derived from a designed procedure, several
scenarios were constructed, of which four were selected to be embedded in the survey instrument. With
441 returned questionnaires in hand, the collected data were analyzed using the partial least squares
estructural equation modeling technique. The results showed attitude to be the most inuential factor,
followed by subjective norms, perceived behavioral control, personal normative beliefs, and moral intensity. Level of education, age, and scenario also showed a signicant effect on the relation between
predictors and behavioral intention.
2016 Elsevier Ltd. All rights reserved.
Keywords:
Ethical behavior
Theory of Planned Behavior
Social networking sites
Computer ethics
1. Introduction
During recent years, Social Networking Sites (SNSs) have
become a major communication vehicle in all societies. Facebook
alone has more than 2.2 billion users which means one-third of the
world's population are using it (Edwards, 2014). The uptake of SNSs
has attracted many concerns about the different ethical issues
regarding their usage, such as employee surveillance (Byrnside,
2007; Rothberg, 2008), application of users' data for advertising
purposes (Reynolds, 2011), privacy (Hull, Lipford, & Latulipe, 2011),
identity theft (Bilge, Strufe, Balzarotti, & Kirda, 2009), inappropriate
prole content (Peluchette & Karl, 2009), cyberstalking (Haron &
Yusof, 2010), cyberbullying (Hinduja & Patchin, 2010; Huang &
Chou, 2013; Kwan & Skoric, 2013), cyberloang (Krishnan, Lim, &
* Corresponding author.
E-mail address: karimi@damavandiau.ac.ir (H. Jafarkarimi).
http://dx.doi.org/10.1016/j.chb.2016.04.024
0747-5632/ 2016 Elsevier Ltd. All rights reserved.
Teo, 2010; Liberman, Seidman, Mckenna, & Buffardi, 2011; Tang &
Austin, 2009), self-disclosure (Livingstone, 2008) spreading rumors and false news (Leitch & Warren, 2009) and SNS addiction
(Balci and Glc, 2013). SNSs are among the powerful Information
and Communication Technology (ICT) tools that link people to each
other and provide a vast range of information and entertainment on
the Internet. This potential makes them a ground for new ethical
challenges (Light and Mcgrath, 2010).
In order to increase the moral responsibilities that consequently
cause ethical actions, the only option is to help individuals' development through initial stages to higher stages of moral development (Kohlberg, 1975). This aim could be achieved by moral
discussions and education based on ethical dilemmas. Making
ethical decisions is a complicated process which an individual goes
through to come out with a do or don't intention. However, this
might be a strenuous task since there is neither generally recognized approach to teaching ethical issues nor a generally accepted
theory of Computer Ethics (Floridi & Sanders, 2002). In such case,
546
know, like their family and friends, or even those who they do not
know and want to make new personal and professional relationship with. The most popular SNS is Facebook which has more than 1
billion active users (Whittaker, 2012). Other popular SNSs are
Google, Twitter, Qzone (popular in China), LinkedIn, Instagram,
etc.
2.2. Theory of Planned Behavior
Different theories and models have been proposed to examine
the relationship between different factors and behaviors (Olson &
Zanna, 1993). Among them, Wicker (1969) looked at the relationship of attitude and behavior and concluded that these two probably have no relation. Out of frustration with traditional attitudebehavior researches that much of them found weak correlation
between attitude and performance of a volitional behavior (Hale,
Householder, & Greene, 2002. p. 259), Fishbein and Ajzen (1975)
proposed the Theory of Reasoned Action (TRA) which aimed to
explain volitional behaviors. Later, they extended TRA and the
outcome was the TPB (Ajzen, 1985). TRA and TPB became the most
widely researched models of behavior (Armitage & Conner, 2001).
TPB suggests that Attitude, Subjective Norms (SN) and Perceived
Behavioral Control (PBC) are good predictors of intention and
consequently of behavior. The TPB claims that the performance of
deliberate intentional behaviors could be predicted with high accuracy from intentions and perceptions of behavioral control
(Ajzen, 1991). As Ajzen (1985) suggested, this theory could be
applied to all volitional activities and behaviors such as skiing,
swimming and mountain climbing. TPB can be used to predict
specic behaviors across occasions, situations and different forms
of action. TPB claims that intention to a great extent inuences the
actual behavior (Beck & Ajzen, 1991). The ve constituents of TPB
can be dened as below:
i. Behavior is an individual's act in a certain case which is the
result of intention and PBC, such that a desired intention will
only lead to behavior if PBC is strong.
ii. Intention is the core element of TPB and refers to individuals'
readiness to perform a given behavior.
iii. Attitude refers to the evaluation of an individual about the
question of how favorable or unfavorable an act is. It depends
on an individual's beliefs and the evaluation of those beliefs.
iv. SN is dened as an individual's perception about what most
others who are important to him or her think he or she
should do.
v. PBC refers to the perceived easiness or difculty of an act to
be done.
