Content uploaded by Soudabeh Ghoroghi
Author content
All content in this area was uploaded by Soudabeh Ghoroghi on Jan 03, 2017
Content may be subject to copyright.
Abstract—This study attempts to develop a predictive model
to enhance the understanding of cybersex by testing
hypotheses on the relationships between religiosity, risk taking,
and cybersex engagement. Using multistage proportional
sampling, 256 postgraduate students who completed an online
survey sent to their email addresses were randomly selected
from five Malaysian universities. Results provided support for
the proposed theoretical model by explaining 22% of variance
in endogenous variable. Statistically significant negative
relationship was found to exist between religiosity and
cybersex engagement. The study also demonstrated a positive
significant relationship between risk taking and cybersex
engagement. University counselors would do well to be aware,
and develop accurate and general knowledge about online
sexual activities, address and prevent the probability of it
becoming an addiction with serious life consequences for
students.
Index Terms—Cybersex, high order reflective construct,
postgraduate students, religiosity.
I. INTRODUCTION
Technological progress of the Internet has led to the
emergence and rapid growth of adult entertainment such as
the delivery of sexual materials ranging from text, stories, to
pictures, and videos. Statistics related to Internet sexuality
have pointed out that about 42.7% of Internet users visit
pornographic web sites [1].
In the present study, “Cybersex” is simply defined as
“The use of the Internet for sex” [2], and comprises
engagement in diverse sexually motivated behaviors in
interactive or solitary form including: viewing text, pictures,
videos, sexchats, using webcam, specific information search
about sexual issues, establishing sexually-biased contacts
online, etc. [3]-[5]. These activities can be grouped into
solitary-arousal (e.g. watching and downloading
pornography), partnered-arousal or interactive, (e.g. sex
chats, webcamming, sharing sexual fantasies), and
non-arousal activities (e.g. directed information search) [5],
[6].
Several studies conclude that cybersex can alter the user’s
Manuscript received July 25, 2016; revised October 21, 2016.
S. Gh and Siti Aishah Hassan are with the Department of Counselor
Education and Counseling Psychology, Faculty of Educational Studies,
Universiti Putra Malaysia, Malaysia (e-mail: sudygh@yahoo.com,
siti_aishahh@upm.edu.my).
Ahmad Fauzi Mohd Ayub is with Department of Foundation Studies,
Faculty of Educational Studies, Universiti Putra Malaysia, Malaysia (e-mail:
afmy@upm.edu.my).
mental, emotional, and social characteristics. While it is true
that most Internet users use it for recreational or utilitarian
purposes, there are individuals who develop an addiction to
cybersex [7]. Cybersex addiction is defined as a compulsion
accompanied by extreme Internet sexually-oriented
behavioral patterns that prevail over an individual’s life,
thoughts, and feelings [8].
Students, especially university students, spent more time
online and are in centre attention of researchers because of
their potential ability to engage in compulsive Internet
behaviors. Undoubtedly, students are vulnerable in the face of
problems related to Internet use, especially to excessive
Internet use. Malaysia’s level of social networking activity has
shown an exceptionally high level of engagement. In 2010,
Malaysia was number nine in the world and in terms of the
number of Facebook users, the country is ranked third highest
in the Asia Pacific region [9]. Pornography and other forms of
online sexual activities are known to be rampant as reported in
the media [10] and especially among university students in
Malaysia [11].
It is stated that influential factors such as personal
characteristics have been a comparatively neglected area of
research in the field of cybersexual studies [12]. Therefore,
exploring the possibility that students with certain personal
characteristics are more likely to be excessive cybersex users
might help us to better predict and intervene prior to the
development of addictive behaviors. In view of that, students’
risk taking propensity, and religiosity is highlighted as
possible influencers in cybersex engagement.
II. LITERATURE REVIEW
A. Religiosity
For a long time now, psychologists have shown great
interest, with varying degrees of favor or disfavor, in the role
that religion plays in interpreting and responding to life events
and psychological adjustment [13]. However, there has been
little attention paid to the relationship between technology
adoption and religion and the tension between technological
development and religious beliefs [14]. The study of religion
and the Internet, which is a subfield of Internet Studies can
improve our knowledge and discussion of the larger social and
cultural shifts at work within networked society [15].
Malaysia is an excellent example of a multiethnic and
multi-religious country in Southeast Asia with a relatively
positive level of cultural and religious tolerance [16]. Ethnic
Malays and Bumiputeras comprise 78% of the total
population of about 30 million [17], and almost all of them are
The Influence of Religiosity and Risk Taking on Cybersex
Engagement among Postgraduate Students: A Study in
Malaysian Universities
Soudabeh Ghoroghi, Siti Aishah Hassan, and Ahmad Fauzi Mohd Ayub
International Journal of Social Science and Humanity, Vol. 7, No. 3, March 2017
143
doi: 10.18178/ijssh.2017.7.3.810
Sunni Muslims and follow the Shafi’i legal school of Islam
[18]. Huff [19] maintains that the experience of Malaysia,
with its Muslim majority population and rapid endorsement
of technology, provides a strong contrast to Muslim nations
in the Middle East, where political sensitivities have
impeded the development of information technology [19].
According to Campbell [20] the rise of religious
fundamentalism globally within traditional religions such as
Islam, Judaism, and Christianity, reveals reactions to
globalization and technology. However, various religious
communities have adopted and in some cases embraced the
Internet as part of their contemporary religious mission and
strategy for growth [21].
This contention was echoed by a study on a sample of
2,698 Turkish Muslims by Hesapçı Sanaktekın Aslanbay,
and Gorgulu [14]. The result of the study shows that the
degree of religiosity significantly affected the patterns of
Internet consumption. High believers, compared to
moderate or non-believer groups, use the Internet more for
the purpose of searching for information [14]. Policy-wise,
there is a relationship between religious belief and
advocating greater restriction on access to Internet
pornography [22]. Religious individuals tend to disapprove
of pornography use and support pornography censorship
[23]. In some related studies, it was found that pornography
use was lower in religious populations than non-religious
and secular populations [24], [25].
