Addictive Behaviors 42 (2015) 1–8

Contents lists available at ScienceDirect

Addictive Behaviors

Multi-family group therapy for adolescent Internet addiction: Exploring

the underlying mechanisms
Qin-Xue Liu a,b, Xiao-Yi Fang c,d,⁎, Ni Yan e, Zong-Kui Zhou a,b, Xiao-Jiao Yuan f, Jing Lan c, Chao-Ying Liu c
a
Key Laboratory of Adolescent Cyberpsychology and Behavior (CCNU), Ministry of Education, Wuhan 430079, China
b
School of Psychology, Central China Normal University, Wuhan 430079, China
c
Institute of Developmental Psychology, Beijing Normal University, Beijing 100875, China
d
Academy of Psychology and Behavior, Tianjin Normal University, Tianjin 300387, China
e
Faculty of Psychology, Southwest University, Beibei 400700, China
f
School of Sociology and Psychology, Southwest University for Nationalities, Chengdu 610041, China

H I G H L I G H T S

• We apply the MFGT to the Internet addiction for the first time.
• We examine changes in the measured variables to determine underlying mechanism.
• MFGT shows a significant effect on reducing Internet addiction.
• Parent–adolescent interaction and need satisfaction contribute to the effect.

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

Available online 30 October 2014 Objective: Internet addiction is one of the most common problems among adolescents and effective treatment is
needed. This research aims to test the effectiveness and underlying mechanism of multi-family group therapy
Keywords: (MFGT) to reduce Internet addiction among adolescents.
Multi-family group therapy Method: A total of 92 participants consisting of 46 adolescents with Internet addiction, aged 12–18 years, and 46
Internet addiction
their parents, aged 35–46 years, were assigned to the experimental group (six-session MFGT intervention) or a
Family relationships
Need satisfaction
waiting-list control. Structured questionnaires were administered at pre-intervention (T1), post-intervention
Effectiveness mechanism (T2) and a three-month follow-up (T3).
Results: There was a significant difference in the decline both in the average score and proportion of adolescents
with Internet addiction in MFGT group at post-intervention (MT1 = 3.40, MT2 = 2.46, p b 0.001; 100 versus 4.8%,
p b 0.001) maintained for three months (MT3 = 2.06, p b 0.001; 100 versus 11.1%, p b 0.001). Reports from both
adolescents and parents were significantly better than those in the control group. Further explorations of the
underlying mechanisms of effectiveness based on the changed values of measured variables showed that the
improvement in adolescent Internet use was partially explained by the satisfaction of their psychological
needs and improved parent–adolescent communication and closeness.
Conclusions: The six-session multi-family group therapy was effective in reducing Internet addiction behaviors
among adolescents and could be implemented as part of routine primary care clinic services in similar populations.
As family support system is critical in maintaining the intervention effect, fostering positive parent–adolescent
interaction and addressing adolescents' psychological needs should be included in preventive programs for Internet
addiction in the future.
© 2014 Elsevier Ltd. All rights reserved.

1. Introduction addiction, also known as Pathological Internet Use, Problematic Internet

With the rapid development of the Internet, Internet addiction
has become a widespread and problematic phenomenon. Internet
⁎ Corresponding author at: Institute of Developmental Psychology, Beijing Normal
University, Beijing 100875, China. Tel.: +86 10 5880 8232; fax: +86 10 5880 8232.
E-mail address: fangxy@bnu.edu.cn (X.-Y. Fang).
Use and Compulsive Internet Use, is characterized by excessive and
compulsive Internet use and a preoccupation with and loss of control
over this use that interferes with individuals' daily functioning
(Caplan, 2002; Davis, 2001; Van den Eijinden, Meerkerk, Vermulst,
Spijkerman, & Engles, 2008; Young & Abreu, 2011). Currently, it is one
of the most common behavioral problems for adolescents, who are
more exposed to Internet use and consequently more vulnerable than

