Predicting Compulsive Internet Use: It’s All about Sex!
GERT-JAN MEERKERK, M.A., REGINA J.J.M. VAN DEN EIJNDEN, Ph.D., and
HENK F.L. GARRETSEN, Prof. Dr.
The objective of this research was to assess the predictive power of various Internet applica-
tions on the development of compulsive Internet use (CIU). The study has a two-wave longi-
tudinal design with an interval of 1 year. The first measurement contained 447 adult heavy
Internet users who used the Internet at least 16 h per week and had Internet access at home
for at least 1 year. For the second measurement, all participants were invited again, of whom
229 responded. By means of an online questionnaire, the respondents were asked about the
time spent on various Internet applications and CIU. On a cross-sectional basis, gaming and
erotica seem the most important Internet applications related to CIU. On a longitudinal basis,
spending a lot of time on erotica predicted an increase in CIU 1 year later. The addictive po-
tential of the different applications varies; erotica appears to have the highest potential.
CYBERPSYCHOLOGY & BEHAVIOR
Volume 9, Number 1, 2006
© Mary Ann Liebert, Inc.
THE CONSTRUCT of compulsive Internet use
(CIU)—also referred to as Internet addiction,1,2
Internet dependence,3,4 problematic Internet use,5,6
or pathological Internet use,7,8 has gained consid-
erable acceptance within the last decade. Since the
first parodying report by Goldberg,9some 10 years
ago, increasing numbers of studies have addressed
the phenomenon that certain persons use the Inter-
net compulsively, which can lead to serious prob-
lems with regard to psychosocial and professional
functioning. Most commonly, the behavior is re-
ferred to as “Internet addiction,” suggesting that it is
the Internet in itself that is addictive, rather than the
actual application with which the user is involved.
On the other hand, several researchers have differ-
entiated between various forms of CIU. Young et
al.,10,11 for example, conducted a survey among 35
therapists who have treated clients suffering from
cyber-related problems. Qualitative results gleaned
from the study suggest that five specific subtypes of
CIU can be categorized: cyber-sexual addiction,
cyber-relationship addiction, net compulsion (obses-
sive online gambling, shopping, or day trading), in-
formation overload (compulsive web surfing or
database searches), and computer addiction (obses-
sive computer game playing). Similarly, Davis7dis-
tinguishes (in his Cognitive-Behavioural Model of
Pathological Internet Use) specific pathological In-
ternet use (PIU) and generalized PIU, where the for-
mer refers to pathological use of the Internet for a
particular purpose (such as online sex or online
gambling), and the latter to a general, multidimen-
sional overuse of the Internet. According to Davis,
specific PIU is content-specific and exists indepen-
dent of multiple Internet functions; it would also
exist in the absence of the Internet. Generalized PIU,
on the other hand, involves a general, multidimen-
sional overuse of the Internet and may include on-
line procrastination. Generalized PIU is often
associated with chatting and related to the social as-
pect of the Internet. Davis7says, “The need for social
contact and reinforcement obtained online results in
an increased desire to remain in a virtual social life.”
Several authors have suggested that particular ap-
plications that involve social interaction constitute a
risk for developing CIU. For example, Caplan12
found in a sample of 386 undergraduate students
that the preference for social benefits available
IVO, Addiction Research Institute, Rotterdam, The Netherlands.