In contrast, the results from two studies of undergraduate
samples (Study1, N=331; Study2, N=97) of Grubbs et al.
[26] showed that religiosity and perceived addiction to
pornography were strongly and positively related and that
this relationship was mediated by moral disapproval of
pornography use. These results persisted even when actual
use of pornography was controlled.
B. Risk Taking
As a personality characteristic or trait, risk-taking
propensity is a form of individual difference [27], and can be
identified by a natural tendency to the seeking of varied,
novel, complex and intense experiences and engaging in
potentially harmful behaviors [28].
In recent years, there have been growing concerns about
risk-taking behaviors by students. In early studies, the
prevalence of adolescent risk-taking behaviors has been
addressed through numerous epidemiological surveys.
These studies have repeatedly focused, almost exclusively,
on behaviors, including such activities as driving a car after
drinking, riding with a drunk driver, shoplifting, having sex
without a condom, etc. [29]. Nowadays, the spread of risky
behaviors in society make social scientists continue to
devote considerable attention to spillover effects from risky
behaviors especially among college and university students
[30].
Risk taking is part of life, but people differ in their
risk-taking propensity. Some people enjoy risky pursuits
while others dislike such activities [31]. It is very important
to see association between risk taking behaviors and
cybersex involvement because risk taking online can
translate into actual risky sexual encounters.
Several survey results suggest that more frequent
engagement in cybersex has been positively associated with
risky behaviors. Users feel safe using the Internet for sexual
purposes, encouraging more adventurous and riskier
behaviors [32], [33].
In accordance with the mentioned arguments we proposed
the following hypotheses:
H1. Religiosity negatively influences cybersex
engagement.
H2. Risk taking positively influences cybersex
engagement.
III. METHOD
We conducted a quantitative online survey among 256
master’s and PhD. students. Due to the sensitivity of the
research topic, before recruiting participants for the study,
ethical approval for this study was granted by the Ethics
Committee for Research Involving Human Subjects
Universiti Putra Malaysia (JKEUPM). Further, prior to
answering the survey questions, an electronic consent form
was used as a tool to provide assurance to participants that
their privacy, confidentiality, and participation would be fully
anonymous with no link to their email address. The consent
form further assured that their participation was voluntary and
withdrawal would have no consequences. Sample size when
conducting PLS path modeling adheres to guidelines from
Chin [34] that suggested a minimum sample size equal to or
exceeding 10 times the larger of the largest number of
structural paths leading to a latent variable. In this study the
largest number of structural paths belongs to religiosity scale
(LPCG dimension) with 8 indicators. Therefore, a sample
comprising 256 respondents who completed questionnaires
was acceptable. This study uses multistage proportional
sampling. First, five universities were randomly selected
among Klang Valley universities (the states of Selangor and
Kuala Lumpur) including two public and three private
universities. Second, proportional sampling was applied to
come up with appropriate sample sizes for each university. In
the third stage, having students email addresses, there was the
possibility for random selection of emails in each university
based on proportional sample size in the second stage.
Considering the issues of confidentiality, selected universities
are called A, B, C, D, and E.
A. Instrumentation
The study applied three instruments to measure the
variables of the study. These include an online questionnaire
containing questions about respondents’ age, gender, study
level, sexual orientation as well as the following
questionnaires:
Internet Sex Screening Test (ISST) Delmonico and Miller
[35]: Among the instruments, it is an excellent measure for
evaluating online sexual behaviors [36], [37]. The rating ISST
scale administered is a 25-item, True-False measure of online
sexual behavior developed by Delmonico [38]. ISST has been
utilized as a self-administered, screening instrument to assist
individuals in determining whether their Internet sexual
behavior has reached the stage of being clinically problematic.
Delmonico and Miller [35] used factor analysis to empirically
establish sub-scales, of which there are five: (a) online sexual
International Journal of Social Science and Humanity, Vol. 7, No. 3, March 2017
144
compulsivity (OSC), (b) online sexual behavior-social
(OSB-S), (c) online sexual behavior-isolated (OSB-I), (d)
online sexual spending (OSS), and (e) interest in online
sexual behavior (IOSB). ISST produces a total score
evaluating problematic online sexual behavior as well as the
scores of the five subscales. Sample items of ISST are: “I
have a sexualized username or nickname that I use on the
Internet”, “I have made promises to myself to stop using the
Internet for sexual purposes,” and “I have stayed up after
midnight to access sexual material online”. For this study,
the scale was transformed into a 5-point Likert scale. The
ISST has proof of convergent validity, and has a relation
with another measure of sexual addiction and internal
consistency for sub-scales in the study reported as α =0.51 to
0.86 [35]. The present study resulted in α = 0.64 to 0.83.
Domain-Specific Risk-Taking Scale (DOSPERT) Blais
and Weber [39]: Risk taking is often domain specific,
meaning that somebody’s ethical risk taking may not predict
his or her health or social risk taking [39]. The risk-taking
responses of the 18-item version of the DOSPERT Scale
assess behavioral intentions or how likely it is that
respondents might participate in risky activities/behavior
developed by Blais and Weber [39]. The risk-taking scale
assesses behavioral tendencies, and risky behaviors derived
from five domains of life (ethical, financial, health/safety,
social, and recreational risks) on a 30-item scale. This study
however uses only the ethical, health/safety and social
domains. Sample items use a 7-point rating scale ranging
from 1 (extremely unlikely) to 7 (extremely likely), and
include "Revealing a friend’s secret to someone else" and
"Engaging in unprotected sex". Blais and Weber [39]
provided evidence proof of the factorial and
convergent/discriminant validity of the scores regarding
constructs such as sensation seeking, dispositional risk
taking, intolerance for ambiguity, and social desirability.
