Original Research

SCL-90-R and 16PF Profiles of Senior High School

Students With Excessive Internet Use
Chang-Kook Yang, MD, PhD1, Byeong-Moo Choe, MD, PhD1, Matthew Baity, PhD2,
Jeong-Hyeong Lee, PhD3, Jin-Seok Cho, MA4

Objective: To investigate the psychiatric symptomatology and personality characteristics of Korean senior
high school students considered to use the Internet to excess.
Method: We administered a questionnaire packet to students that included 4 measures. These measures
included a questionnaire on Internet use patterns during the previous month, the Internet Addiction Test
(IAT), the Symptom Checklist-90-R (SCL-90-R), and the Sixteen Personality Factor Questionnaire (16PF).
A total of 328 students, aged 15 to 19 years, participated in the study.
Results: Students were divided into 4 Internet user groups according to their IAT total scores: nonusers
(n = 59, 18.0%), minimal users (n = 155, 47.3%), moderate users (n = 98, 29.9%), and excessive users (n =
16, 4.9%). The SCL-90-R showed that the excessive users group, when compared with the other groups in
this study, reported the highest levels of symptomatology. The 16PF also revealed that excessive users
were easily affected by feeling, emotionally less stable, imaginative, absorbed in thought, self-sufficient,
experimenting, and preferred their own decisions.
Conclusions: This study suggests that senior high school students who use the Internet to excess report and
subsequently exhibit significantly more psychiatric symptoms than students who use the Internet less
frequently. In addition, excessive users appear to have a distinctive personality profile when compared with
nonusers, minimal, and moderate users.
(Can J Psychiatry 2005;50:407–414)
Information on author affiliations appears at the end of the article.

Clinical Implications
· Senior high school students whose self-described Internet use was considered to be excessive reported
significantly higher scores on most of the SCL-90-R scales and appeared to have personality traits that were
distinctive from the other groups.
· Four personality factors of the 16PF (low scores on factor C, high scores on factor M, high scores on factor Q2,
and high scores on factor CRE) were identified only among excessive Internet users.
· The above findings provide some basic information for understanding psychopathological and personality
profiles of senior high school students who use the Internet to excess.

Limitations
· This was a retrospective study, and we used a 2-stage sampling method to select subjects. In addition, we used a
questionnaire method rather than a direct in-depth interview.
· The sample size of excessive Internet users was small, which may limit the power of our study. It is difficult to
determine whether the behaviours and attitudes expressed by the students led to or were a product of their
Internet-dependent behaviour.

Key Words: high school students, excessive Internet use, psychiatric symptomatology, personality traits
he recent surge of affordable computers on the market has levels. Specifically, the Internet has become indispensable to
T established that personal computers are both easily
obtained and an essential learning tool at many educational
teachers and students, as it provides a plethora of information
as well as various programs for entertainment and easy inter-