Armitage and Conner (2001) conducted a review of 185 studies
and found evidence to support the TPB as an efcient predictor of
intention and behavior. In addition, many studies claim that this
theory is appropriate in ICT contexts (Banerjee, Cronan, & Jones,
1998; Chatterjee, 2008; Inedo, 2012; Leonard, Cronan, & Kreie,
2004; Namlu & Odabasi, 2007). Hence, the TPB is applied in this
study to address the factors that are important in people's behavior
in the SNS context.
In line with TPB's variables, and based on a literature review,
another three constructs namely personal normative belief (PNB),
Moral Intensity (MI) and perceived threat of legal punishment
(PTLP) were included to be embedded in the TPB to investigate
behavioral intention in SNS dilemma cases. PNB refers to an individual's feeling about his or her moral obligation to do or not to
do an act. MI is a construct that relates to issues in terms of their
perceived moral signicance (Jones, 1991). Finally, PTLP is dened
as the probability of legal punishment in the case of doing an act
547
the vast amount of dilemmas faced by their users every day. This
study focuses on SNSs and the probable dilemmas that may happen
to any individuals who use it on a daily basis and investigates the
role of different factors that have an effect on individuals' behavioral intention. To this end, it is necessary to discover the probable
factors that hypothetically could inuence people's behavioral
intention. For this purpose, the related models in the eld of ICT
were reviewed and the relevant factors were extracted to be used in
the proposed conceptual model. Hence, the main objectives of our
research are:
i. To identify the important factors that inuence an individual's intention to behave unethically in SNSs.
ii. To design a scenario-based instrument to measure the relation of the factors and behavioral intention.
iii. To propose and evaluate a general model for unethical use of
SNSs among students.
In comparison with research efforts found in the literature, our
work has the following signicances. Firstly, Surveys reect that the
teaching of online manners is not taken seriously in Malaysia. According to the Global Youth Online Behavior Survey, conducted by
Microsoft (2012), only 18% of Malaysian parents teach online
manners to their children which is far below the average of the
countries included in the survey (39%). This reects the gap in the
felt signicance of Computer Ethics (CE) in Malaysia and other
countries. To address this issue, it is important to have solutions
based on Malaysian culture instead of borrowing strategies from
studies conducted in other contexts. Other solutions may not work
in different cultures. The present study was carried out in Malaysia
which is considered as one of the important countries in Asia in the
eld of ICT. Secondly, this research model investigates a set of
moderators which were not included in one place in previous
studies. The investigated variables altogether can be considered as
a comprehensive set of constructs in comparison with earlier
studies. In addition, this research is among the pioneering works in
the eld of CE in the context of SNS in Malaysia. It is also among
very few attempts to understand inuential factors of unethical
behavior in the context of SNSs in the world. The ndings of this
research will provide some theoretical information to understand
the signicant factors that cause ethical or unethical behavior in
ethical dilemma cases in the context of SNSs. The outcomes of this
study can support the development of educational plans to increase
the level of awareness among people to control the rate of unethical
behavior using ICT. As well, the outcomes of this study can help to
improve solutions for security and privacy concerns that are one of
the main issues in CE.
4. Hypothesis development
The actual behavior of people in ethical dilemma cases, especially in the SNS context, is almost impossible to observe. As there is
a need to see their actions when facing these dilemmas, and due to
the fact that it is not possible to observe individuals' decisions in
ethical dilemma cases, this research employed designed scenarios
to check individual intentions in the presented scenarios. The
following hypotheses were formed to investigate the impact of a set
of variables and unethical behavioral intention. These hypotheses
are the basis for the structured equation model of unethical
behavior in SNS context.