14192c11.pgs 2/2/06 2:41 PM Page 95
online accounted significantly for the negative out-
come of Internet use and suggested that the prefer-
ence for computer-mediated social interaction plays
a role in the etiology, development, and outcomes of
generalized PIU. Chou and Hsiao13 found in a large
sample of 910 university and college students that
the Internet Communication Pleasure Score (a mea-
sure relating, among others, to the use of the Internet
for interpersonal communication) was the most
powerful predictor of Internet addiction. Li and
Chung14 studied in a relatively small sample of 76
college students the relationship between Internet
function and Internet addictive behavior, and found
that the social function played the core role in Inter-
net addictive behavior. Ward15 studied 112 under-
graduate and graduate students and found that
communication applications were the central focus
associated with problematic use. Young16 found in a
convenience sample of 396 dependent Internet users
and a control group of 100 non-dependent Internet
users that non-dependents predominantly used
those aspects of the Internet that allowed them to
gather information (i.e., Information Protocols and
the World Wide Web) and e-mail, whereas depen-
dents predominantly used the two-way communi-
cation functions available on the Internet (i.e., chat
rooms, MUDs, news groups, or e-mail). Finally, a
longitudinal study among 663 Dutch adolescents
from our own research group showed that instant
messenger (IM) use and chatting in chat rooms were
related to increases in CIU 6 months later.17 Contrary
to these findings, Widyanto and McMurran18 found
no correlation between the type of Internet functions
and participants’ CIU in a convenience sample of 86
self-selected Internet users. In general, several stud-
ies have shown associations between the social func-
tion of the Internet and CIU; however, as far as we
know only one study used a longitudinal design, al-
lowing for more definite conclusions on the direc-
tion of causation. Therefore, the present study aims
to assess the addictive potential of the various Inter-
net applications by examining the predictive power
of the time spent on the various applications on the
development of CIU within a longitudinal design.
The results may contribute to the further under-
standing of the mechanisms behind CIU.
This study had a two-wave longitudinal design
with an interval of 1 year. The data were gathered
in the Netherlands by means of two online mea-
surements (at T1 and at T2), carried out among a
representative sample of adult and experienced
heavy users of the Internet: aged 18 years and
older, having access to the Internet at home for at
least 1 year, and using the Internet on average
16–100 h per week.19 In November 2002, partici-
pants received an email which invited them to surf
to a website where the questionnaire could be com-
pleted in about 10 min. Non-responders received
reminders after 2 and 4 weeks. At 1 year after the
first measurement, the procedure was repeated
(T2), and all respondents to the first measurement
received an email, inviting them to visit a website
to fill out an online questionnaire following the
same procedure as during the first measurement.
The online questionnaires at T1 and T2 con-
tained, among others, the following variables: de-
mographics, Internet use, and the Compulsive
Internet Use Scale (CIUS20).
Internet use was measured by asking the respon-
dents “How many days per week are you online for
private purposes?” (8-point scale: “every day” to
“less than once a week”) and “How many hours do
you spend online for private purposes on a typical
day that you use the Internet?” (8-point scale:
“seven hours or more” to “less than one hour”).
Based on these two questions, the average number
of hours per week was calculated by multiplying
the number of days per week by the number of
hours per typical day. Furthermore, the respondents
were asked how much time they spent on 11 (12 at
T2) specific Internet applications (7-point scale:
“none” to “more than 40 hours per week”)—that is,
email, searching for information on the Internet,
surfing the Internet, online gaming, chatting, buy-
ing on the Internet, gambling on the Internet, down-
loading from the Internet, Usenet, searching for
erotic stimuli (erotica), dating on the Internet, and,
at T2, participation in an online forum.
A recently developed and validated scale, the
CIUS,20 assessed CIU (see Appendix). The CIUS has
14 items on a five-point Likert scale (“Never” to
“Very often”) and scores of 0–56. The scale has a
high reliability (T1 and T2 Alpha = 0.89) and in-
cludes the aspects of loss of control, preoccupation,
withdrawal symptoms, coping, and conflict with
regard to the use of the Internet.20
Sample and non-response
Of the 1,000 participants who received an invita-
tion to volunteer in the first measurement (at T1),
96 MEERKERK ET AL.
14192c11.pgs 2/2/06 2:41 PM Page 96
447 (44.7%) responded. Because all participants
were part of an access panel, information on age,
gender, and education level was available from
previous surveys. An attrition analysis was con-
ducted to test for possible differences between the
responders and non-responders. Logistic regres-
sion analyses revealed significant differences be-
tween responders and non-responders for all three
variables. Responders were slightly older (38.5 vs.
36.9 years, OR 1.01, 95% CI 1.00, 1.02), more often
female (51% vs. 43%, OR 1.37, 95% CI 1.06, 1.77),
and slightly higher educated (4.2 vs. 4.0, OR 1.09,
95% CI 1.00, 1.18, on a 7-point scale ranging from
“Lower education” to “University education”).