Can J Psychiatry, Vol 50, No 7, June 2005 W 407

The Canadian Journal of Psychiatry—Original Research
Abbreviations used in this article
16PF Sixteen Personality Factor Questionnaire
ANOVA analysis of variance
GSI Global Severity Index
IAT Internet Addiction Test
PSDI Positive Symptom Distress Index
PST Positive Symptom Total
SCL-90-RSymptom Checklist-90-R
SD standard deviation
personal communication. Unfortunately, for many of the
same reasons that the Internet has expanded and enriched the
lives of school-aged children, some unwanted effects have
begun to appear in the form of overreliance or extended
Internet use.
Although there is no official diagnostic classification for a
pathological preoccupation with the Internet, some profes-
sionals have proposed that this phenomenon be considered an
Internet addiction (1–3). The condition is characterized as an
impulse-control disorder that primarily involves psychologi-
cal dependence on the Internet (3). Internet addiction is con-
sidered a behavioural addiction akin to pathological
gambling, but further in-depth studies are needed to clarify
whether it is a new diagnostic entity. In addition, there
remains some hesitation about the use of the term “addiction,”
despite the apparent descriptive fit between what constitutes
an addictive disorder and what has been observed
behaviourally in individuals with extensive and pathological
Internet use. For this study, we use the term “excessive
Internet use” to describe an overindulgence or abuse of the
Internet to the extent that such use results in psychological
dependence and a significant decline in sociopsychiatric
functioning (4).
The literature on excessive Internet use has begun to report on
the various clinical problems associated with this behaviour.
One study has suggested that the Internet is used to counteract
other deficiencies in the person’s life, such as poor social rela-
tionships, lack of friends, dissatisfaction with physical
appearance, disability, and poor coping (1). The range of clin-
ical and environmental problems that have been associated
with excessive Internet use include functional impairment and
an increased likelihood of Axis I psychiatric disorders (5),
increased social isolation (6), and a decreased sense of
well-being (7). In regard to premorbid conditions, another
study has suggested that preoccupation with the Internet may
be related to various clinical problems that predate Internet
use, including depression, anxiety, and low self-esteem (3).
There also exists a small body of literature examining the per-
sonality traits of addiction-prone individuals, which include
408
increased shyness, introversion, and social withdrawal
(8,9–11). The results of the literature thus far suggest not only
that excessive Internet use can have potential negative effects
on a person’s functioning but also that certain symptom clus-
ters may predispose some individuals to preoccupation with
the Internet.
The frequency and intensity of Internet use seems to vary
widely among adolescents. Although this age group tends to
be quite impressionable and may use the Internet extensively,
not all overindulge or abuse it. Even fewer adolescents seem
to experience detrimental side effects as a result of their
Internet use; however, cases have begun to appear in aca-
demic and mental health settings. The research to date indi-
cates that significant impairments in personality and
psychiatric symptomatology are associated with excessive
Internet use. Accordingly, our objective in this study was to
further contribute to the literature by examining the psychiat-
ric symptoms and personality profiles of excessive Internet
users in an adolescent population. More specifically, this
study attempts to describe the clinical profile of adolescents to
help identify both the consequences of excessive Internet use
and the personality traits that may make some adolescents
more vulnerable to Internet abuse.
Materials and Methods
Subjects
Busan City, Korea, has 15 districts that can be divided into 3
groups: upper, middle, and lower economic status. One dis-
trict was selected randomly from the upper, 2 from the middle,
and 1 from the lower group. We than used a 2-stage sampling
method to select 4 senior high schools from each of the 4 dis-
tricts. From the selected schools, 337 first- to third-year stu-
dents were approached with the opportunity to participate in
the current research study. Of the original pool of participants,
we used data from 328 volunteers (9 students did not complete
the questionnaire). The mean age was 16.2 years, SD 0.95
(range 15 to 19 years). The subjects were evenly distributed in
their grades and sex (Table 1).
Methods
The questionnaire package comprised 4 measures: a brief per-
sonal profile including Internet use patterns during the previ-
ous month, the IAT, the SCL-90-R, and the 16PF. The
questionnaires were administered in the middle of May 2002
because Korean students are not expected to take any kind of
regular school examination for approximately 1 month before
and after this point. We hypothesized that the Internet use pat-
terns of students at this point might represent students’ usual
patterns of Internet use. We obtained written informed con-
sent from the participating students and from homeroom
teachers and principals, since they represent the students’
W Can J Psychiatry, Vol 50, No 7, June 2005
 SCL-90-R and 16PF Profiles of Senior High School Students With Excessive Internet Use

Table 1 Grade and sex distribution of subjects

Grade Age Male students Female students Total
years n (%) n (%) n (%)

10 15 to 17 56 (17.1) 54 (16.5) 110 (33.6)

11 16 to 18 54 (16.5) 55 (16.8) 109 (33.2)
12 17 to 19 54 (16.5) 55 (16.8) 109 (33.2)
Total 164 (50.0) 164 (50.0) 328 (100.0)

parents at school in Korea. The ethical committee of our uni-
versity medical centre approved the study protocol, and all
data were collected simultaneously.
Internet Addiction Test
We used the IAT (3) to measure levels of Internet use during
the previous month. The IAT has 20 items associated with
Internet use, including psychological dependence, compul-
sive use, and withdrawal, as well as related problems of
school, sleep, family, and time management. For each item, a
graded response is selected (1 = “not at all” to 5 = “always”),
and total scores can range between 20 and 100. The internal
consistency (Cronbach’s alpha) of the IAT was 0.92, and its
test–retest reliability, performed biweekly, was also satisfac-
tory (r = 0.85, n = 60, P < 0.001). In a previous study (4),
cut-off scores for the IAT were developed to divide Internet
users into minimal (scores 20 to 39), moderate (scores 40 to
59), and excessive (scores 60 to 100) users on the basis of the
severity of their Internet addictive behaviour (that is, total IAT
scores) and the scores of a questionnaire probing
Internet-induced behavioural alterations that are most likely
associated with problems in the 5 areas of daily life. These
same cut-off scores were used in the current study.
Symptom Checklist-90-Revision
The SCL-90-R is one of the most widely used and well-
validated self-report symptom inventories designed to reflect
the psychological symptoms seen in psychiatric and medical
patients (12). Respondents rate 90 items using a 5-point scale
(1 = “no problem” to 5 = “very serious”) to measure the extent
to which they have experienced the listed symptoms in the last
7 days. The items are divided into 9 subscales: Somatization,
Obsessive–Compulsive, Interpersonal Sensitivity, Depres-
sion, Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation,
and Psychoticism. Higher scores on the SCL-90-R indicate
greater psychological distress. The SCL-90-R also has 3
global indexes: the Global Severity Index (GSI) measures the
extent or depth of the individual’s psychiatric disturbance; the
Positive Symptom Total (PST) counts the total number of
questions rated above 1 point; and the Positive Symptom
Distress Index (PSDI) represents the intensity of symptoms.
Sixteen Personality Factor Questionnaire
The 16PF, developed from the theories of Cattell and others,
measures an individual’s underlying “normal” personality
traits (13). Cattell and others’ measure is based on 16 primary
factors or source traits that were believed to identify an indi-
vidual’s total personality. Items for each factor are scored on a
bipolar scale, and the primary factors are then combined to
form global secondary factors.
Data Analyses
We performed all statistical analyses using SAS for Windows
(14). In addition, we performed the chi-square test to find out
whether statistical differences existed among the groups. We
used 1-way ANOVA to find out whether differences existed
among the groups on the SCL-90-R and 16PF. We made mul-
tiple comparisons and used the Bonferroni test when there
was a significant difference among the groups. Linear and
quadratic trend analysis were also examined to test the trends
among the groups. A score of 70 is 2 SDs above the mean of
each SCL-90-R scale and is considered clinically significant.
We calculated the percentage of students with a score above
70 on the SCL-90-R and tested differences among the groups
for significance, using the Cochran–Mantel–Haenszel test.
Results
Status of Internet Use
Of the 328 students, 269 (82.0%) reported use of the Internet
during the previous month, and 59 students (18.0%) denied
any use. Of the students who reported Internet use, 146 were
male (89.0%) and 123 were female (75.0%), a significant dif-
ference (P2 = 17.2, P = 0.001). There were no differences in
Internet use among the 3 grades.
Length of Internet Use
Average daily length of Internet use during weekdays was 0.8
hours, SD 1.1, for minimal users; 1.5 hours, SD 1.5, for mod-
erate users; and 2.3 hours, SD 1.7, for excessive users. Results
indicated significant differences among the 3 user groups (F =