4.1. Attitude
Although some researchers have reported attitude as an insignicant factor in TPB (Banerjee et al., 1998), many researchers claim
548
549
4.7.5. Age
Based on Kohlberg's (1975) theory of moral development, it is
expected that older people are more ethical. This was supported in
other research where it has been shown that age does matter in
ethical decision making and that older people are subject to less
inuence from external factors (Peterson et al., 2001). As a result,
other scholars have included age as an inuential variable in ethical
decision making and examined its role (Deshpande, 1997; Gupta,
Swanson, & Cunningham, 2010; Honeycutt, Glassman, Zugelder,
& Karande, 2001; Moores & Chang, 2006; Weeks, Moore,
Mckinney, & Longenecker, 1999).
4.7.7. Citizenship
Since students in the sample were from different countries, we
were able to compare the results based on subsamples from
different countries. This aimed to understand whether or not differences in culture may have an impact on the unethical behavioral
intention in context of SNS.
550
5. Method
5.1. Instrument design
Same as previous studies on ethical behavior, a questionnaire
was designed to measure discussed constructs to capture each individual's intention to behave unethically in a provided SNS related
scenario. Scenarios included previously studied ethical issues like
privacy, cyberbullying, accuracy etc. and 23 possible dilemma cases
were designed and presented to a panel of experts including 4
psychologists, 3 IS professors and 6 PhD students in the eld of IS.
The panel was presented with a questionnaire which asked them to
evaluate the scenarios in terms of ethicality and importance to this
research. Based on this procedure 4 scenarios were selected to be
included in the survey (Appendix A). In line with the scenarios, this
questionnaire employed several previously validated measures and
instruments which are presented in Appendix B.
The designed questionnaire which included demographic
questions, scenario and its related questions, Ego test and LOC test,
were presented to 3 IS Professors for content validity. After revising
the requested changes, a pilot test was done on 48 individuals who
were selected on a convenience basis, each answering 2 sets of
questions regarding 2 different scenarios and the questionnaire
was nalized for the main survey. Due to the complaints received
by respondents due to the length of questionnaire, it was discovered that more than one scenario in a questionnaire can decrease
the quality of answers and hence, only one scenario was presented
to each respondent in the main survey.
5.2. Sample
6.1. Measurement validity
The respondents were selected from students in different elds
and majors from Universiti Teknologi Malaysia.
To ensure the adequacy of the sample size, G*power software
(Faul, Erdfelder, Lang, & Buchner, 2007) was used. The settings
proposed by Dattalo (2008) were used (a 0.05 and b 0.80 for
error type one and two as stated in Dattalo's study) and for six
551
Table 1
Respondents' prole.
Gender
Age
Citizenship
Religion
SNS type
a
b
Male
Female
<20
(20e22]
(22e25]
(25e29]
>29
Malaysian
Non-Malaysian
Muslim
Buddhist
Hindu
Christian
Others
SPMa & STPMb
Diploma
Bachelor
Master
PhD
Others
<500
501e1000
1001e2000
2001e3000
3000<
Facebook
Twitter
Google
LinkedIn
Other SNSs
None
<30
[30 mine1 hour)
[1e3) hours
[3e4) hours
[4e5) hours
7< hours
Frequency
Frequency
Questionnaire type
Total 441
V1
V2
V3
54
59
18
37
21
16
21
80
33
93
7
3
6
4
7
3
63
23
16
1
36
35
18
14
9
109
38
43
20
21
1
9
24
31
20
11
17
60
51
21
29
20
26
15
72
39
90
10
1
6
4
4
3
56
29
16
3
22
49
20
8
11
106
36
39
25
11
0
5
22
44
17
7
15
59
53
33
18
24
21
16
75
37
94
51
54
26
23
16
18
22
69
36
87
1
5
4
4
1
54
40
9
4
34
38
23
7
8
99
40
48
21
18
0
13
28
35
17
6
13
1
5
4
2
2
58
25
16
2
27
43
15
8
11
102
38
47
22
19
0
12
26
31
20
7
9
Percentage
V4
224
217
98
107
81
81
74
296
145
364
33
6
22
16
17
9
231
117
57
10
119
165
76
38
39
416
152
177
88
69
1
39
100
141
74
31
54
51%
49%
22%
24%
18%
18%
17%
67%
33%
83%
7%
1%
5%
4%
4%
2%
52%
27%
13%
2%
27%
37%
17%
9%
9%
94%
34%
40%
20%
16%
0%
9%
23%
32%
17%
7%
12%
552
Table 2
Reliability and convergent validity check results for reective measures.