http://dx.doi.org/10.1016/j.addbeh.2014.10.021
0306-4603/© 2014 Elsevier Ltd. All rights reserved.
2 Q.-X. Liu et al. / Addictive Behaviors 42 (2015) 1–8
adults (Lortie & Guitton, 2013), with a prevalence rate higher than 8% in
some countries (Cho, Kim, Kim, Lee, & Kim, 2008; Kuss, Griffiths, &
Binder, 2013; Van den Eijnden, Spijkerman, Vermulst, van Rooij, &
Engels, 2010). In China, approximately 10% of adolescents (approxi-
mately 20 million teenagers) reported a tendency towards or current
diagnosis of Internet addiction (Block, 2008; China Internet Network In-
formation Center, 2013). Internet addiction may cause psychological
distress, personality development problems, social problems and poor
school performance (Brezing, Derevensky, & Potenza, 2010; Young,
Pistner, O'Mara, & Buchanan, 2000). In addiction, high comorbidity
with effective disorders, impulse control disorders and substance
abuse disorders has been reported (Petersen, Weymann, Schelb, Thiel,
& Thomasius, 2009; Weinstein & Lejoyeux, 2010). There is significant
research around the diagnosis, epidemiology, predicting factors and
negative outcomes of Internet addiction, but little is known about
treating it, which is an imperative for adolescents, families, schools
and society, especially in China (King, Delfabbro, Griffiths, & Gradisar,
2011;Winkler, Dörsin, Rief, Shen, & Glombiewski, 2013).
Petersen et al. (2009) conducted a survey at the request of the
German health department and argued that clinical recommendations
are not possible due to the lack of studies and that further research is
urgently needed. In a systematic review of Internet addiction treatment,
only eight studies were included. Half of them were psychological
approaches, and two utilized cognitive-behavioral therapy (King
et al., 2011). Peukert, Sieslack, Barth, and Batra (2010) also indicate
cognitive-behavioral and pharmacological approaches as potentially
effective treatments in their review. They suggest that interventions
with family members could be useful. Winkler et al. (2013) further
examine the efficacy of different treatments for Internet addiction
(13 studies included) in their meta-analysis, and their results show
that CBT did not perform significantly better than other psychological
treatments, even though it appears to be the predominant approach
for treating Internet addiction. They also suggest that both individual
counseling and group therapy have their shortcomings and that further
research around different approaches and modalities is needed. Howev-
er, there is no study that examines which factors contribute to the
efficacy of treatment or what predictors cause the behavior change to
happen, which is very important to evaluate and improve interventions
(Liu, Fang, & Zhou, 2011).
Family plays a central role in the socializing process for adoles-
cents, and parents provide emotional connection, behavioral con-
straints and modeling (Gray & Steinberg, 1999; Lau, Quadrel, &
Hartman, 1990). Family-based intervention is the most thoroughly
studied treatment modality for adolescent substance dependence
and addiction, and there is a large body of research to support its
efficacy (for a review, see Liddle, 2004). Previous research also
proved that a good relationship and communication with parents
are protective factors for adolescents from Internet addiction (Kim,
Jeong, & Zhong, 2010; Van den Eijnden et al., 2010). Family members
involved in interventions facilitate the process of recovery and help
the addict maintain a lasting effect of intervention after sessions
(Liddle, 2004; Zhong et al., 2011). Grounded in family system theory
and integrated in family and group therapy, multi-family group ther-
apy (MFGT) was proposed as a promising new approach to treat In-
ternet addiction behaviors, but no empirical study was conducted
(Liu et al., 2011). The effectiveness of MFGT has been empirically
demonstrated among adolescents with psychological disorders
(Chien & Chan, 2013; McDonell & Dyck, 2004), children at risk for
special educational services (Kratochwill, McDonald, Levin, Scalia,
& Coover, 2009) and in addiction related areas (Conner et al., 1998;
Zubrick et al., 2005). In this field, Zhong et al. (2011) found that
family-based intervention is more effective in reducing Internet ad-
diction than group therapy that involved only the adolescents. The
multi-family group offers both adults and adolescents the advan-