Construct validity was also gauged using correlations with
the findings of a risky gambling task and also through
testing gender differences. Hanoch, Johnson, and Wilke [40]
have also provided proof of the DOSPERT Scale's construct
validity by using it to illustrate that those chosen to show
high levels of risk taking in one content area can be risk
averse in other risky domains. The DOSPERT scale has
been observed to possess high internal consistency
(Cronbach's α=0.71–0.86) and moderate test-retest
reliability estimates [39]. Results of the study by Buelow
and Brunell [41] showed moderate to high Internal
consistency (α = 0.65 –0.89) for the scale. The present study
resulted in α = 0.70 to 0.79.
Strength of Religious Faith Questionnaire (SRFQ):
Respondents’ conviction of religious faith is assessed using
the instrument adapted and rewritten using the following
religiosity questionnaires:
a) Muslim Religiosity-Personality Inventory (MRPI) [42],
b) Hoge’s Intrinsic Religiosity Scale (IRS) [43] and c) Age
Universal I/E-Revised Scale (I/E-R) [44].
The instrument has 12 items on a 5-point scale ranging
from 1 (strongly disagree) to 5 (strongly agree), with three
items being reverse-scored. Higher scores are indication of
higher levels of religiosity. Because the instrument does not
contain references to any specific religious orientation, it
may be utilized with students of all religious affiliations as
well as for those without any interest in or affiliation with
religious traditions and perspectives. Sample of items includes:
“In my life, I experience the presence of the Divine (i.e.,
God)” and “I try to understand the meaning of religious
words/verses”. Since SRFQ was adapted to fit the study,
factor analysis was developed for the newly-adapted SRFQ
for the purpose of determining the number of distinct
constructs required to explain the pattern of correlations
among a set of measures [45]. The constructs identified are:
Level of people’s consciousness of God (LPCG) and
Faith-based religious values (FBRV). For the process of
validating the newly adapted Strength of Religious Faith
Questionnaire (SRFQ) instrument, and to prove the validity,
after revision, rewording, and approval for completeness by
the first expert, copies of the scale were given to three other
experts. They studied and certified that the scale was good
enough to measure what it intended to measure. The feedback
received from the expert review panel was used to revise,
re-word, and improve the content of the adapted survey
instrument. The present study resulted in α = 0.80 to 0.93.
B. Statistical Tools
The research hypotheses have been tested using partial least
squares-structural equation modeling (PLS-SEM) to assess
both the measurement and structural models. This study uses
SmartPLS 3.0 [46]. The use of PLS is justified for the
following reasons: 1) research objective would be prediction
of the dependent variables rather than confirmation of
structural relationships; 2) this study uses latent variable
scores as indicators of the second-order construct 3) data in
this study exhibit skewness and non-normality that is not an
issue when the study uses PLS [47]-[49]. All variables used in
the study are reflective. Reflective indicators are seen as
functions of the latent construct, and changes in the latent
construct are expressed in the changes in the indicator
(manifest) variables [50].
In line with what has been suggested by Becker, Klein, and
Wetzels [51] and Hair, Ringle, and Sarstedt [52], and having
the multi-dimensional reflective-reflective constructs, a
two-stage approach was used in the study. The construct
scores of the first-order constructs will be estimated in a
first-stage model in the absence of the second-order construct.
The model subsequently utilizes these first-stage construct
scores to indicate the higher order latent variable in a separate
second-stage analysis [51].
IV. ANALYSIS AND FINDINGS
A. Descriptive Analysis
Participants ranged in age from 22 to 51 years (M = 30.84,
SD = 6.23). Table I summarizes the demographic profile of
respondents. Gender was relatively evenly distributed and the
sample included 130 (50.8%) males, while their female
counterparts comprised 48.8% (n = 125). Only a single
student introduced himself as transgender. A total of 129
(50.4%) of them were master and 127 (49.6%) PhD students.
The majority (226) of participants were heterosexual (88.3%),
16 (6.3%) bisexual, and 10 (3.9%) were uncertain about their
International Journal of Social Science and Humanity, Vol. 7, No. 3, March 2017
145
sexual orientation. Only 4 (1.6%) respondents indicated that
they were homosexual and were sexually attracted to members of the same sex.
TABLE I: DEMOGRAPHIC PROFILE OF RESPONDENT
Profile of respondents
Study level
Master
PhD
129
127
Percentage %
50.4
49.6
Gender
Male
Female
Transgender
130
125
1
Percentage %
50.8
48.8
.4
Sexual orientation
Heterosexual
Bisexual
Uncertain
Homosexual
226
16
10
4
Percentage %
88.3
6.3
3.9
1.6
N = 256
B. Assessment of the Model Using PLS-SEM
The assessment of a model in PLS-SEM requires a
two-step process that involves measuring the
outer/measurement and inner/structural model [53]-[55].
The initial step in PLS analysis is to analyze the
measurement model to measure loading of indicators
(specific questions) on the theoretically defined constructs.
The structural model assesses the relationships between the
constructs through path analysis [53], [54].
C. Measurement Model
To assess the measurement model a two-step analysis was
conducted. Initially, the first-order factors were analyzed for
all constructs. According to Chin [53], and Hair et al. [54], to
define the reliability and validity of the reflective
measurement, the Composite Reliability (CR) and Average
Variance Extracted (AVE) were used.
To establish reliability of the reflective measurement
model, indicator reliability and construct reliability were
assessed. In assessing a model's indicator reliability, the
loading of each indicator must be higher than 0.7 to be
considered acceptable [54]. A loading lower than 0.4
indicates that an item should be considered for removal, and
items with a loading of 0.4-0.7 should be considered for
removal if their removal increases the CRs and AVEs above
the threshold [53], [47].