Construct
Convergent Validity
Name
Short form
AVE (>0.5)
Attitude
ATT
0.860
0.895
0.780
Intention
INT
0.860
0.897
0.781
PTLP
0.844
0.906
0.762
PNB
0.651
0.811
0.590
PBC
0.778
0.870
0.691
Subjective norms
SN
0.684
0.816
0.609
SNS addiction
SAD
0.815
0.850
0.477
ATT01
ATT02
ATT03
INT01
INT02
INT03
PTLP01
PTLP02
PTLP03
PNB01
PNB02
PNB03
PBC01
PBC02
PBC03
SN01
SN02
SN03
SAD01
SAD02
SAD03
SAD04
SAD05
SAD06
0.909
0.884
0.857
0.897
0.849
0.903
0.850
0.910
0.858
0.739
0.733
0.828
0.823
0.857
0.813
0.655
0.875
0.899
0.482
0.805
0.541
0.744
0.801
0.704
Table 3
FornelleLarcker criterion analysis.
SAD
ATT
INT
PBC
PNB
PTLP
SN
MI
SAD
ATT
INT
PBC
PNB
PTLP
SN
MI
0.732
0.171
0.132
0.045
0.105
0.082
0.101
0.01
0.883
0.723
0.422
0.638
0.357
0.562
0.406
0.884
0.538
0.612
0.29
0.639
0.419
0.831
0.343
0.227
0.442
0.223
0.768
0.428
0.529
0.465
0.873
0.291
0.468
0.78
0.364
Formative
Table 4
Results of factor analysis for formative construct.
Collinearity statistics
Signicance
Name
Construct
Items
Tolerance
(>0.2)
VIF
(<5)
Outer weight
Outer loading
t-statistic (>1.96)
p value
Moral intensity
MI 01
MI 02
MI 03
0.760
0.588
0.638
1.316
1.700
1.566
0.329
0.681
0.496
0.640
0.917
0.850
0.877
4.772
3.399
0.381
0.000 (***)
0.000 (***)
Note:
***
<0.01.
Table 5
Summary of results.
HI
H2
H3
H4
H5
H6
Note:
**
<0.05,
R2 0.657
Path
Attitude
Subjective norms
Perceived behavioral control
Personal normative beliefs
Moral intensity
Perceived threat of legal punishment
ATT / INT
SN / INT
PBC / INT
PNB / INT
MI / INT
PTLP / INT
***
Hypothesis test
Path coefcient
p-value
0.398
0.215
0.216
0.159
0.096
0.074
0.000
0.000
0.000
0.000
0.019
0.021
(***)
(***)
(***)
(***)
(**)
(**)
t-statistic
Result
8.176
5.204
5.495
3.122
2.069
2.030
Supported
Supported
Supported
Supported
Supported
Not supported
<0.01.
p < 0.001 for H1 to H4 and p < 0.01 for H5) except H6. For the effect
553
As it was discussed, all the collected data were based on the four
scenarios that were designed and embedded within the questionnaire. Therefore, it was a crucial step to investigate the role of those
scenarios on the observed relationships in the model. To this end, at
rst, the results of the hypotheses testing based on different scenarios were gured (see Appendix C) and another MGA-PLS was
then conducted. As there were four groups of subsamples regarding
scenarios, four sets of MGA-PLS were done. Each time, the responses to one scenario were excluded from the sample and the
results of the tests for that particular scenario were compared with
the whole dataset excluding that particular scenario. Table 11
represents the results for PLS-MGA based on different scenarios.
The empirical evidence indicates a marginal difference (p < 0.1)
for the statistical signicance of SN in Scenario 1, in comparison
with the rest of the responses. In conclusion, based on the results in
Table 11 and Appendix C, without trying to propose a general rule, it
Table 6
Effect size results for the conrmed hypotheses.
Path
Attitude / intention
Subjective norms / intention
Perceived behavioral control / intention
Personal normative beliefs / intention
Moral intensity / intention
Note: NS nonsignicant.
Effect size
f2
Category
0.22
0.08
0.09
0.03
0.01
Medium
Small
Small
Small
NS
Table 7
Predictive relevance results.
Path
Attitude / intention
Subjective norms / intention
Perceived behavioral control / intention
Personal normative beliefs / intention
Moral intensity / intention
Predictive relevance
q2
Category
0.13
0.04
0.05
0.02
0.01
Small
Small
Small
Small
NS
Note: NS nonsignicant.