tages of a peer group, which help them to get support and learn
from peer confrontation. Transferential reactions occur not only
within one family but also across family lines, facilitating the group
to serve both as an arena for cross transferences based on each
person's introject and as a reality tester (Leichter & Schulman,
1974). Connection within family members is also helpful for high
treatment attendance (Nieter, Thornberry, & Brestan-Knight,
2013). Moreover, family-oriented intervention might be particularly
effective in Chinese culture, where the cohesion between family mem-
bers is highly emphasized. Therefore, the present study aims to explore
both the effectiveness of MFGT on Internet addiction and the underlying
mechanisms of the effectiveness.
One mechanism through which MFGT may effectively reduce
Internet addiction is improving parent–adolescent communication
and closeness. Compared with non-addicts, adolescents with Inter-
net addiction have poorer communication with their parents (Park,
Kim, & Cho, 2008) and are more likely to receive rejection and
negative feedback from their parents (Van den Eijnden et al.,
2010). Poor parent–adolescent communication and low perceived
parent–adolescent closeness, in turn, predicted adolescents' Internet
addiction (Liu, Fang, Deng, & Zhang, 2012; Liu, Fang, Zhou, Zhang, &
Deng, 2013). According to the Circumplex Model of Marital and Family
Systems proposed by Olson (Olson, 2000; Olson, Sprenkle, & Russell,
1974), family communication is critical in facilitating intimacy
among family members and strengthening the family's adaptability
to change. MFGT emphasizes improving family cohesion and motiva-
tion to change within the family; it not only focuses on the parent–
adolescent interaction but also values the style and strength of attach-
ment between family members (Dickerson & Crase, 2005). Therefore,
it could be a well-suited approach to treat Internet addiction among
adolescents.
The second mechanism through which MFGT may take effect in
treating adolescents' Internet addiction is by fulfilling their psychologi-
cal needs through strengthening their communication and relationship
with their parents. Psychological need is considered one of the most
important driving forces that promotes behavioral change. Fulfillment
of psychological needs through Internet use has been proposed as an in-
ternal motive in adolescents' Internet addiction (Morris & Ogan, 1996;
Suler, 1999). Adolescents' unfulfilled needs for competence and related-
ness in life and perceived need satisfaction online are the major precur-
sors of their excessive Internet use (Cai, Cui, & Li, 2007; Shen, Liu, &
Wang, 2013; Wan, Zhang, Liu, Deng, & Fang, 2010). Compared with
non-addicts, Internet addicts perceived higher need satisfaction online
and lower need satisfaction in real life (Deng, Fang, Wan, Zhang, &
Xia, 2012). Therefore, if parent–adolescent communication practices
and relationships are improved, adolescents’ psychological needs for re-
latedness or competence might be more easily fulfilled through their
daily life interactions with their parents, which, in turn, could be helpful
to reduce their reliance on the Internet for fulfilling their needs. As
fundamental as these two underlying mechanisms appear to be in af-
fecting adolescents' Internet addiction, they have nevertheless been
barely examined explicitly in prior Internet addiction intervention
studies. In this study, we include these two underlying mechanisms
as major intervening variables to examine whether the effectiveness
of family group intervention for adolescent Internet addiction de-
pends on them.
Based on a quasi-experimental design, the present study exam-
ines the effectiveness of the MFGT for adolescent Internet addiction
among 46 pairs of adolescents and their parents. The study aims
to examine three hypotheses: First, the intervention group shows a
reduction in Internet addiction both at the end of the intervention
and at a three-month follow-up compared with the control group.
Second, adolescents in the intervention group show improved com-
munication and relationship with their parents and psychological need
satisfaction in real life. Third, the effectiveness of the intervention is par-
tially explained by the improved parent–adolescent relationship and
communication, and adolescents' psychological need satisfaction in
real-life.
 Q.-X. Liu et al. / Addictive Behaviors 42 (2015) 1–8 3