Can J Psychiatry, Vol 50, No 7, June 2005 W 409

The Canadian Journal of Psychiatry—Original Research

Table 2 SCL-90-R clinical scale scores

Scales Nonusers Minimal users Moderate users Excessive users F (P)
n = 59 n = 155 n = 98 n = 16
Mean (SD) Mean (SD) Mean (SD) Mean (SD)
Somatization 50.3 (9.9)a,b 48.3 (9.1)a 51.0 (9.6)a,b 54.7 (16.9)b 2.84 (0.038)
a,b a a,b b
Obsessive–Compulsive 49.9 (11.3) 46.7 (10.1) 49.8 (10.3) 54.7 (14.1) 3.89 (0.009)
a,b a a,b b
Interpersonal Sensitivity 49.5 (11.4) 45.7 (10.0) 48.4 (10.4) 53.3 (14.1) 3.76 (0.011)
a a a b
Depression 48.1 (10.6) 46.1 (9.9) 48.4 (11.1) 54.8 (14.2) 3.57 (0.014)
a,b a a,b b
Anxiety 49.4 (9.8) 45.6 (8.6) 47.8 (8.6) 53.0 (13.8) 4.72 (0.003)
Hostility 48.5 (8.6)a 47.5 (8.9)a 49.6 (9.6)a,b 54.6 (11.3)b 3.03 (0.030)
a,b a a,b b
Phobic Anxiety 46.9 (7.8) 45.1 (5.5) 46.7 (7.3) 50.4 (10.9) 3.52 (0.015)
Paranoid Ideation 51.0 (12.3) 48.0 (10.8) 51.8 (12.0) 54.5 (16.2) 2.96 (0.032)
a,b a a,b b
Psychoticism 50.4 (10.5) 47.6 (8.9) 51.2 (11.0) 54.4 (13.3) 3.93 (0.009)
a,b a a,b b
GSI 49.1 (10.7) 46.0 (9.3) 49.2 (10.2) 54.6 (16.4) 4.58 (0.004)
PSDI 49.5 (9.7)a 48.0 (10.3)a 52.8 (12.3)a,b 57.6 (15.3)b 5.81 ( 0.001)
PST 48.2 (10.5) 44.8 (11.2) 47.0 (10.6) 50.6 (12.7) 2.35 (0.072)
The F tests determined whether there were significant differences among all the groups (df for all tests = 3, 324). Post hoc tests were then used to distinguish
among the groups. Groups with different superscripts are significantly different.