554
Table 8
The results of MGA-PLS for categorical moderators.
Independent variable
ATT
MI
PBC
PNB
PTLP
SN
Note:
<0.10,
**
Undergraduates vs.
Postgraduates
t-statistic
p value
t-statistic
p value
t-statistic
p value
t-statistic
p value
0.396
0.32
0.361
0.096
1.638
0.289
0.692
0.749
0.718
0.923
0.102
0.772
1.049
0.321
0.92
0.654
1.358
0.173
0.295
0.749
0.358
0.514
0.175
0.863
1.274
0.043
0.758
0.489
0.123
1.379
0.204
0.966
0.449
0.625
0.902
0.169
2.179
0.622
1.582
1.972
0.751
0.03
0.03 (**)
0.534
0.114
0.095 (*)
0.453
0.976
<0.05.
Table 9
Results of PIA-PLS for continuous moderators.
Moderator
Age
Path coefcient
t-statistic
p value
Signicance level
Path coefcient
t-statistic
p value
Signicance level
Path coefcient
t-statistic
p value
Signicance level
Path coefcient
t-statistic
p value
Signicance level
Ego strength
LOC
SNS addiction
Note:
**
<0.05,
***
ATT / INT
MI / INT
PBC / INT
PNB / INT
PTLP / INT
SN / INT
0.077
2.606
0.009
(***)
0.01
0.288
0.773
NS
0.01
0.298
0.766
NS
0.091
1.176
0.24
NS
0.035
0.826
0.409
NS
0.022
0.567
0.571
NS
0.3
0.724
0.469
NS
0.045
0.571
0.568
NS
0.002
0.047
0.963
NS
0.011
0.347
0.728
NS
0.004
0.13
0.897
NS
0.102
0.991
0.322
NS
0.065
2.003
0.046
(**)
0.027
0.763
0.445
NS
0.016
0.478
0.633
NS
0.1
1.188
0.235
NS
0.092
0.268
0.788
NS
0.000
0.008
0.994
NS
0.012
0.463
0.643
NS
0.035
0.998
0.318
NS
0.032
0.839
0.402
NS
0.03
1.08
0.281
NS
0.004
0.11
0.912
NS
0.063
0.948
0.344
NS
<0.01.
Table 10
Direct impact of SNS addiction, age, ego strength and LOC on behavioral intention.
Path
Path coefcient
t statistics
p value
Result
0.047
0.032
0.037
0.018
1.384
0.565
1.191
0.673
0.167
0.572
0.234
0.501
NS
NS
NS
NS
Table 11
PLS-MGA results for different scenarios.
1 vs. 2, 3, 4
ATT
MI
PBC
PNB
PTLP
SN
Note:
<0.10,
**
2 vs. 1, 3, 4
3 vs. 1, 2, 4
4 vs. 1, 2, 3
t-statistic
p value
t-statistic
p value
t-statistic
p value
t-statistic
p value
0.367
0.074
1.18
1.103
1.413
1.679
0.714
0.941
0.239
0.271
0.158
0.094 (*)
0.007
0.264
0.562
0.007
2.368
0.322
0.994
0.792
0.574
0.994
0.018 (**)
0.748
1.371
0.961
0.337
0.698
1.583
0.751
0.173
0.338
0.737
0.486
0.114
0.454
0.952
0.231
1.15
0.951
0.176
0.231
0.342
0.818
0.248
0.343
0.860
0.818
<0.05.
555
556
coefcient value reected the contrary. This was the same case
with PTLP's role in two of four scenarios (as presented in
Appendix C), and in another two scenarios, the path coefcient
was not even signicant. To explain this result, several reasons
can be considered. The rst reason could be that the measures
failed to reect the PTLP construct. There is a possibility that this
issue arose and it is a normal occurrence in research. However,
since the PTLP construct indices were based on previous studies
(Grasmick & Green, 1980), and they passed different steps of
validly and reliability tests, it is very unlikely that the weakness
of the measurement caused this nding. The second reason could
be the fact that people engage in online unethical behavior more
easily, even if there is a strong rule to avoid it. For instance,
although there are strong legal acts to ght cybercrimes, the rate
of cybercrime victimization is at least two times higher than
conventional crimes (UNODC, 2013). To be more specic, under
the Malaysian Copyright Amendment Act (2003), copyright offences have nes up to RM20,000 (for each infringing copy).