2. Method 2.2. Procedures

2.1. Participants
Participants were recruited through advertisements on school
websites in Baotou City of Inner Mongolia in China. Related information
about the research and a simplified scale of Internet addiction, which is
used for clinical diagnosis, were included in the advertisement. Families
who were interested and matched the diagnosis were welcome to sign
up and have a face-to-face interview one by one. Among the 55 families
who signed up for the intervention study, 46 families were selected
based on the Adolescent Pathological Internet Use Scale (APIUS; detailed
information about APIUS is provided in the measurement section) and
inclusion criteria. The body screening scale, SCL-90 and simplified
addiction screening scale were used to exclude participants who pos-
sessed physical disabilities, mental disorders or other addictive behav-
iors. Only one boy was excluded for depression. Twenty-one families
were assigned to the intervention group because their schedules
matched with the intervention arrangement and the other 25 families
were included in the control group because they could not set up a con-
tinued intervention schedule. Families in the control group were added
to the waiting-list for the intervention study after the informed consent
from the parents and the adolescents.
The intervention group had a dominant proportion of male (n = 17)
over female (n = 4) with an average age of 15 years old (SD = 1.73).
The female-to-male ratio of parents in the intervention group was
16:5 with an average age of 40.9 years old (SD = 2.85). Nine of them
held a degree of college or above (42.9%); six of them obtained a high
school degree (28.6%) and six of them did not obtain a high school de-
gree (28.6%). The average monthly income ranges from 2000 to
10,000 Yuan with an average of 4685, which indicated middle-income
families in the city. All parents were first-time married.
The demographic compositions in the control group resembled
those in the intervention group. Adolescents in the control group had
a male-to-female ratio of 21:4 and the average age was 15.7 years old
(SD = 1.2). Families in the control group did not show significantly dif-
ferences from those in the intervention group.
First, the Manual of Adolescents Internet Addiction Family Group Ther-
apy was developed with precision based on the theoretic framework of
family group therapy, previous intervention practices, and empirical
studies. Before the intervention was launched, a pilot study was imple-
mented among six adolescents and their families to assess the operabil-
ity of the intervention design, potential problems in administering the
intervention, and smooth transitions between activity themes in each
intervention session. With preliminary results, interview feedback
from the pilot study and consultation with experts on the intervention
team, the Manual of Adolescents Internet Addiction Family Group Therapy
was modified and finalized. Then, recruited parents provided informed
consent for their adolescent children's and their own participation.
All participants were asked to complete assessments both before
(T1) and after the intervention (T2), and at a three-month follow-up
(T3) as well. Participants were assured of the confidentiality of their re-
sponses. Procedures were approved by the Institutional Review Board of
the Institute of developmental Psychology, Beijing Normal University.
The details of the procedures are presented in Fig. 1.
2.3. Intervention
After the intervention began, the 21 families in the intervention con-
dition were randomly divided into three intervention groups with
seven families in each group. Two therapists were assigned to each
group randomly and all therapists had the same clinical background
under family and group therapy training. The intervention was given
every three days, with each session lasting 2 h. Six sessions were admin-
istered for each grouped families.
The intervention was tailored to strengthen parent–adolescent com-
munication and relationships and shift adolescents' fulfillment of psy-
chological needs from the Internet to interactions and building
relationships with family members. Specific topics and activities were
designed for each intervention session and connected with each other
across six sessions, each of which included five parts in 2 h: a warm-
up exercise, feedback on homework from the last session (except the

Research design and intervention plan based on previous research and theory framework

Pilot study (n=6 family) Expert consulting

Family Group Intervention Manual for Internet Addiction

Base-line assessment (N=46 Family)

Intervention group (n= 21) Control group (n=25)

Post-assessment (N=46 Family)

3 month- follow up assessment

9
Intervention group (n=18) Control group (n=23)

Fig. 1. The intervention process and participants flow diagram.

4 Q.-X. Liu et al. / Addictive Behaviors 42 (2015) 1–8

Table 1
Comparisons of measured variables between the intervention group and control group at T1, T2 and T3.