Internal consistency reliability was established (Table II)
because:
1) All the Cronbach’s alpha values for the constructs were
between 0.62 and 0.96 showing reliability of the
measures or survey instrument.
2) Composite reliabilities (CRs) exceeded 0.8, suggesting
that the scale items for the constructs are reliable.
Indicator reliability and convergence validity were
established because:
3) Examining the loadings for 10 constructs, 40 of the
items had loadings of 0.56 to 0.92. Items listed in Table
III were removed to raise the AVEs and CRs achieved
acceptable convergent validity [47], [53].
All AVEs values were greater than 0.5 and considered
acceptable except items reported in Table III.
Discriminant validity of the latent variables was
established:
1) Using the Heterotrait-Monotrait Ratio (HTMT), variables
showed acceptable discriminant validity (HTMT of
below 0.85) according to the HTMT.85 criterion [56]
(Table IV);
2) Using the Fornell and Larcker criterion:
The AVE of each latent construct was higher than the
construct’s highest squared correlation with any other latent
construct [53], [57]. See Table V. Moreover, all loadings were
high and cross-loadings were low in comparison with the
loadings [53], [54], [58].
Table VI shows the results of the assessment of religiosity,
risk taking, and cybersex as second order reflective
constructs.
The results presented in Tables VI and VIII indicate that the
reliability, convergent validity, and discriminant validity for
the three second order constructs met acceptability criteria.
D. Structural Model
Two criteria should be considered and interpreted in using
PLS-SEM: the path coefficients and the R2 measure for the
endogenous constructs [53], [54]. Structural model
assessment determines some results such as the variance
explanation of endogenous constructs, effect sizes, and
predictive relevance [49]. In accordance with Hair et al. [2014)
the study takes 5,000 resamples to determine statistical
significance.
Before looking at the results of the path model,
Standardized Root Mean Square Residual (SRMR) criterion
should be reported for model validation purposes [59]. In this
study SRMR is 0.058 (below cut-off of 0.08) indicating a
satisfying overall goodness of model fit.
Using PLS algorithm R2 measure was calculated for the
endogenous latent variable cybersex engagement to measure
the model's predictive accuracy (Fig. 1). R2 values of 0.67,
0.33, and 0.19 are considered substantial, moderate, and weak
respectively by Chin [53]. R2 measure was 0.22 for cybersex
engagement. R2 = 0.220 is considered relatively high by
International Journal of Social Science and Humanity, Vol. 7, No. 3, March 2017
146
behavioral research [47], [57]. To avoid bias toward
complex models, Hair et al. [47] suggest the criterion adjusted R2 needs to be calculated. R2adj for cybersex was
0.214.
T
ABLE II: RESULTS OF ASSESSMENT OF MEASUREMENT MODEL FOR FIRST ORDER CONSTRUCTS
Construct
Item
Factor
Loading
CR
Cronbach's
alpha
AVE
Religiosity-LPCG
0.968
0.962
0.789
REL11
0.865
REL12
0.921
REL13
0.920
REL14
0.917
REL15
0.858
REL16
0.887
REL17
0.864
REL18
0.871
Religiosity-FBRV
0.949
0.928
0.823
REL21
0.934
REL22
0.916
REL23
0.899
REL24
0.878
RISK Social
0.849
0.786
0.492
RISK-S1
0.763
RISK-S2
0.560
RISK-S4
0.711
RISK-S5
0.866
RISK-S6
0.759
RISK Health/Safety
0.775
0.652
0.371
RISK-HS2
0.700
RISK-HS3
0.634
RISK HS6
0.664
RISK Ethical
0.792
0.683
0.392
RISK-E2
0.624
RISK-E3
0.705
RISK-E6
0.728
Cybersex - IOSB
0.863
0.685
0.760
IOSB1
0.852
IOSB2
0.891
Cybersex - OSS
0.792
0.617
0.560
OSS1
0.772
OSS2
0.719
OSS3
0.752
Cybersex - OSB-S
0.831
0.749
0.497
OSB-S1
0.758
OSB-S3
0.670
OSB-S4
0.693
OSB-S5
0.743
Cybersex- OSB-I
0.878
0.815
0.644
OSB-I1
0.759
OSB-I2
0.820
OSB-I3
0.796
OSB-I4
0.832
Cybersex- OSC
0.829
0.753
0.449
OSC1
0.745
International Journal of Social Science and Humanity, Vol. 7, No. 3, March 2017
147
Construct
Item
Factor
Loading
CR
Cronbach's
alpha
AVE
OSC3
0.716
OSC4
0.667
OSC6
0.724
Using one-tailed P values for causal links in the model,
path coefficients have been calculated [59]. With regard to
the religiosity construct, results show that this variable
contributes to a significant negative effect on cybersex
engagement (β = -0.224, p = 0.001) and positive and
significant path confirmed for the relation of risk taking and
cybersex engagement (β = 0.351, p = <0.01).
Using Blindfolding in SmartPLS [49], [52], the
cross-validated redundancy measure (Q2) was calculated to
be 0.155, thus giving an acceptable predictive validity of
exogenous latent variable which is well above zero,
indicating the predictive relevance of the PLS path model.
Values “greater than zero imply that the exogenous constructs
have predictive relevance for the endogenous construct under
consideration [48].
Finally, the effect size for each path model can be obtained
by calculating Cohen’s f2, where 0.02, 0.15, and 0.35 have
been suggested as small, moderate, and large effects
respectively [60]. The f2 presented in Table IX indicates
values of 0.060 and 0.15, which are considered as small and
moderate respectively. Therefore the results show that the
effect of risk taking on cybersex engagement was higher than
the effect of religiosity on cybersex engagement.