However, pirated movies, music and software are easily accessible and widely used all over the country (Bernama, 2010). In
these cases, although people know that the law is strict, they still
use infringing copies. Based on these facts, it can be said that
although the threat of legal punishment might be high for a
certain activity, people still commit the illegal act. In fact, in this
case, the PTLP should have no impact on the behavioral intention.
However, the t-statistic and path coefcient reected a positive
relationship which was statistically signicant and could not be
the outcome of chance. Therefore, the third and best justication
is to argue that a missing factor is required to mediate the impact
of PTLP on intention. A mediating effect occurs when a third
variable, which represents the generative mechanism through
which the main independent variable is affected, is able to inuence the dependent variable of interest (Baron & Kenny, 1986).
Although the probable mediator of this relationship is outside
the scope of this research, overall gain is a candidate construct
which can be studied in future research. This suggests that, since
people do a costebenet analysis prior to a behavior (Schweitzer,
n
~ ez, & Douma, 2004), they behave in a way to secure the
Ordo
highest gain. Taking overall gain as the mediator of PTLP means
that even though the risk of legal punishment for some acts increases, the gain from the act is far more, which makes PTLP
insignicant. To put it in other words, in the investigated scenarios, although people believe that if they get caught, they will
face a problem, the pleasure or benet from those particular
activities is far more than the risk. Therefore, as PTLP increases,
the pleasure in these activities also increases. As a result, when
the impact of this hypothetical mediator is deleted, the path
between PTLP and intention seems to be a positive and strong
relationship. This is the most probable cause of the positive
signicance of PTLP on intention.
Among the potential moderators including gender, religion,
ego strength, LOC, SNS addiction, age, citizenship and level of
education, only the last two reected a signicant moderating
effect. Interestingly, the level of education and age were both
shown to be moderators of attitude and PNB. The results suggest
that older people care more about their feelings of how morally
obligated they are if they engage in an action. If they feel more
morally obligated in an unethical behavior, they would not be very
interested in performing that act. Instead, if they believe that
something is good to do, they would be more interested to engage
in it and would be less inuenced by other forces such as SN. The
same is the case for level of education. For instance, if a case
happens in which sharing a particular comment or picture is
among the options, older people and more educated people are
more likely to follow their own attitude and moral obligation
557
558
Appendix B
Instruments and measures used in this research
Construct
Item
Adapted from
Intention
Attitude
Subjective norms
Moral intensity**
Ego strength
Locus of control
*
(Ajzen, 2005)
(Barron, 1953)
(Rotter, 1966)
In the questionnaire, depending on the scenario that was used, different names were replaced by X.
since MI was a multi dimension construct, it was considered as formative construct.
**
Appendix C
Hypothesis
Path
Path coefcient
t-statistic
p-value
1
R2 0.669
H1
H2
H3
H4
H5
H6
H1
H2
H3
H4
H5
H6
H1
H2
H3
H4
H5
H6
H1
H2
H3
H4
H5
H6
ATT / INT
SN / INT
PBC / NT
PNB / INT
MI / INT
PTLP / INT
ATT / INT
SN / INT
PBC / NT
PNB / INT
MI / INT
PTLP / INT
ATT / INT
SN / INT
PBC / NT
PNB / INT
MI / INT
PTLP / INT
ATT / INT
SN / INT
PBC / NT
PNB / INT
MI / INT
PTLP / INT
0.367
0.088
0.284
0.225
0.047
0.001
0.395
0.237
0.235
0.13
0.031
0.231
0.233
0.266
0.191
0.214
0.128
0.19
0.453
0.183
0.143
0.071
0.048
0.025
4.877
1.161
3.654
2.852
1.104
0.012
5.3
3.06
3.204
1.91
1.143
2.702
1.959
3.23
2.692
2.511
2.404
1.876
4.962
1.906
2.143
0.944
1.167
0.308
0.000
0.248
0.000
0.005
0.272
0.991
0.000
0.003
0.002
0.059
0.256
0.008
0.053
0.002
0.008
0.013
0.018
0.063
0.000
0.059
0.034
0.347
0.246
0.759
2
R2 0.683
3
R2 0.696
4
R2 0.642
Result
***
NS
***
**
NS
NS
***
**
***
*
NS
NS
*
***
***
**
**
NS
***
*
**
NS
NS
NS
Note:
NS not supported.