T1 T2 T3

M(SD) M(SD) M(SD) F1 F2

Average APUIS Intervention 3.40(.27) 2.46(.61) 2.06(.73) 38.31⁎⁎⁎ 65.98⁎⁎⁎

Control 3.38(.20) 3.59(.31) 3.27(.26) 7.62⁎⁎ 2.50
t −0.54 −7.79⁎⁎⁎ −6.72⁎⁎⁎
Internet use time Intervention 26.38 (9.6) 11.43(5.75) 7.08(3.98) 40.16⁎⁎⁎ 56.65⁎⁎⁎
Control 27.08(11.1) 27.52(11.40) 22.29(6.0) 2.21 3.03
t −0.21 −4.73⁎⁎⁎ −9.39⁎⁎⁎
Parental reports Intervention 3.37(.48) 3.13(.66) 2.70(.44) 6.05⁎⁎ 21.10⁎⁎⁎
control 3.36(.52) 3.45(.72) 3.20(.57) 1.43 1.97
t 0.01 −4.12⁎⁎⁎ −5.26⁎⁎⁎
Addiction ratea Intervention 100% 4.8% 11.1%
Control 100% 96% 87%

Note. F1 indicates F statistics from comparison among three assessments; F2 indicates the F statistics from the linear test. aFive parcitipants were missing at T3.
⁎⁎ p b 0.01.
⁎⁎⁎ p b 0.001.

first session), a main structured activity, a brief summary and the family
assignment. During the sessions, the following topics were focused on:
understanding a family with Internet addict (session 1), parent–adoles-
cent communication skills training (session 2), parent–adolescent com-
munication practices on Internet addiction (sessions 2 and 3), parent–
adolescent relationship building skills training (session 4), associations
between psychological needs and Internet use and how to satisfy the
unfulfilled need in the family relationships (session 5) and setting up
appropriate and healthy expectations for the family system (session
6). One additional session at the three-month follow-up was designed
to target potential relapse, discuss new issues and generate solutions
to maintain the effectiveness of the intervention.
2.4.2. Parent–adolescent relationship
Adolescents reported their relationships with the parent who partic-
ipated in the intervention on nine items from the Closeness to Parents on
a scale (Buchnan, Maccoby, & Dornbush, 1991) from 1 (not at all true) to
5 (very much true). Sample questions included “How openly do you talk
with your mother (father)?” or “How close do you feel to your mother
(father)?” A Chinese adaption of the scale has been used in previous
study (Liu et al., 2013) and it has a high internal consistency in the
study (α = .91). Average scores across the nine items were used to rep-
resent the adolescent's relationships with the mother or the father dur-
ing the intervention.