For the second measurement (at T2), at 1 year
after the first measurement, all 447 respondents of
the first measurement were approached again and
invited to fill out the second online questionnaire.
About half of them (51%, n= 229) responded and
filled out the questionnaire. An attrition analysis
was conducted to test for possible differences be-
tween the responders and dropouts (n= 218). The
logistic regression analyses revealed no differences
in the demographic variables of age, gender, and
education level, nor on the score on the CIUS, score
on the OCS,21 and number of years with Internet
connection at home. A small difference was found
for number of hours online per week; responders
spent on average more time online than non-
responders (26.6 h per week vs. 24.2 h per week,
OR 1.02, 95% CI 1.00, 1.04).
To analyze the addictive potential of the different
applications, first Pearson correlation analyses
were conducted with time spent on the different
applications at T1 and T2, duration of Internet ac-
cess at home, and CIUS scores as variables. Dura-
tion of Internet access at home was included
because this variable may have an influence on CIU
as exemplified in “beginner’s fascination.” To
check for multicollinearity, correlations between
the various applications were calculated. Next,
cross-sectional predictors of CIU were determined
at T1 and T2, by conducting linear regression
analyses with CIU at T1 and T2, respectively, as de-
pendent variable, and time spent on the 11 Internet
applications as independent variables. To control
for demographic factors and duration of Internet
access at home, the demographic variables gender,
age and educational level, and access time were en-
tered in step 1 of the regression equation. To deter-
mine possible longitudinal predictors of CIU,
linear regression analyses were conducted with
CIU at T2 as dependent variable, and CIU at T1 and
time spent on the 11 Internet applications at T1, as
independent variables. Again, to control for demo-
graphic factors and duration of Internet access at
home, the demographic variables gender, age and
educational level, and access time were entered in
step 1 of the regression equation. In all analyses, p<
0.05, unless otherwise noted. Statistical program
was SPSS 12.0.
There were large differences in the time spent on
the various Internet applications. Table 1 shows
that some applications are hardly used (e.g., 97.5%
of the respondents never gambles online and 84.5%
never dates online), whereas other applications are
used by almost all respondents (e.g., email, infor-
mation searching, and surfing). Much time is spent
on email, downloading, chatting, and surfing. Re-
markable is the fact that relatively few respondents
report spending a lot of time on searching the Inter-
net for sexual stimuli, although the pursuit of
sexual interests over the Internet is reported to be
very common among Internet users.22,23 The inter-
correlations between the various applications are
generally weak (Table 2), ranging from 0.431
(email–information seeking) to near zero (e.g.,
gaming–erotica). This indicates that the applica-
tions are relatively independent from each other
and that multicollinearity will not disturb the pre-
between Internet applications and CIU
Cross-sectional Pearson correlation analyses
demonstrated large differences in the correlation
between time spent on the applications and CIU
(Table 3). Relatively high cross-sectional correla-
tions (from 0.261 to 0.203 at T1, and from 0.270 to
0.204 at T2) were found between CIU and chatting,
gaming, and, dating.
Cross-sectional linear regression analyses were
conducted to find predictors of CIU in terms of
time spent on the different applications. The results
(Table 4) showed positive associations at T1 for
gaming, chatting, and erotica. In addition, there
was a negative association between age and CIU,
indicating that the older the Internet user, the less
likely that person is to show signs of CIU. The ap-
plication factors explained 14% of the variance in
PREDICTING COMPULSIVE INTERNET USE 97
14192c11.pgs 2/2/06 2:41 PM Page 97
CIU at T1. The results indicate that the more time
spent on gaming, chatting, and erotica, the more
likely it is that the Internet user shows signs of CIU.
The same analysis conducted with T2 variables
showed a somewhat different pattern of results.
Again, positive associations were found for gaming
and erotica, but no association was found for chat-
ting; however, a positive association was found for
dating. No effects were found for the demographic
variables. The application factors explain 15% of
the variance in CIU at T2. The results of both cross-
sectional regression analyses indicate that particu-
larly gaming and erotica are associated with CIU.
In other words, those who spent a lot of time on
gaming and erotica have a higher risk to show
signs of CIU. The evidence for chatting and dating
is less evident.