14.5, P = 0.001). The average daily length of Internet use dur-
ing weekends was 1.6 hours, SD 1.3, for minimal users, 2.4
hours, SD 1.6, for moderate users, and 3.5 hours, SD 3.6, for
excessive users, which also revealed significant differences
among the 3 groups (F = 13.7, P < 0.001). The multiple com-
parisons revealed a significant difference in the daily length of
Internet use during weekdays and weekends among the 3
groups, with significantly longer use by excessive users, fol-
lowed by moderate and minimal users.
Ongoing Main Internet Activities
According to the students’ self-report, the Internet was used
mainly for virtual games (72.1%), chat room activity (12.3%),
school-related activities (7.1%), file downloads (5.2%), news
groups (1.9%), and other (1.4%). There was no significant dif-
ference among the 3 user groups in the principal Internet
activities. The most frequent Internet activities reported by
excessive users were virtual games (87.5%) and the use of
chat rooms (12.5%). Principal activities differed between the
sexes, with male students preferring games and female stu-
dents preferring chatting (P2 = 171.9, P < 0.001).
Internet Addiction Test
According to the IAT cut-off score, 59 students (18.0%) were
nonusers, 155 students (47.3%) were minimal users, 98 stu-
dents (29.9%) were moderate users, and 16 students (4.9%)
were excessive users. The 59 nonusers comprised 18 male stu-
dents (30.5%) and 41 (69.5%) female students; the 155 mini-
mal users comprised 74 (47.7%) male students and 81
(52.3%) female students; the 98 moderate users comprised 62
(63.3%) male students and 36 (36.7%) female students; and
the 16 excessive users comprised 10 (62.5%) male students
and 6 (37.5%) female students. A comparison of sex represen-
tation showed that male students significantly outnumbered
female students in both the moderate and excessive user
groups (P2 = 17.2, P = 0.001).
Comparison of Average SCL-90-R Scores by Groups
Table 2 represents the means, SDs, and ANOVAs with post
hoc comparisons for each group. Some significant
between-group differences were found on 8 subscales and 2
global indexes (the GSI and PSDI). Overall, excessive users
had significantly higher scores than minimal users on all
subscales except Paranoid Ideation, as well as on 2 of the
global indexes (GSI and PSDI), which suggests higher levels
of pathology among students who report excessive Internet
use. We found no statistically significant differences among
the subscales and global indexes for the nonuser, minimal,
and moderate user groups. However, when we examined the
trend analysis, the subscales Obsessive–Compulsive (F =
5.72, P = 0.017), Interpersonal Sensitivity (F = 6.54, P =
0.011), and Anxiety (F = 9.46, P = 0.002), as well as the PST
(F = 3.93, P = 0.048), showed a U function. Conversely, the
subscales Depression (F = 5.89, P = 0.016), Hostility (F =
6.02, P = 0.015), and Phobic Anxiety (F = 4.07, P = 0.045), as
well as the GSI (F = 0.90, P = 0.028) showed an offset U func-
tion. The PSDI (F = 8.79, P = 0.003) showed a linear increase.

410
W Can J Psychiatry, Vol 50, No 7, June 2005
 SCL-90-R and 16PF Profiles of Senior High School Students With Excessive Internet Use

Table 3 Percentage of subjects with high scores (T > 70) on the SCL-90-R scales
Scales Non-users Minimal users Moderate users Excessive users c2 (P)
n = 59 n = 155 n = 98 n = 16
Somatization 6.8 1.9 5.1 12.5 5.78 (0.123)
a b a c
Obsessive–Compulsive 3.4 1.9 6.1 25.0 18.4 (< 0.001)
Interpersonal Sensitivity 5.1 2.6 6.1 12.5 4.28 (0.233)
Depression 3.4 3.9 6.1 12.5 2.94 (0.401)
Anxiety 3.4 3.2 3.1 12.5 3.73 (0.292)
Hostility 1.7a 3.9a 7.1a,b 18.8b 8.74 (0.033)
Phobic Anxiety 5.1 0.7 2.0 6.3 5.39 (0.145)
Paranoid ideation 10.2 4.5 10.2 18.8 6.12 (0.106)
Psychoticism 8.5 2.6 9.2 12.5 6.69 (0.083)
a,b a a,b b
GSI 5.1 1.3 5.1 18.8 12.59 (0.006)
PSDI 3.4 4.5 7.1 6.3 1.33 (0.723)
a,b a a b
PST 1.7 0.0 0.0 6.3 11.06 (0.011)

The c tests determined whether there were significant differences among all the groups. Post hoc tests were then used to distinguish among the
2

groups. Groups with different superscripts are significantly different.