2.4. Measurement 2.4.3. Parent–adolescent communication

2.4.1. Adolescent Internet addiction
Two indicators of Internet addiction were reported by adolescents
themselves. First, they reported their average number of hours spent
on the Internet per week over the past month. Second, they reported
on their Internet addiction behaviors with the Adolescent Pathological In-
ternet Use Scale (Lei & Yang, 2007). This scale contains 38 items with
each item being rated from 1 to 5 (1 = not true at all; 5 = true all of
the time). Six subscales were included: salience, social comfort, mood
alteration, tolerance, compulsive Internet use and negative outcomes.
The average scores across the 38 items were used as indicators of Inter-
net addiction with higher scores indicating more serious Internet addic-
tion. This scale has high internal consistency (α = .95 for the whole
scale; α between .81 and .91 for all subscales) and high test–retest reli-
ability (r = .86). Based on the APIUS average scores, adolescents with
scores below 3 were considered to be normal Internet users, adoles-
cents with scores between 3 and 3.15 were considered to have a ten-
dency towards Internet addiction, and adolescents with scores higher
than 3.15 were defined as having Internet addiction (Lei & Yang,
2007). The Internet addiction rate was calculated based on the number
of adolescents defined as having a tendency towards or having Internet
addiction (APIUS N 3) and is considered as one of the intervention effec-
tiveness indicators in the study.
Second, parents' reports of children's Internet use behaviors in the
last month were used as a supplementary measure of adolescents' Inter-
net use behaviors. Three items were reported: Internet use frequency
(1 = “very rare” to 5 = “very frequently”), parents' observation about
adolescents' Internet use appropriateness (1 = “very appropriate” to
5 = “not appropriate at all”), and parents' satisfaction towards
children's Internet use behaviors (1 = “very satisfied” to 5 = “not sat-
isfied at all”). The average score across the three items was used to pres-
ent the parents' evaluation of children's Internet use behaviors. A higher
score indicates more Internet use and lower behavior control.
The Parent–Child Communication Scale (Barnes & Olson, 1985) was
used to assess the adolescents' perception of their communication
with the parent in the intervention. This scale contains 20 items on a
scale from 1 (never) to 5 (always). It is composed of two dimensions
that measure the degree of openness and the extent of problems in fam-
ily communications. The responses were identified separately for fa-
thers and mothers. The average score across both dimensions was
used to represent the average level of parent–child communication in
this study. A Chinese adaption of the scale has been used in previous
study (Liu et al., 2012) and the α for the scale is 0.82 in the study.
2.4.4. Adolescent psychological needs
Adolescents rated their psychological needs using a scale modi-
fied from the College Students' Psychological Needs and Fulfillment
Scale (Wan et al., 2010). The scale is composed of three subscales:
the degree of psychological needs; need satisfaction from real life,
and need satisfaction from the Internet. Each subscale includes 35
items that tap into eight dimensions of the targeted subscale: need
for autonomy, need for entertainment, need for interaction, need
for achievement, need for impact, need for acknowledgement, need
for expression and need for information. The rating for each item in
the degree of psychological needs subscale ranges from 1 (not strong
at all) to 5 (extremely strong). Ratings in the other two need satisfac-
tion subscales range from 1 (very low) to 5 (very high). Unsatisfied
psychological needs were calculated by subtracting the subscale
scores on need satisfaction in real life from the subscale scores on
the degree of psychological needs. The advantage of the Internet in
satisfying needs was calculated by subtracting the subscale scores
on need satisfaction in real life from the subscale scores on need sat-
isfaction from the Internet. These two scores were two major process
variables of interest in the study. The α for the scale is 0.97 in the
study.
 Q.-X. Liu et al. / Addictive Behaviors 42 (2015) 1–8 5

Table 2
Comparisons on intervening process variables.

T1 T2 T3

M(SD) M(SD) M(SD) F1 F2

P–A relationship Intervention 2.99(0.45) 3.72(0.81) 3.79(0.64) 14.86⁎⁎⁎ 29.94⁎⁎⁎

Control 2.96(0.44) 2.99(0.49) 3.05(0.48) 1.45 2.84
t 0.21 2.89⁎⁎ 2.96⁎⁎
P–A communication Intervention 2.96(0.71) 3.71(0.49) 3.83(0.62) 13.78⁎⁎⁎ 30.57⁎⁎⁎
Control 2.94(0.44) 2.98(0.47) 3.03(0.42) 0.08 0.11
t 0.19 4.08⁎⁎⁎ 3.97⁎⁎⁎
Unsatisfied needs Intervention 0.67(0.58) 0.19(0.56) −0.01(0.45) 11.56⁎⁎⁎ 19.85⁎⁎⁎
Control 0.65(0.37) 0.47(0.25) 0.48(0.29) 3.48 4.65
t 0.18 −2.57⁎ −3.87⁎⁎
Advantage of Internet in satisfying need Intervention 0.59(0.57) −0.24(0.59) −0.81(0.58) 23.68⁎⁎⁎ 39.15⁎⁎⁎
Control 0.58(0.44) 0.46(0.25) 0.47(0.29) 3.41 2.61
t 0.09 −4.83⁎⁎⁎ −7.78⁎⁎⁎

Note. P–A stands for parent–adolescent; F1 indicates F statistics from comparison among three assessments; F2 indicates the F statistics from the linear test.
⁎ p b 0.05.
⁎⁎ p b 0.01.
⁎⁎⁎ p b 0.001.