TABLE III: REMOVED INDICATORS FROM THE MEASUREMENT MODEL
Construct
Removed Items
RISK Social
RISK-S3
RISK Health/Safety
RISK-HS1, RISK-HS4, RISK-HS5
RISK Ethical
RISK-E1, RISK-E4, RISK-E5
Cybersex- OSB-S
OSB-S2
Cybersex- OSC
OSC2, OSC5
Cybersex-OTHERES
OTHERS1-5
.
TABLE IV: DISCRIMINANT VALIDITY ASSESSMENT HETEROTRAIT-MONOTRAIT RATIO (HTMT.85) FOR FIRST ORDER CONSTRUCTS
HTMT
1
2
3
4
5
6
7
8
9
10
1
Religiosity-LPCG
2
Religiosity-FBRV
0.327
3
RISK Social
0.223
0.128
4
RISK Health/Safety
0.270
0.222
0.313
5
RISK Ethical
0.231
0.124
0.363
0.806
6
Cybersex- IOSB
0.431
0.155
0.200
0.308
0.401
7
Cybersex - OSS
0.328
0.136
0.133
0.455
0.539
0.734
8
Cybersex - OSB-S
0.300
0.090
0.202
0.349
0.541
0.746
0.775
9
Cybersex - OSB-I
0.347
0.071
0.174
0.298
0.381
0.686
0.728
0.768
10
Cybersex - OSC
0.393
0.083
0.152
0.409
0.479
0.789
0.813
0.789
0.831
International Journal of Social Science and Humanity, Vol. 7, No. 3, March 2017
148
V. DISCUSSION
This study was the first of its kind to examine the
influence of religiosity and risk taking on engagement in
cybersex among university students in Malaysia with
analysis done using PLS-SEM.
Although previous studies might not have examined
higher order constructs of variables used in this study, in
general, the results of the current study are consistent with
the findings of previous studies. The result of strong
negative effect of religiosity on cybersex engagement is in
line with [22], [23]. The result of influence of risk taking on
cybersex engagement was consistent with the study of
Prause and Finn [32], and Young [33] that found risky
behaviors have been positively associated with more frequent
engagement in cybersex.
This study used a multi-dimensional scale to assess the
relationship between religiosity, risk taking and cybersex
engagement. While previous studies have also shown the
association of religiosity and online sexual activities [23-26],
and risk taking and cybersex [32], [33], few studies have
developed an integrated predictive model using higher order
constructs. This modeling approach reduces model
complexity, achieves theoretical parsimony, and can avoid
multicollinearity [47], [61]. Therefore, in summarizing the
results of this study, it can be said that the students’
religiousness is a protective factor against cybersex
engagement and more frequent engagement in cybersex has
been positively associated with risk taking propensity.
TABLE V: RESULT OF DISCRIMINANT VALIDITY FOR MODEL (FORNELL AND LARCKER CRITERION) FOR FIRST ORDER CONSTRUCTS
1
2
3
4
5
6
7
8
9
10
Religiosity-LPCG
0.888
2
Religiosity-FBRV
0.312
0.907
3
RISK Social
-0.198
-0.108
0.743
4
RISK Health/Safety
-0.151
-0.149
0.256
0.729
5
RISK Ethical
-0.182
-0.083
0.496
0.217
0.757
6
Cybersex- IOSB
-0.347
-0.119
0.124
0.212
0.263
0.872
7
Cybersex - OSS
-0.260
0.022
0.015
0.289
0.339
0.503
0.749
8
Cybersex - OSB-S
-0.248
-0.046
0.159
0.221
0.358
0.536
0.544
0.739
9
Cybersex - OSB-I
-0.305
0.021
0.127
0.212
0.275
0.517
0.544
0.616
0.802
10
Cybersex - OSC
-0.330
-0.031
0.117
0.280
0.321
0.567
0.561
0.586
0.652
0.751
NOTE: SQUARE ROOTS OF AVES SHOWN DIAGONALLY IN BOLDFACE
TABLE VI: ASSESSMENT OF MEASUREMENT MODEL FOR SECOND-ORDER CONSTRUCTS
Construct
Item
Factor Loading
CR
AVE
Religiosity
0.741
0.610
LPCG
0.973
FBRV
0.522
Risk Taking
0.783
0.555
Social
0.527
Health/Safety
0.802
Ethical
0.863
Cybersex
0.902
0.649
IOSB
0.773
OSS
0.771
OSB-S
0.826
OSB-I
0.810
OSC
0.847
TABLE VII: DISCRIMINANT VALIDITY ASSESSMENT (FORNELL AND LARCKER CRITERION) FOR SECOND-ORDER CONSTRUCTS
1
2
3
1
Religiosity
0.781
2
RISK
-0.241
0.745
3
Cybersex
-0.343
0.397
0.806
International Journal of Social Science and Humanity, Vol. 7, No. 3, March 2017
149
TABLE VIII: RESULTS OF HETEROTRAIT–MONOTRAIT RATIO (HTMT.85) ANALYSIS FOR SECOND-ORDER CONSTRUCTS
HTMT
1
2
3
1
Religiosity
2
RISK
0.461
3
Cybersex
0.416
0.521
TABLE IX: RESULTS OF HYPOTHESIS TESTING
Hypothesis
Path
coefficient
t
value
P value
Effect size
Supported
Religiosity Cybersex
-0.224
3.051
0.001
0.059
Yes
Risk taking Cybersex
0.351
4.574
<0.01
0.145
Yes
Note: all relationships are calculated one-tallied, statistical significance was set at 5 %.
Fig. 1. Assessment of structural model.
VI. IMPLICATIONS AND FUTURE RESEARCH
In this study, we investigated the effects of factors
influencing residents' engagement in cybersex at
postgraduate level in selected Malaysian universities. We
used a framework to conceptualize the relationships
between religiosity, risk taking, and cybersex engagement.
However, studies in this regard has been widely criticized in
terms of their incompleteness [12], [14]. Therefore, in
acknowledging these insufficiencies, we have applied a
predictive knowledge-based conceptual framework. This
study is a pioneering work in examining variables used in an
integrated model using higher order reflective constructs.