Longitudinal associations between
Internet applications and CIU
The longitudinal design of the study enables to
determine predictors of CIU over a 1-year period.
First, Pearson correlation analyses were conducted
showing significant correlations between chatting,
gaming, dating, buying, and erotica at T1, and CIU
at T2 (Table 3). The results of the subsequent longi-
tudinal regression analyses are shown in Table 4.
The factors explain 61% of the variance in CIU at
T2. Not surprisingly, the strongest association was
found for the CIUS score at T1. In addition, a posi-
tive association was found for erotica. Apparently,
spending a lot of time searching for erotic stimuli
predicts an increase in CIU 1 year later. None of the
other application factors reached significance, nor
did the demographic factors add to the prediction
The main goal of the present study was to assess
the relative addiction risk of several Internet appli-
cations. First of all, large differences were found in
the popularity of the various applications. In terms
of time spent on the application, e-mailing, down-
loading, chatting, and surfing are among the most
popular. Consistently over the two measurements,
the cross-sectional analyses demonstrated that CIU
was associated with gaming and searching for
erotic stimuli. In addition, CIU was associated with
chatting at the first measurement, and dating at the
second measurement. It appears that Internet users
who spent a lot of time on particularly gaming and
erotica, are at higher risk to use the Internet com-
pulsively. The results of the longitudinal analyses
are partly in line with these conclusions and dem-
onstrated a clear association between CIU and
searching for erotic stimuli; searching for erotic
stimuli predicted an increase in CIU 1 year later.
Using the Internet for sexual gratification should
therefore be regarded as the most important risk
factor for the development of CIU.
These findings are only partly in line with the
findings of other studies, which reported that par-
ticularly applications that involve social interaction
are associated with CIU.12–17 Clearly, searching for
sexual stimuli on the Internet may involve social
98 MEERKERK ET AL.
TABLE 1. TIME SPENT ON INTERNET APPLICATIONS IN PERCENTAGE OF
RESPONDENTS AT T1
Time (in hours) spent on application per week
0 <4 5–10 11–20 >21
Email 0.2 48.8 27.4 9.8 13.8
Information 1.8 61.6 23.3 9.7 3.6
Surfing 3.4 58.8 21.9 10.0 5.9
Gaming 37.8 41.4 11.0 5.9 3.8
Chatting 30.6 36.7 13.3 8.6 10.8
Buying 42.4 54.6 2.0 0.7 0.2
Gambling 97.5 2.2 0.2 0 0
Downloading 17.4 46.4 15.6 11.1 9.5
Usenet 51.7 38.1 4.3 3.8 2.0
Erotica 65.7 28.7 3.6 1.6 0.4
Dating 84.5 12.1 1.8 1.1 0.4
T1, time 1.
14192c11.pgs 2/2/06 2:41 PM Page 98
TABLE 2. CORRELATIONS BETWEEN TIME SPENT ON INTERNET APPLICATIONS AT T1
Email Information Surfing Gaming Chatting Buying Gamble Downloading Usenet Erotica Dating
Information 0.431** 1
Surfing 0.267** 0.386** 1
Gaming 0.077 0.067 0.135* 1
Chatting 0.363** 0.202** 0.343** 0.250** 1
Buying 0.058 0.170** 0.182** 0.034 0.104 1
Gamble 0.017 0.008 0.008 0.003 0.070 0.136* 1
Downloading 0.220** 0.320** 0.222** 0.047 0.244** 0.203** 0.031 1
Usenet 0.261** 0.316** 0.128 0.027 0.076 0.162* 0.010 0.193** 1
Erotica 0.019 0.039 0.242** 0.003 0.043 0.088 0.014 0.164* 0.110 1
Dating 0.060 0.058 0.110 0.013 0.176** 0.114 0.039 0.028 0.144* 0.263** 1
T1, time 1.