2.5. Analysis plan
The data analysis proceeded in three steps. First, T tests and
Repeated-Measures ANOVA analyses were conducted to test the effec-
tiveness of the intervention based on the comparison among adoles-
cents' Internet addiction measures in the intervention and control
groups at T1, T2 and T3. Second, comparisons on all intervening vari-
ables between the two groups at T1, T2 and T3 were conducted, and
changed values of the Intervening variables from T1 to T2 and T3 were
created. Third, hierarchical multiple regression analyses were conduct-
ed to examine whether the change in adolescents' Internet addition be-
haviors is explained by the change in the measured intervening
variables.
3. Results
3.1. Effectiveness of the intervention
Before examining the effectiveness of the intervention, the Internet
addiction measures at T1 were compared between the intervention
and control groups and no significant difference was detected, indicat-
ing that the two groups were at the same or similar level of Internet ad-
diction at the baseline of the study (see F values at T1 in Table 1). The
effectiveness of the intervention was manifested in three aspects:
First, adolescents in the intervention group significantly reduced the
time they spent on the Internet by the end of the intervention, spending
about half of the time as adolescents in the control group did (see F
values at T2 in Table 1). Second, comparisons of APIUS scores demon-
strated that the intervention group experienced a decrease in their aver-
age APIUS scores from T1 to T2. Third, parents in the intervention group
reported more satisfaction with adolescents' Internet use behaviors at
the end of the intervention compared with both their satisfaction at
the baseline and parents' satisfaction in the control group. In sum,
based on reports from both adolescents and their parents, the interven-
tion was effective in terms of reducing adolescents' Internet addiction
behaviors by the end of the intervention.
Results from the repeated measures ANOVA (see Table 1) showed
that the differences in APIUS scores across the three measurements
were significant(F (2, 78) = 38.31, p b .001) and also displayed a linear
decrease over time from Time 1 to Time 3 (F (1, 39) = 65.98, p b .001),
which indicates that the intervention effects remained at T3. Time
spent on the Internet also displayed a significant decrease from T1 to T3
(F (2, 78) = 40.16, p b .001) with a linear decrease from T1 to T3 as
well (F (1, 39) = 56.65, p b .001). In the control group, significant dif-
ferences were found among the three APIUS assessments and a further
paired sample t test showed that APIUS scores at T2 were significant
higher than at T1(t = − 4.15, p b 0.001), but there was no linear ten-
dency of APIUS scores from T1 to T3 (F (1, 39) = 2.50, p = 0.13).

Table 3
Regressions for changed values of measured variables at T2 and T3

ΔX = XT2 − XT1 ΔX = XT3 − XT1

β t R2 F ΔR2 ΔF β t R2 F ΔR2 ΔF

Step 1
Δ P–A relationship −.56 −2.93⁎⁎ .31 8.55⁎⁎ .31 8.55⁎⁎ −.61 −2.62⁎ .36 6.76⁎ .36 6.76⁎

Step 2
Δ P–A relationship −.03 −.13 .65 16.99⁎⁎⁎ .34 17.87⁎⁎ −.03 −.11 .55 6.79⁎ .19 4.74⁎
Δ P–A communication −.78 −4.22⁎⁎ −.73 −2.19⁎

Step 3
Δ P–A relationship .03 .17 .70 7.07⁎⁎ .05 .82 .00 .01 .83 10.01⁎⁎ .27 6.49⁎
Δ P–A communication −.51 −1.98⁎ −.53 −1.98⁎
Δ Unsatisfied needs .02 .13 .34 1.00
Δ Advantage of Internet in satisfying needs .39 1.33 .88 3.44⁎⁎

Note. P–A stands for parent–adolescent.