Furthermore, this study was the first attempt to explore these
relationships employing the powerful PLS-SEM statistical
method, which is well suited for model development.
The findings of this study also provide and confirm
knowledge based on the literature of religiosity and risk
taking propensity as individual variations in the likelihood
of engaging in cybersex. Indeed, the present study reinforces
and highlights the importance of addressing research to
examine these variables. The results of this study have some
important practical implications for counselors to encourage
constructive discussion of the topic in order to minimize its
harmful consequences. Considering religiosity was shown
to play a significant role against cybersex, belief systems
from the students’ respective religious traditions can serve in
helping them in both prevention and therapy.
Notwithstanding, this study is not without its limitations.
The present study relied on self-report questionnaires that
measure perceptions and intentions of individuals at a single
point in time. The fact that perceptions and intentions change
over time as individuals gain experience, point to the need for
a longitudinal study to validate the findings. Future research
may investigate other combinations of individual differences
such as psychological triggers of trauma that may potentially
mediate the structural relationship in this study.
REFERENCES
[1] I. Xiao, C. Trestian, and A. Kuzmanovic, “A glance at an overlooked
part of the world wide web,” presented at the 22nd International
Conference on World Wide Web Companion, Rio de Janeiro, Brazil,
May 13–17, 2013.
[2] D. D. Rimington, “Examining the perceived benefits for engaging in
cybersex behavior among college students,” Master dissertation, Utah
State Univ, 2008.
[3] N. M. Döring, “The internet’s impact on sexuality: A critical review of
15 years of research,” Computers in Human Behavior, vol. 25,
pp.1089-1101, Sept.2009
[4] M. B. Short, L. Black, A. H. Smith, C. T Wetterneck, and D. E Wells,
“A review of Internet pornography use research: Methodology and
content from the past 10 years,” Cyberpsychology, Behavior, and Social
Networking,” vol. 15, pp. 13-23, Jan. 2012.
[5] Z. C. Valkyrie, “Cybersexuality in MMORPGs: Virtual sexual
revolution untapped,” Men and Masculinities, vol.14, pp. 76-96, April
2011.
[6] K. Shaughnessy, S. Byers, and S. J. Thornton, “What is cybersex?
Heterosexual students’ definitions,” International Journal of Sexual
Health, vol. 23, pp. 79-89, Jun. 2011.
[7] N. M. Döring, “Internet sexualities,” International Handbook of Internet
Research, ch. 10, pp. 171-185, 2010.
[8] W. L. Yarber, B. W. Sayad, and B. Strong, Human Sexuality: Diversity
in contemporary America, 8th ed. McGraw-Hill, 2013.
[9] N. M Almadhoun, D. Dominic, P. Dhanapal, and L. F. Woon,
“Perceived security, privacy, and trust concerns within social
networking sites: The role of information sharing and relationships
development in the Malaysian higher education institutions’ marketing,”
presented at the IEEE International Conference on Control System,
Computing and Engineering, Penang, Malaysia, 426-431, 2011.
[10] W. Y. Low, “Malaysian youth sexuality: Issues and challenges,”
Journal of the University of Malaya Medical Centre (JUMMEC),” vol.
12, pp. 3-14. Jun.2009.
[11] M. Z. B. Zakaria and D. F. B Baharuddin, “Cybersex addiction treatment:
A Malaysian perspective on the needs for counselors’ training,”
presented at the PERKAMA International Conference, Bandung,
Indonesia1-15, 2011.
[12] K. Shaughnessy and E. S. Byers, “Contextualizing cybersex experience:
Heterosexually identified men and women’s desire for and experiences
with cybersex with three types of partners,” Computers in Human
Behavior, vol. 32, pp. 178-185, Mar. 2014.
[13] C. H. Hackney, and G. S. Sanders, “Religiosity and mental health: A
meta-analysis of recent studies,” Journal for the Scientific Study of
Religion, vol. 42, pp. 43-56, Mar. 2003.
International Journal of Social Science and Humanity, Vol. 7, No. 3, March 2017
150
[14] O. Hesapçı Sanaktekın, Y. Aslanbay, and V. Gorgulu, “The effects of
religiosity on internet consumption: A study on a Muslim country,”
Information, Communication & Society, vol. 16, pp. 1553-1573, Sep.
2012.
[15] H. A. Campbell, “Religion and the internet: A microcosm for studying
internet trends and implications,” New Media and Society, vol. 15, pp.
680-694, Nov. 2013.
[16] G. K. Brown, “Making ethnic citizens: The politics and practice of
education in Malaysia,” International Journal of Educational
Development, vol. 27, pp. 318-330, May. 2007.
[17] Department of Statistics Malaysia. Population by states and ethnic
group. [Online]. Available:
http://www.statistics.gov.my/portal/index.php?option=com_content
&view=article&id=471&lang=en&Itemid=0
[18] Y. A. Momtaz, T. Hamid, R. Ibrahim, N. Yahaya, and S. T. Chai,
“Moderating effect of religiosity on the relationship between social
isolation and psychological well-being,” Mental Health, Religion &
Culture, vol.14, pp. 141-156, Jul. 2010.
[19] T. E. Huff, “Globalization and the internet: Comparing the Middle
Eastern and Malaysian experiences,” The Middle East Journal, vol.55,
pp. 439-458, 2001.
[20] H. A. Campbell, Exploring Religious Community Online: We Are One
in the Network, New York: Peter Lang, 2005.
[21] R. Kluver and P. H. Cheong, “Technological modernization, the
internet, and religion in Singapore “Journal of Computer-Mediated
Communication, vol. 12, pp. 1122-1142. Jun. 2007.