Table 3 shows the results of analyses that used the
Cochran–Mantel–Haenszel test to compare the 4 groups with
scores over 70 on the SCL-90-R subscales. Means are
reported in terms of the percentage of students in each group
with scores 2 SDs above the mean. We found significant dif-
ferences among the groups for 2 subscales (Obsessive–
Compulsive and Hostility) and 2 indexes (GSI and PST). In all
cases where differences were found, the excessive users
group had a significantly higher percentage of students with
scores above 70 than did the other 3 groups. Among excessive
users, the percentage of students scoring above 70 was shown
to be highest for Obsessive–Compulsive symptoms (25.0%);
followed by Hostility, Paranoid Ideation, and GSI (all
18.8%); and Somatization, Interpersonal Sensitivity, Depres-
sion, Anxiety, and Psychoticism (all 12.5%). An interesting
pattern among the nonusers emerged, wherein students in this
group reported greater levels of pathology on 7 of 9 SCL-90-R
subscales and had significantly higher GSI and PST scores
than did the minimal users group. This finding suggests a cer-
tain degree of symptomatology in students who do not have at
least some exposure to the Internet.
Comparison of Average 16PF Scores by Groups
Table 4 represents the mean, SD, and ANOVA, with post hoc
comparison, for each of the 4 study groups. The mean score
for every factor is 5.5, with scores of 8 and above and 3 and
below considered significant. Because the 16PF measures
normal personality, the individual factor scores are used to
indicate the direction of traits. The mean scores of all the
groups fell within the middle range (between 4.2 and 7.5). The
16PF profiles of excessive users differed significantly from
those of the nonusers and minimal users on factors C, M, and
Q2 among the primary factors and factor CRE among the sec-
ondary factors. When we examined the trend analysis, all fac-
tors that showed significant differences in mean values among
the groups—M (F = 7.03, P = 0.008), Q2 (F = 18.72, P <
0.0001), and CRE (F = 7.92, P = 0.005)—showed a linear
increase, whereas factor C (F = 8.93, P = 0.003) showed a lin-
ear decrease. In addition to these findings, factors that did not
show significant differences in mean values among the
groups—E (F = 3.96, P = 0.048), Q1 (F = 4.43, P = 0.036),
and IND (F = 4.98, P = 0.026)—did show a linear increase.
Factor Q4 showed a linear increase tendency, even though it
failed to reach significance (F = 3.75, P = 0.054). Table 4
shows that excessive users revealed that they were easily
affected by feelings, emotionally less stable, imaginative,
absorbed in thought, self-sufficient, experimenting, and pre-
f e r r e d th e ir o w n d e c is io n s , co mp a r e d w ith th e
remaining groups.
Discussion
This study primarily focused on exploring the psychiatric
symptomatology and personality characteristics of Korean
senior high school students who report excessive Internet use.
Our results indicate that some students use the Internet exces-
sively and that their main ongoing Internet activities are vir-
tual gaming and chatting. Moreover, these excessive Internet
users reveal significantly more psychiatric symptoms and

Can J Psychiatry, Vol 50, No 7, June 2005 W 411

The Canadian Journal of Psychiatry—Original Research

Table 4 Means and SD for 16PF

Factors Nonusers Minimal users Moderate users Excessive users F (P)
n = 59 n = 155 n = 98 n = 16
Mean (SD) Mean (SD) Mean (SD) Mean (SD)

Primary factors
A: reserved vs outgoing 5.9 (2.2) 6.2 (2.3) 5.9 (2.1) 5.8 (3.3) 0.35 (0.793)
B: concrete thinking vs abstract thinking 6.9 (1.8) 6.7 (1.8) 6.2 (1.9) 6.3 (2.1) 2.11 (0.099)
a a a,b b
C: affected by feelings vs emotionally stable 6.6 (2.2) 6.8 (2.2) 6.1 (2.3) 4.9 (2.7) 4.35 (0.005)
E: submissive vs dominant 5.5 (2.3) 5.8 (2.3) 6.0 (2.5) 6.8 (2.4) 1.39 (0.246)
F: sober vs enthusiastic 6.1 (1.9) 6.3 (2.3) 6.4 (2.1) 6.6 (2.5) 0.30 (0.822)
G: expedient vs conscientious 5.2 (2.3) 5.0 (2.3) 5.1 (2.4) 4.6 (2.5) 0.30 (0.827)
H: shy vs bold 5.5 (2.4) 6.0 2.3) 6.2 (2.4) 6.6 (2.8) 1.31 (0.271)
I: tough-minded vs tender-minded 4.4 (2.0) 4.2 (2.1) 4.4 (1.9) 4.3 (2.4) 0.21 (0.887)
L: trusting vs suspicious 5.7 (2.0) 5.6 (2.1) 5.6 (2.2) 5.6 (2.1) 0.01 (0.998)
M: practical vs imaginative 4.6 (2.5)a 4.4 (2.1)a 5.1 (2.2)a,b 6.1 (2.6)b 4.15 (0.007)
N: forthright vs shrewd 6.6 (2.3) 6.3 (2.2) 6.5 (2.1) 5.7 (1.9) 1.00 (0.394)
O: self-assured vs apprehensive 5.2 (2.1) 4.8 (2.2) 5.0 (2.1) 5.6 (2.2) 0.89 (0.448)
Q1: conservative vs experimenting 6.3 (2.0) 6.4 (2.0) 6.6 (1.9) 7.5 (1.8) 1.66 (0.175)
a a a b
Q2: group-oriented vs self-sufficient 4.2 (2.3) 5.0 (2.3) 5.3 (2.3) 7.0 (1.6) 6.70 ( 0.001)
Q3: undisciplined self-conflict vs following 5.3 (2.0) 5.2 (2.2) 5.3 (2.2) 4.9 (2.6) 0.14 (0.935)
self-image
Q4: relaxed vs tense 5.0 (2.1) 4.8 (2.2) 5.2 (2.1) 6.1 (2.5) 1.70 (0.167)
Secondary factors
EXT: introversion vs extroversion 56.1 (25.8) 55.1 (26.8) 52.7 (24.8) 48.5 (29.4) 0.51 (0.679)
ANX: low anxiety vs high anxiety 38.3 (24.4) 35.0 (24.1) 39.6 (22.9) 47.3 (28.5) 1.59 (0.191)
TOU: emotional sensitivity vs tough poise 57.5 (24.3) 59.4 (20.4) 54.7 (20.4) 48.8 (22.8) 1.52 (0.209)
IND: subduedness vs independence 47.2 (26.0) 51.5 (27.0) 54.5 (27.2) 63.8 (26.8) 1.84 (0.140)
SUP: low control vs high control 40.6 (26.9) 38.6 (28.8) 40.0 (30.0) 30.9 (29.2) 0.47 (0.701)
a a a,b b
CRE: low creativity vs high creativity 42.7 (8.0) 43.1 19.0) 46.3 (6.5) 56.4 (8.1) 3.02 (0.030)