14192c11.pgs 2/2/06 2:41 PM Page 99
100 MEERKERK ET AL.
TABLE 3. PEARSON CORRELATIONS BETWEEN TIME SPENT ON APPLICATIONS
AND COMPULSIVE INTERNET USE (CIU) AT T1 AND T2
Cross-sectional Cross-sectional Longitudinal
correlations with correlations with correlation with
CIU T1 CIU T2 CIU T2
Chatting 0.261** 0.223** 0.226**
Gaming 0.216** 0.204** 0.173**
Dating 0.203** 0.270** 0.158*
Email 0.199** 0.163* 0.124 n.s.
Erotica 0.189** 0.193** 0.147*
Surfing 0.171** 0.152* 0.051 n.s.
Information search 0.165** 0.127 n.s. 0.089 n.s.
Usenet 0.122* 0.099 n.s. 0.056 n.s.
Downloading 0.115* 0.048 n.s. 0.021 n.s.
Buying 0.105* 0.149* 0.155*
Gambling 0.066 n.s. 0.108 n.s. 0.044 n.s.
Forum n.a. 0.189**
Access time at home 0.088 n.s. 0.093 n.s. —
T1, time 1; T2, time 2 (1 year after T1).
TABLE 4. CROSS-SECTIONAL AND LONGITUDINAL LINEAR REGRESSION ANALYSES FOR T1 AND T2
Cross-sectional T1 Cross-sectional T2 Longitudinal T1–T2
Age 0.110* 0.045 0.044
Gender 0.031 0.046 0.032
Education 0.057 0.000 0.008
Access time 0.068 0.088 0.086
0.023 0.015 0.012
CIU T1 0.761**
Email 0.064 0.026 0.009
Information 0.084 0.002 0.021
Surfing 0.026 0.017 0.086
Gaming 0.143** 0.163* 0.043
Chatting 0.130* 0.083 0.017
Buying 0.070 0.072 0.033
Gambling 0.057 0.067 0.025
Downloading 0.022 0.015 0.042
Usenet 0.067 0.035 0.048
Erotica 0.124* 0.175* 0.132*
Dating 0.090 0.175* 0.011
Forum n.a. 0.119 n.a.
0.141 0.149 0.612
T1, time 1; T2, time 2 (1 year after T1).
14192c11.pgs 2/2/06 2:41 PM Page 100
interaction, but may also exclusively involve non-
interactive searching for pornography. Gaming
may also involve social interaction, and social in-
teraction appears to be one of the factors that moti-
vates people to continue gaming (even when no
monetary reward is involved)24,25; however, not all
gaming implies social interaction. The finding that
chatting was not always associated with CIU raises
doubts about the relationship between social inter-
action and CIU. Further research is needed to study
what qualities and aspects of social interaction con-
tribute to the addictive potential of certain Internet
The most relevant question with regard to the re-
sults of the present study relate to the how and
why of the observed addictive potential of online
sexual behavior. First, it is important to distinguish
between the various sex-related uses of the Inter-
net. Griffiths26 describes a number of different
ways the Internet can be used for sexually related
purposes—for example, seeking out sexually re-
lated material for educational use, buying or selling
sexually related goods for further use offline, seek-
ing out material for entertainment/masturbatory
purposes for use online, engaging in and maintain-
ing online relationships via email and/or chat,
seeking out sexual partners for a transitory or en-
during relationship, seeking out individuals who
then become victims of sexually related crime (e.g.,
online sexual harassment, cyber stalking), and ex-
ploring gender and identity roles. Not all of these
activities may be done to excess or are potentially
addictive; most likely using pornography for mas-
turbatory purposes, engaging in online relation-
ships, and engaging in sexually related Internet
crime may be addictive.26
The specific features of the Internet that make
sexuality on the Internet so tempting have been de-
scribed by, for example, Cooper23 and Young et al.11
Cooper’s “Triple A engine” describes three typical
features of sexual behavior via the Internet that
contribute to its tempting qualities: Access, Afford-
ability, and Anonymity. Access refers to how easy it
is to connect to the Internet and to find, with a fin-
ger click, a variety of sexually stimulating audio,
video, or text items. Moreover, these sexual stimuli
are in abundance, replenished daily, and often at no
or little charge. Most importantly, one can engage
in online sexual behavior anonymously (at least
subjectively), which lowers thresholds and fosters
disinhibition27 without having to fear negative con-
sequences. Young’s “ACE model” (Anonymity,
Convenience, and Escape), shares the anonymity
feature and stresses furthermore the convenience of
meeting others or finding sexually stimulating ma-
terial on the Internet within the safe environment
of one’s own house. In addition, Young stresses
that sexuality on the Internet can be used as a cop-
ing strategy to escape daily sorrows, or ameliorate
a negative mood. Some even experience a kind of
“high” (see also the flow experience28).