[22] A. Baltazar, H. W. Helm, D. McBride, G. Hopkins, and J. J. Stevens,
“Internet pornography use in the context of external and internal
religiosity,” Journal of Psychology & Theology, vol. 38, pp. 32-40,
2010.
[23] J. N. Thomas, “Outsourcing moral authority: The internal
secularization of evangelicals’ anti-pornography narratives,” Journal
for the Scientific Study of Religion, vol.52, pp. 457-475. Sep. 2013.
[24] J. S. Carroll, L. M Padilla-Walker, L. J. Nelson, C. D. Olson, C.
McNamara Barry, and S. D. Madsen, “Generation XXX: Pornography
acceptance and use among emerging adults,” Journal of Adolescent
Research, vol. 23, pp. 6-30, Jan. 2008.
[25] P. J.Wright, R. S.Tokunaga, and S. Bae, “Pornography consumption
and US adults' attitudes toward gay individuals' civil liberties, moral
judgments of homosexuality, and support for same-sex marriage:
Mediating and moderating factors,” Communication Monographs, vol.
81,pp. 79-107, Feb. 2014.
[26] J. Grubbs, J. Exline, K Pargament, J Hook, and R. Carlisle,
“Transgression as addiction: Religiosity and moral disapproval as
predictors of perceived addiction to pornography,” Archives of Sexual
Behavior, vol. 144, pp. 125-136. Feb. 2014.
[27] K. Greene, M. Krcmar, L. H. Walters, D. L. Rubin, and L. Hale,”
Targeting adolescent risk-taking behaviors: The contributions of
egocentrism and sensation-seeking," Journal of Adolescence, vol. 23,
pp. 439-461, Aug. 2000.
[28] M. Gerrard, F. X. Gibbons, A. F. Houlihan, M. L. Stock, and E. A.
Pomery,” A dual-process approach to health risk decision making:
The prototype willingness model,” Developmental Review, vol. 28, pp.
29 – 61, Mar. 2008.
[29] J. T. Parsons, A. W. Siegel, and J. H. Cousins, “Late adolescent
risk-taking: Effects of perceived benefits and perceived risks on
behavioral intentions and behavioral change,” Journal of Adolescence,
vol. 20, pp. 381-392, Aug. 1997.
[30] D. Eisenberg, E. Golberstein, and J. L Whitlock, “Peer effects on risky
behaviors: New evidence from college roommate assignments,”
Journal of Health Economics, vol. 33, pp. 126-138, Jan. 2014.
[31] H. Szrek, L. Chao, S. Ramlagan, and K. Peltzer, “Predicting (un)
healthy behavior: A comparison of risk-taking propensity measures”
Judgment and Decision Making, vol. 7, pp. 716-727, Nov. 2012.
[32] D. Ley, N. Prause, and P. Finn, “The emperor has no clothes: A review
of the ‘Pornography addiction’ model,” Current Sexual Health
Reports, vol. 1, pp. 94-105, Feb. 2014.
[33] K. S. Young, “Internet sex addiction: Risk factors, stages of
development, and treatment,” American Behavioral Scientist, vol. 52,
pp. 21-37. Sep. 2008.
[34] W. W. Chin, “The partial least squares approach to structural equation
modeling,” in Modern Business Research Methods, G. A.
Marcoulides, Ed., Mahwah, NJ: Lawrence Erlbaum Associates, 1998,
ch. 10, pp. 295-336.
[35] D. L. Delmonico and J. Miller, “The internet sex screening test: A
comparison of sexual compulsives versus non-sexual compulsives,"
Sexual and Relationship Therapy, vol. 18, pp. 261-276, 2003.
[36] J. Grubbs, F. Volk, J. J. Exline, and K. I. Pargament, “Internet
pornography use: Perceived addiction, psychological distress, and the
validation of a brief measure,” Journal of Sex and Marital Therapy, vol.
41, pp.83-106, Dec. 2013.
[37] J. N. Hook, J. P. Hook., D. E. Davis, E. L. Worthington, and J. K.
Penberthy, “Measuring sexual addiction and compulsivity: A critical
review of instruments,” Journal of Sex and Marital Therapy, vol. 36, pp.
227-260, 2010.
[38] D. L. Delmonico. (1997). Internet sex screening test. [Online]. Available:
http://www.sexhelp.com
[39] A. R. Blais and E. U. Weber, “Domain-specific risk-taking (DOSPERT)
scale for adult populations,” Judgment and Decision Making, vol. 1, pp.
33-47. July 2006.
[40] Y. Hanoch, J. G. Johnson, and A. Wilke, “Domain specificity in
experimental measures and participant recruitment: An application to
risk-taking behavior,” Psychological Science, vol. 17, pp. 300-304. Apr.
2006.
[41] M. T. Buelow and A. B. Brunell, “Facets of grandiose narcissism predict
involvement in health-risk behaviors,” Personality and Individual
Differences, vol. 69, pp. 193-198, June 2014.
[42] S. E. Krauss, “The Muslim religiosity personality inventory (MRPI)
scoring manual,” Institute for Social Science Studies, 2011.
[43] R. Hoge, “A validated intrinsic religious motivation scale,” Journal for
the Scientific Study of Religion, vol.11, pp. 369-376, Dec.1972.
[44] R. L. Gorsuch and S. E. McPherson, “Intrinsic/extrinsic measurement:
I/E-revised and single-item scales,” Journal for the Scientific Study of
Religion, vol. 28, pp. 348-354, Sep. 1989.
[45] L. R. Fabrigar and D. T. Wegener, Exploratory Factor Analysis, Oxford
University Press, 2011.
[46] C. M. Ringle, S. Wende, and J. M. Becker. SmartPLS 3. Boenningstedt:
SmartPLS GmbH, [Online]. Available: http://www.smartpls.com
[47] J. F. Hair, G. T. M. Hult, C. M. Ringle, and M. Sarstedt, A Primer on
Partial Least Squares Structural Equation Modeling (PLS-SEM),
Thousand Oaks: Sage, 2014.