The F tests determined whether there were significant differences among all the groups (df for all tests = 3, 324). Post hoc tests were then used to distinguish
among the groups. Groups with different superscripts are significantly different.

have some highly characteristic personality profiles, com-
pared with Korean high school students who report less
frequent Internet use.
The SCL-90-R is a subjective measure of the degree of psy-
chiatric symptoms. With the SCL-90-R, it is possible to differ-
entiate the students into 2 groups in terms of the frequency of
subjective psychiatric complaints (that is, neurotic and psy-
chotic symptoms) and severity of use (that is, excessive vs
moderate, minimal, and no use). In this study, the average
score on the symptom dimensions reported by students who
used the Internet excessively was not high enough to indicate
a clinically significant abnormality. However, the results
themselves are important because the excessive users
revealed more psychiatric symptoms than did the remaining
groups. Compared with other groups, a significantly higher
percentage of students reporting excessive use had scores
above 70 on the Obsessive–Compulsive and Hostility
subscales and the GSI, as well as on the PST. These results
indicate that students in this study who report excessive
Internet use are characterized by complaints of indecisive-
ness, preoccupation with details, nervousness, irritability,
aggressiveness, and impulsivity. The high percentage of
excessive users with significant scores on the GSI suggests
that the extent or depth of the present psychiatric disturbance
is severe in this group. It was interesting to find that 3 of the 9
clinical scales and 1 scale of 3 global indexes in the SCL-90-R
showed a U function, and 3 of the 9 clinical scales and 1 scale
of 3 global indexes in the SCL-90-R showed an offset U