*
<0.10, ** <0.05, *** <0.01.
References
Adam, A. (2001). Cyberstalking: gender and computer ethics. In E. Green, & A. Adam
(Eds.), Virtual gender: Technology, consumption and identity (1st ed., pp.
559
560
561
245e255.
Sparks, P., Shepherd, R., Wieringa, N., & Zimmermanns, N. (1995). Perceived
behavioural control, unrealistic optimism and dietary change: an exploratory
study. Appetite, 24(3), 243e255.
Speri, A. (2014). ISIS ghters and their friends are total social media pros.
Stone, M. (1974). Cross-validatory choice and assessment of statistical predictions.
Journal of the Royal Statistical Society, 36(2), 111e147.
Strader, T. J., Fichtner, J. R., Simpson, L. A., & Clayton, S. R. (2011). The impact of
context on employee perceptions of acceptable non-work related computing.
International Journal of Technoethics, 2(2), 30e44.
Tait, R. (2006). Censorship fears rise as iran blocks access to top websites. Retrieved
from.
website:
http://www.guardian.co.uk
http://www.guardian.co.uk/
technology/2006/dec/04/news.iran.
Tan, B. (2002). Understanding consumer ethical decision making with respect to
purchase of pirated software. Journal of Consumer Marketing, 19(2), 96e111.
Tang, T. L.-P., & Austin, M. J. (2009). Students' perceptions of teaching technologies,
application of technologies, and academic performance. Computers & Education,
53(4), 1241e1255.
Thompson, L., Dawson, K., Ferdig, R., Black, E., Boyer, J., Coutts, J., et al. (2008). The
intersection of online social networking with medical professionalism. Journal
of General Internal Medicine, 23(7), 954e957.
Tramow, D., & Finlay, K. A. (1996). The importance of subjective norms for a minority of people: between subjects and within-subjects analyses. Personality
and Social Psychology Bulletin, 22(8), 820e828.
Trevino, L. K. (1986). Ethical decision making in organizations: a person-situation
interactionist model. The Academy of Management Review, 11(3), 601e617.
Trottier, D., & Lyon, D. (2012). Key features of social media surveillance. Internet and
Surveillance: The Challenges of Web 2.0 and Social Media, 16, 89e105.
UNODC. 2013. Comprehensive study on cybercrime (U. N. O. o. D. a. Crime, Trans.).
In 1 (Ed.), 1 (Vol. 1, pp. 320). Vienna: United Nations.
Vitell, S. J., Nwachukwu, S. L., & Barnes, J. H. (1993). The effects of culture on ethical
decision-making: an application of hofstede's typology. Journal of Business
Ethics, 12(10), 753e760.
Wang, X., & Mcclung, S. R. (2011). Toward a detailed understanding of illegal digital
downloading intentions: an extended theory of planned behavior approach.
New Media & Society, 13(4), 663e677.
Wang, J., Tian, J., & Shen, Z. (2013). The effects and moderators of cyber-loang
controls: an empirical study of chinese public servants. Information Technology and Management, 14(4), 269e282.
Weaver, G. R., & Agle, B. R. (2002). Religiosity and ethical behavior in organizations:
a symbolic interactionist perspective. The Academy of Management Review,
27(1), 77e97.
Weeks, W. A., Moore, C. W., Mckinney, J. A., & Longenecker, J. G. (1999). The effects
of gender and career stage on ethical judgment. Journal of Business Ethics, 20(4),
301e313.
Whittaker, Z. (2012). Facebook hits 1 billion active user milestone. Retrieved 12/2/
2014, from http://news.cnet.com/8301-1023_3-57525797-93/facebook-hits-1billion-active-user-milestone/.
Wicker, A. W. (1969). Attitudes versus actions: the relationship of verbal and overt
behavioral responses to attitude objects. Journal of Social Issues, 25(4), 41e78.
Wingrove, T., Korpas, A. L., & Weisz, V. (2010). Why were millions of people not
obeying the law? Motivational inuences on non-compliance with the law in
the case of music piracy. Psychology, Crime & Law, 17(3), 261e276.
Yoon, C. (2011). Theory of planned behavior and ethics theory in digital piracy: an
integrated model. Journal of Business Ethics, 100(3), 405e417.
Zimmer, M. (2010). But the data is already public: on the ethics of research in
facebook. Ethics and Information Technology, 12(4), 313e325.