[48] M. Sarstedt, C. M. Ringle, J. Henseler, and J. F. Jr. Hair, “On the
emancipation of PLS-SEM: A commentary on rigdon 2012,” Long
Range Planning, vol. 47, pp. 154-160, June. 2014.
[49] J. Henseler, C. M. Ringle, and R. R. Sinkovics, “The use of partial least
squares path modeling in international marketing,” Advances in
International Marketing (AIM), vol. 20, pp.277-320, 2009.
[50] A. Diamantopoulos, “The error term in formative measurement models:
Interpretation and modeling implications,” Journal of Modelling in
Management, vol.1, pp. 7-17, 2006.
[51] J. Becker, K. Klein, and M. Wetzels, “Hierarchical latent variable
models in PLS-SEM: Guidelines for using reflective-formative type
models,” Long Range Planning, vol. 45, pp. 359-394, Oct-Dec., 2012.
[52] J. F. Hair, C. M. Ringle, and M. Sarstedt, “Editorial-partial least squares
structural equation modeling: Rigorous applications, better results and
higher acceptance,” Long Range Planning, vol. 46, pp. 1-12, Mar. 2013.
[53] W. W. Chin, “Bootstrap cross-validation indices for PLS path model
assessment,” in Handbook of Partial Least Squares; Concepts, Methods
and Applications, V. Esposito Vinzi, W. W. Chin, J. Henseler and H.
Wang eds, Verlag Berlin Heidelberg: Springer, 2010, ch. 3, pp. 83-97.
[54] J. F. Hair, C. M. Ringle, and M. Sarstedt, “PLS-SEM: Indeed a silver
bullet,” Journal of Marketing Theory and Practice, vol. 19, pp. 139-152,
2011.
[55] J. F. Hair, M, Sarstedt, T. M Pieper, and C. M. Ringle, “The use of
partial least squares structural equation modeling in strategic
management research: A review of past practices and recommendations
for future applications,” Long Range Planning, vol. 45, pp. 320-340,
Oct-Dec. 2012.
[56] J. Henseler, C. M. Ringle, and M. Sarstedt, “A new criterion for
assessing discriminant validity in variance-based structural equation
modeling,” Journal of the Academy of Marketing Science, vol. 1, pp.
115-135, Jan. 2015.
[57] N. Kock, WarpPLS 2.0 User Manual, Laredo, TX: ScriptWarp Systems,
2013.
[58] C. Fornell and D. F Larcker, “Evaluating structural equation models
with unobservable variables and measurement error,” Journal of
Marketing Research, vol. 81, pp. 39-50. Feb.1981.
[59] J. Henseler et al., "Common beliefs and reality about PLS comments on
rönkkö and evermann, 2013,” Organizational Research Methods, vol.
17 pp. 82-209, Apr, 2014.
[60] J. Cohen, Statistical Power Analysis for the Behavioral Sciences,
Hillsdale, New Jersey: Lawrence Erlbaum. 1988.
[61] C. M. Ringle, M. Sarstedt, and D. W. Straub, “Editor’s comments: A
critical look at the use of PLS-SEM in MIS quarterly,” MIS Quarterly,
vol. 36, pp. iii-xiv. Mar. 2012.
International Journal of Social Science and Humanity, Vol. 7, No. 3, March 2017
151
Soudabeh Ghoroghi is a PhD candidate at University of Putra Malaysia
(UPM), Malaysia. She holds a master of science degree in counseling
psychology from UPM, and a bachelor of clinical psychology degree from
Tehran University, Iran.
She has worked as research assistant since 2013. She also has work
experience as a social worker and counselor since 1994. Her areas of
research mainly include family and marital related issues, behavioral
addiction and cybersex addiction.
Siti Aishah Hassan has become a lecturer in the
Faculty of Educational Studies, University Putra
Malaysia UPM since 2006. She holds her doctoral
and master degrees from the International Islamic
University Malaysia in Guidance and Counseling and
a bachelor of Chemistry degree from University of
Missouri St. Louis, USA.
She has organized and chaired workshop and
conduct trainings to educate the licensed counselors
and postgraduate students on systemic theories and practices and has served
as content expert in family counseling and research consultant especially on
Structural Equation Modeling (SEM-AMOS) to several other agencies.
Dr. Hassan is actively involved in various research projects, especially
those related to Marital, Couple and Family Counseling such as cybersex
and intervention, marital conflict, adjustment and satisfaction, parental
spirituality and attachment, maternal spiritual characteristics and quality
time, and parent-teacher collaborative models. She has been involved in ten
local and international non-profit organization and charitable works. Some
of them are: a member of Malaysian Counseling Association (PERKAMA),
Malaysian Psychometrics Association (MPA), American Counseling
Association (ACA), Asian Psychological Association (ApsyA), and Asia
Pacific Counselor Association (APECA), and a board member of
Association for Marriage and Family Therapy Malaysia (AMFTM).
Ahmad Fauzi Mohd Ayub is an associate professor at
Faculty of Educational Studies, University Putra
Malaysia. He hold a PhD degree in educational
technology from National University of Malaysia in
2008.
His major field is in information and communication
technology in education and mathematics education.
He has been working as matriculation teacher since
1994 and joined as lecturer in 2002. As an associate
prof, he has taught many courses mainly related to information and
communication technology in education, leading to the Bachelor of Education
(Information Technology) degree. The courses he conducted involved both
theoretical concepts and practical laboratory work. Besides that, he also taught
other subjects such as testing and measurement courses and research methods.
He has had hundred articles published in 6 cited journals, non-cited journals
chapters of books and conferences proceeding both at international and
national level. Besides being books editor and Editorial Committee for
conference proceedings.
International Journal of Social Science and Humanity, Vol. 7, No. 3, March 2017
152