Putnam29 gives a good description of the patho-
genesis of online sexual compulsions for persons
who are vulnerable through biological, psychologi-
cal and/or social characteristics, and how the be-
havior is reinforced through operant conditioning
and classical conditioning learning mechanisms. In
brief, Putnam states that the vulnerability for the
development of compulsive sexual behavior may
originate from biological factors such as deviant
testosterone and serotonin levels, or may develop
in response to physical, sexual, family, or social
trauma. In addition, personality disorders, mood
and anxiety disorders, and substance abuse and de-
pendence may contribute to the vulnerability to de-
velop compulsive sexual behavior. These personal
factors can make a person vulnerable to develop
compulsive sexual behavior; however, the compul-
sive behavior may stay latent in the “normal” of-
fline world. The unique factors of the Internet (as
described above as the Triple A engine and the ACE
model) may trigger the latent compulsive behavior
to become manifest when a predisposed person en-
gages in sexual behavior on the Internet. Through
operant conditioning, the online sexual behavior
increases in frequency and duration. The sexual
arousal (possibly followed by masturbation and or-
gasm) serves as a positive reinforcer, and the dis-
traction from negative mood states (coping) serves
as a negative reinforcer. The reinforcement may be
particularly strong due to the variable-ratio sched-
ule of reinforcement. Eventually, classical condi-
tioning occurs when the online sexual behavior is
repeated and computer use is paired to sexual
arousal. As a result, using the computer may elicit
craving to engage in online sexual behavior.29
The above makes plausible the notion that for
some vulnerable persons the specific qualities of
the Internet facilitates the development of sexual
compulsive behavior or a sex addiction. Indications
for this personal sensitivity or vulnerability may be
found in the psychosocial problems like loneliness,
low self-esteem, or depressive symptoms, often re-
lated to CIU.6,30–33 In line with this are the state-
ments of, for example, Shaffer et al.34,35 and
Griffiths,36 who proclaim that CIU merely reflects
other forms of psychopathology. This reasoning
also shows that the term “Internet addiction” is in-
appropriate and misleading, as it is not the Internet
in itself that is addictive, but the specific applica-
PREDICTING COMPULSIVE INTERNET USE 101
14192c11.pgs 2/2/06 2:41 PM Page 101
tion (e.g., searching for sexual stimuli). However, it
is the Internet that is used compulsively to perform
these behaviors, which legitimates the use of the
term “Compulsive Internet Use.”
Finally, some limitations of the present study need
addressing. The first shortcoming concerns the divi-
sion of the Internet activities into 11 (12 at T2) appli-
cations. Although carefully constructed, the division
did not result in 100% unique and non-overlapping
applications. For example, searching for sexual stim-
uli may include solistic activities like searching for
erotic pictures, as well as engaging in a more
interaction-oriented activity like maintaining an
erotic online relationship through chat or other on-
line communication channels. Furthermore, chatting
was presented as one application, not differentiating
between online communication with total strangers
in public chat rooms, and online communication
with friends through the use of messengers like
MSN or Yahoo Messenger. Therefore, it is recom-
mended that future research should further differen-
tiate between the various applications to identify
which aspects of Internet use are potentially addic-
tive. A second limitation regards the reliability of the
self-reports on searching sexual stimuli. Considering
the vast supply of pornographic websites and the
popularity of sex on the Internet,22,23 one would ex-
pect spending time searching for erotic stimuli on
the Internet to be mentioned more often than was
the case in the present study. Socially desirable an-
swer tendencies may have caused underreporting
on searching for erotica and may have masked the
effects of searching for erotica on CIU.