412
W Can J Psychiatry, Vol 50, No 7, June 2005
 SCL-90-R and 16PF Profiles of Senior High School Students With Excessive Internet Use
function. These findings show that 1) the nonusers had higher
scores than the minimal users, 2) the moderate users reported
greater symptomatology than the minimal users, and 3) the
excessive users showed the highest level of psychiatric symp-
toms. Internet use is common and widespread among adoles-
cents. Therefore, not using the Internet may represent a trend
deviating from the usual pattern of teenagers’ behaviour, and
it might be expected that there will be some difference in psy-
chiatric symptoms between students reporting no Internet use
and students reporting at least some Internet use. These trend
analysis results suggest a need for further studies to examine
psychiatric symptoms of students who do not use the Internet.
The Internet addictive process is not fully understood. How-
ever, there is a suggestion that personality traits may predis-
pose certain individuals to overuse the Internet (15). In our
study, the 4 personality factors most associated with excessive
Internet use were factors C, M, Q2, and CRE. Individuals with
low scores on factor C tend to have low tolerance for unsatis-
factory conditions, to be changeable, to evade the necessary
demands of reality, to be irritable, and to show some classic
neurotic symptoms (for example, phobias, sleep disturbance,
and psychosomatic complaints) (13). The Internet may be
used as an escape from reality and, further, may be a tonic for
people with inner conflicts. Students with such traits may use
the Internet to counteract psychological distress rooted in
their personality.
Individuals with high scores on factor M can be described as
imaginative, absent-minded, absorbed in thought, and
impractical. They tend to be unconventional, unconcerned
with everyday matters, self-motivated, imaginatively cre-
ative, concerned with essentials, often absorbed in thought,
and oblivious of particular people and physical realities. The
self-guided interests of high-M individuals may lead to unre-
alistic situations accompanied by expressive outbursts. Their
individuality can lead to their rejection from group activities
(13). This cluster of traits reflects a personality profile that
may predispose an individual to derive a great deal of satisfac-
tion from Internet activities rather than from more conven-
tional social activities or friendships. The main ongoing
Internet activity in the present study was game playing. Stu-
dents who play virtual games excessively may have a more
schizoid personality than other students; because of their
inward-looking nature, the virtual games themselves are
essentially solitary activities.
High scores on factor Q2 suggest an individual who is
self-sufficient, resourceful, and independent. Potential nega-
tive consequence of these traits are that these individuals can
become accustomed to taking action on their own and can
have some difficulties compromising with others. Although
people with high factor Q2 scores do not dislike others, they
simply do not feel the need to obtain their agreement or
Can J Psychiatry, Vol 50, No 7, June 2005 W
support (13). This trait is consistent with the results from pre-
vious studies that have identified both pride and frequent
intellectualization as personality characteristics commonly
associated with problematic Internet use (8,16). On the
Internet, people have the freedom to explore cyberspace on
their own, irrespective of social norms. Thus cyberspace may
become an attractive place for a student who has the traits of
egocentricity, familiarity with having his or her own way, and
intellectual curiosity.
High CRE scorers are imaginative and willing to explore new
ideas. The characteristics of high CRE scorers are closely
related to those associated with the high M and Q2 factors
described earlier. According to Cattell and others (13), high
CRE indicates people who are usually self-sufficient, who are
often serious in their demeanour, and who generally prefer
solitary activities. Sometimes high CRE scorers are overly
imaginative, and their expectations or beliefs can become fan-
tastical. These individuals may fail to see the practical limita-
tions to their ideas. This cluster follows a personality pattern
similar to that of the shy, scientific, less empathic, and intro-
verted intellectual who is drawn to the Internet (8,9).
A noticeable finding was that factors E, Q1, Q4, and IND
showed a linear increase trend even though there were no sig-
nificant differences in mean values among the groups.
According to Cattell and others (13), individuals who score
high on factor E tend to be dominant, assertive, self-assured,
and independent-minded. Individuals who score high on fac-
tor Q1 tend to be experimenting, liberal, critical, open to
change, and seeking intellectual stimulation. High scores on
factor Q4 are an indication that one tends to be tense, fretful,
impatient, and intensely motivated, while high scores on the
IND factor suggest increased levels of aggressiveness.
Although the small sample size likely contributed to the find-
ings’ lack of statistical significance, the trend analysis helps to
further describe characteristics of those who report excessive
Internet use. It seems that all these personality traits appear
contextually appropriate in the kind of environment provided
by the Internet. Our study results suggest that personality pro-
files should be considered in accounts of the etiology of
excessive Internet use among high school students. Future
studies should continue to examine how personality traits
influence problematic Internet use and should investigate
whether a similar personality profile may be an etiologic fac-
tor in the development of any addictive syndrome, whether it
be to alcohol, gambling, or the Internet.
This study has several limitations. First, we used a question-
naire method rather than a direct, in-depth interview. Second,
the sample size of excessive Internet users is too small to rep-
resent the overall characteristics of the population of exces-
sive users. Third, our results do not clearly indicate whether
the symptoms identified in this study preceded the
413
The Canadian Journal of Psychiatry—Original Research

development of Internet-dependent behaviour or were a con-
sequence of Internet use. There is also a need for more studies
of adolescents who visit psychiatric clinics and seek profes-
sional help for problematic Internet use. Despite this study’s
limitations, the picture has begun to emerge that adolescents
who overuse the Internet not only have an increased likeli-
hood of psychiatric symptoms but also tend to be socially iso-
lated from their peers. This isolation may not be strictly a
result of spending more time on the Internet; it could in fact be
an expression of preexisting personality traits. While defini-
tive conclusions should be reserved until future research can
produce similar results, an interesting profile has emerged
from the results of this study, suggesting that certain personal-
ity traits may predispose some adolescents to become
“addicted” to the Internet.
Conclusion
The present study suggests that Korean senior high school stu-
dents who use the Internet to excess report more psychiatric
symptoms and differ in personality profiles from nonusers,
minimal, and moderate users. The 4 personality factors identi-
fied appear to explain the profile of an individual who is
drawn to the environment of the Internet. The study also pro-
vides some characteristic psychiatric symptomatology and
personality profiles that may more broadly describe other
students who use the Internet to excess.
References
1. Eppright T, Allwood M, Stern B, Theiss T. Internet addiction: a new type of
addiction? Mo Med 1999;96:133– 6.
2. Tsai CC, Lin SS. Internet addiction of adolescents in Taiwan: an interview study.
Cyberpsychol Behav 2003;6:649–52.
3. Young KS. Caught in the net. New York: John Wiley & Sons; 1998. p 12–120.
4. Yang CK. Sociopsychiatric characteristics of adolescents who use computers to
excess. Acta Psychiatr Scand 2001;103:1– 6.
5. Shapira NA, Goldsmith TD, Keck Jr. PE, Khosla UM, McElroy SL. Psychiatric
features of individuals with problematic Internet use. J Affect Disord
2000;57:267–72.
6. Kraut R, Patterson M, Lundmark V, Kiesler S, Mukopadhyay T, Scherlis W.
Internet paradox: a social technology that reduces social involvement and
psychological well-being? Am Psychol 1998;53:1017–31.
7. Moody EJ. Internet use and its relationship to loneliness. CyberPsychol Behav
2001;4:30:393– 401.
8. Shotton MA. The costs and benefits of ‘the Internet addiction.’ Behav Inform
Technol 1991;10:219–30.
9. Douse NA, McManus IC. The personality of fantasy game players. Br J Psychol
1993;84:505–9.
10. Pratarelli ME, Browne BL, Johnson K. The bits and bytes of computer/Internet
addiction: a factor analytic approach. Behav Res Methods Instrum Comput
1999;31:305–14.
11. Young KS. Psychology of computer use: XL. Addictive use of the Internet: a
case that breaks the stereotype. Psychol Rep 1996;79:899–902.
12. Kim KI, Kim JH, Won HT. Korean manual of Symptom Checkilist-90-Revision.
Seoul: Jung Ang Juk Sung Publisher; 1984. p 8–10.
13. IPAT Staff. Administrator’s manual for the 16 PF. Champaign (IL): Institute for
Personality & Ability Testing; 1986.
14. SAS Institute Inc. SAS/STAT for Windows. SAS Proprietary Software Release
8.2. Cary (NC): SAS Institute; 2000.
15. Griffiths MD, Dancaster I. The effect of type A personality on physiological
arousal while playing computer games. Addict Behav 1995;20:543–8.
16. Pocius KE. Personality factors in human–computers interaction: a review of the
literature. The Internets in Human Behavior 1991;7:103–35.
Manuscript received October 2004, revised, and accepted January 2005.
1
Professor, Department of Psychiatry, Dong-A University College of
Medicine, Busan, Korea.
2
Staff Psychologist, Department of Psychiatry, Massachusetts General
Hospital, Harvard Medical School, Boston, Massachusetts.
3
Associate Professor, Division of Management Information Science,
Dong-A University, Busan, Korea.
4
Clinical Psychologist, Department of Psychology, Yang-San
Neuropsychiatric Hospital, Yang-San, Korea.
Address for correspondence: Dr CK Yang, Department of Psychiatry,
Donga-A University College of Medicine, 3 Ga-1, Dongdaesin-dong,
Seo-gu, Busan, 602-715 Korea
e-mail; ckyang@daunet.donga.ac.kr

Résumé : Résultats à la SCL-90-R et au 16 PF d’élèves d’école secondaire de 2e cycle utilisant Internet à

l’excès
Objectif : Rechercher la symptomatologie psychiatrique et les caractéristiques de personnalité d’élèves coréens d’école
secondaire de 2e cycle estimés être des utilisateurs d’Internet excessifs.
Méthode : Nous avons administré aux élèves une série de questionnaires comprenant 4 mesures qui incluaient des
questions sur le mode d’utilisation d’Internet au cours du mois précédent : le test de dépendance à Internet (IAT), la liste de
vérification des symptômes-90-R (SCL-90-R), et le questionnaire des seize facteurs de la personnalité (16PF). En tout, 328
élèves de 15 à 19 ans ont participé à l’étude.
Résultats : Les élèves ont été répartis en 4 groupes d’utilisateurs d’Internet, selon leur score total au IAT : non-utilisateurs
(n = 59, 18,0 %), utilisateurs minimaux (n = 155, 47,3 %), utilisateurs modérés (n = 98, 29,9 %), et utilisateurs excessifs
(n = 16, 4,9 %). La SCL-90-R indiquait que le groupe des utilisateurs excessifs, comparé aux autres groupes de l’étude,
déclarait les niveaux de symptomatologie les plus élevés. Le 16PF révélait également que les utilisateurs excessifs étaient
facilement affectés par leurs sentiments, moins stables émotionnellement, imaginatifs, absorbés dans leurs pensées,
préféraient leurs propres décisions, étaient autosuffisants et expérimentaient.
Conclusions : L’étude suggère que les élèves d’école secondaire de 2e cycle qui utilisent Internet à l’excès déclarent et
manifestent ultérieurement un nombre significativement plus élevé de symptômes psychiatriques que les élèves qui utilisent
Internet moins fréquemment. En outre, les utilisateurs excessifs semblent avoir un profil de personnalité distinct quand on
les compare aux non-utilisateurs, ainsi qu’aux utilisateurs minimaux et modérés

414
W Can J Psychiatry, Vol 50, No 7, June 2005