In conclusion, the present study demonstrated
that not all applications of the Internet have an ad-
dictive potential. Using the Internet for predomi-
nantly sexual gratification could be empirically
linked to an increase of CIU in a 1-year period. It
may be that the persons who engage compulsively
in searching sexual gratification through the Inter-
net had a latent vulnerability for becoming overly
attached to sexual explicit stimuli, but that this vul-
nerability would never have resulted in compulsive
behavior if the Internet had not brought them in
contact with an abundance of sexually explicit stim-
uli. Further research should therefore address the
question whether biological deviances, personality
disorders, and/or psychosocial problems are a priori
more prevalent among persons developing CIU.
We are very grateful to Stichting Volksbond Rot-
terdam for funding this research.
APPENDIX: COMPULSIVE INTERNET USE
Instruction: Please answer the following questions
about your use of the Internet for private purposes.
Answers can be given on a five-point scale: (0) never,
(1) seldom, (2) sometimes, (3) often, or (4) very often.
1. How often do you find it difficult to stop using
the Internet when you are online?
2. How often do you continue to use the Internet
despite your intention to stop?
3. How often do others (e.g., partner, children,
parents, friends) say you should use the Inter-
4. How often do you prefer to use the Internet in-
stead of spending time with others (e.g., part-
ner, children, parents, friends)?
5. How often are you short of sleep because of
6. How often do you think about the Internet,
even when not online?
7. How often do you look forward to your next
8. How often do you think you should use the In-
ternet less often?
9. How often have you unsuccessfully tried to
spend less time on the Internet?
10. How often do you rush through your (home)
work in order to go on the Internet?
11. How often do you neglect your daily obliga-
tions (work, school, or family life) because you
prefer to go on the Internet?
12. How often do you go on the Internet when you
are feeling down?
13. How often do you use the Internet to escape
from your sorrows or get relief from negative
14. How often do you feel restless, frustrated, or
irritated when you cannot use the Internet?
1. Young, K.S. (1999). Internet addiction: symptoms,
evaluation, and treatment. In: VandeCreek, L., Jackson,
T. (eds.), Innovations in clinical practice: a source book.
Sarasota, FL: Professional Resource Press, pp. 19–31.
2. Young, K.S. (2004). Internet addiction: a new clinical
phenomenon and its consequences. American Behav-
ioral Scientist 48:402–415.
3. Wang, W. (2001). Internet dependency and psychoso-
cial maturity among college students. International
Journal of Human Computer Studies 55:919–938.
4. Yuen, C., & Lavin, M.J. (2004). Internet dependence
in the collegiate population: the role of shyness. Cy-
berPsychology & Behavior 7:379–383.
102 MEERKERK ET AL.
14192c11.pgs 2/2/06 2:41 PM Page 102
5. Quayle, E., & Taylor, M. (2004). Model of problematic
Internet use in people with sexual interest in chil-
dren. CyberPsychology & Behavior 6:93–106.
6. Shapira, N.A., Goldsmith, T.D., Keck, P.E., et al.
(2000). Psychiatric features of individuals with prob-
lematic Internet use. Journal of Affective Disorders
7. Davis, R.A. (2001). A cognitive-behavioral model of
pathological Internet use. Computers in Human Behav-
8. Dinicola, M.D. (2004). Pathological Internet use
among college students: the prevalence of pathologi-
cal Internet use and its correlates [Dissertation]. Dis-
sertation Abstracts International Section A: Humanities
and Social Sciences 65:1675.
9. Goldberg, I.K. (1997). Ivan Goldberg discusses Inter-
net addiction [On-line]. Available: www.psycom.
10. Young, K.S. (1999). Internet addiction: evaluation
and treatment. Student BMJ 7:351–352.
11. Young, K.S., Pistner, M., O’Mara, J., et al. (1999). Cyber
disorders: the mental health concern for the new mil-
lennium. CyberPsychology & Behavior 2:475–479.
12. Caplan, S.E. (2002). Problematic Internet use and
psychosocial well-being: development of a theory-
based cognitive-behavioral measurement instru-
ment. Computers in Human Behavior 18:553–575.
13. Chou, C., & Hsiao, M.C. (2000). Internet addiction,
usage, gratification, and pleasure experience: the Tai-
wan college students’ case. Computers & Education
14. Li, S.M., & Chung, T.M. (2005). Internet function and
Internet addictive behavior. Computers in Human Be-
havior. (in press). Available Online: http://www.sci-
15. Ward, D.L. (2001). The relationship between psy-
chosocial adjustment, identity formation, and prob-
lematic Internet use [Dissertation]. Dissertation
Abstracts International: Section B: The Sciences and En-
16. Young, K.S. (1998). Internet addiction: the emergence
of a new clinical disorder. CyberPsychology & Behavior
17. Eijnden R.v.d., Meerkerk, G.-J., Vermulst, A., et al.
(2005). Online communication, compulsive Internet
use and psychosocial well-being among adolescents:
a longitudinal study (submitted).
18. Widyanto, L., & McMurran, M. (2004). The psycho-
metric properties of the Internet Addiction Test. Cy-
berPsychology & Behavior 7:443–450.
19. Meerkerk, G.J., Eijnden, R. v.d., & Garretsen, H.
(2005). Prevalence and stability of compulsive Inter-
net use among the Dutch population (submitted).
20. Meerkerk, G.J., Eijnden, R. v.d., Vermulst, A., et al.
(2005). “Internet Addiction” revisited. The develop-
ment of an instrument for measuring Compulsive In-
ternet Use (submitted).
21. Davis, R.A., Flett, G.L., & Besser, A. (2000). Valida-
tion of a new scale for measuring problematic Inter-
net use: implications for pre-employment screening.
CyberPsychology & Behavior 5:331–345.
22. Cooper, A., McLoughlin, I.P., & Campbell, K.M.
(2000). Sexuality in cyberspace: update for the 21st
century. CyberPsychology & Behavior 3:521–536.
23. Cooper, A. (1998). Sexuality and the Internet: surfing
into the new millennium. CyberPsychology & Behavior
24. Choi, D., & Kim, J. (2004). Why people continue to
play online games: in search of critical design factors
to increase customer loyalty to online contents. Cy-
berPsychology & Behavior 7:11–24.
25. Johansson, A., & Gotestam, K. (2004). Problems with
computer games without monetary reward: similar-
ity to pathological gambling. Psychological Reports
26. Griffiths, M. (2000). Excessive Internet use: implica-
tions for sexual behavior. CyberPsychology & Behavior
27. Suler, J. (2004). The online disinhibition effect. Cy-
berPsychology & Behavior 7:321–326.
28. Chen, H., Wigand, R.T., & Nilan, M. (2000). Explor-
ing Web users’ optimal flow experiences. Information
Technology and People 13:263–281.
29. Putnam, D.E. (2000). Initiation and maintenance of on-
line sexual compulsivity: implications for assessment
and treatment. CyberPsychology & Behavior 3:553–563.
30. Engelberg, E., & Sjoberg, L. (2004). Internet use, so-
cial skills, and adjustment. CyberPsychology & Behav-
31. Moody, E.J. (2001). Internet use and its relationship
to loneliness. CyberPsychology & Behavior 4:393–401.
32. Yang, S.C., & Tung, C.J. (2005). Comparison of Internet
addicts and non-addicts in Taiwanese high school.
Computers in Human Behavior (in press). Available On-
33. Armstrong, L., Phillips, J.G., Saling, L.L. (2000). Poten-
tial determinants of heavier Internet usage. Interna-
tional Journal of Human-Computer Studies 53:537–550.
34. Shaffer, H.J., Hall, M.N., & VanderBilt, J. (2000).
“Computer addiction”: a critical consideration.
American Journal of Orthopsychiatry 70:162–168.
35. Shaffer, H.J. (2002). Is computer addiction a unique
psychiatric disorder? Psychiatric Times 19. Available
36. Griffiths, M. (1999). Internet addiction—Internet
fuels other addictions. Student BMJ 7:428.
Address reprint requests to:
Dr. Gert-Jan Meerkerk
Addiction Research Institute Rotterdam
3021 DM Rotterdam, The Netherlands
PREDICTING COMPULSIVE INTERNET USE 103
14192c11.pgs 2/2/06 2:41 PM Page 103