⁎ p b 0.05.
⁎⁎ p b 0.01.
⁎⁎⁎ p b 0.001.
6 Q.-X. Liu et al. / Addictive Behaviors 42 (2015) 1–8
Parents' reports on adolescent Internet addiction at the three-month
follow-up also substantiated the effect of the intervention. At 3 months
after the intervention, parents still reported a significant reduction in
their perception of their children's Internet addiction, revealing a signif-
icant linear decrease from T1 to T3.
As an important indicator of intervention effectiveness in the whole
intervention group, the Internet addiction rate yielded from the APIUS
average scores was also compared across the intervention and control
groups over time. As displayed in Table 1, all adolescents in both groups
were either addicted to the Internet or had the tendency towards Inter-
net addiction before the intervention. By the end of the intervention,
only 1 out of 21 adolescents (4.8%) in the intervention group was still
addicted to the Internet, compared with as many as 24 out of 25
(96%) in the control group. At the three-month follow-up, two adoles-
cents in the intervention group showed a relapse. However, the inter-
vention effects remained as only 11.1% of adolescents in the
intervention group remained addicted after the intervention ended,
compared with 87% of their counterparts in the control group.
3.2. Examining mediating effects of intervening variables
Because we measured the perception of parent–adolescent interac-
tion and psychological need that is deeply inside of adolescents, we
used the reports from the adolescents themselves to examine the
change process of adolescent Internet behaviors. To examine the second
hypothesis (that the intervention effects are mediated through major
process variables), we first examined the change in parent–adoles-
cent relationships, parent–adolescent communication and adoles-
cents' psychological needs. We measured these attributes at their
baseline, at the end of the intervention, and at a three-month follow-
up for both the intervention and control groups.
As displayed in Table 2, the measured variables did not differ signif-
icantly between the intervention and control groups at T1, but did differ
significantly at T2 and T3. Adolescents in the intervention group dem-
onstrated improvement in their relationship and communication with
their parents, an increase in their fulfillment of needs, a decrease in In-
ternet use' advantage in fulfilling their needs in a linear function.
With the demonstration of improvement on all process variables in
the intervention group, we continued to examine whether the change
in those process variables accounted for change in adolescents' Internet
addiction behaviors. Before examining those mediating effects, change
values (Δ X) were created for adolescent reported Internet addiction
behaviors, parent–adolescent communication and relationship, and
psychological needs by subtracting the post-intervention assessment
values from the baseline values (ΔX = XT2 − XT1).
To examine the contribution of those process variables in the expla-
nation of intervention effectiveness on adolescents' Internet addiction, a
hierarchical multiple regression analysis was performed. Process vari-
ables were entered in three steps: In step 1, ΔAPIUS was the dependent
variable and Δ parent–adolescent relationship was the independent var-
iable. In step 2, Δ parent–adolescent communication was entered into the
step 1 equation. In step 3, two indicators of Δ adolescents' psychological
needs were added in the equation in step 2. The results of step 1
(Table 3) indicated that 31% of the variance in the change in adolescents'
Internet addiction (R2 = .31) was accounted for by improvements in
the parent–adolescent relationship. In step 2, after entering changes in
mother–adolescent communication, an additional 34% of the variance
in the change in adolescents' Internet addiction was explained(ΔR2 =
.34) and this change in R2 was significant. In step 3, the addition of psy-
chological needs did not explain additional variance in the dependent
variable.
To examine the mediating effects of process variables from T1 to T3,
change values (Δ X) were created by following the same procedure:
subtracting the three-month follow-up assessment values from the
baseline values (ΔX = XT3 − XT1). A hierarchical multiple regression
analysis was again performed (see the right half of Table 3). First, the
results of the hierarchical analysis revealed that changes in parent–ado-
lescent communication and relationships explaining significant vari-
ance in adolescents' Internet addiction from T1 to T3. Moreover, when
predicting longer term effects in reducing Internet addiction, changes
in adolescents' psychological needs also added significant prediction
of variance in the Internet addiction change (Δ R2 = .27). This was in
large part due to changes in adolescents' satisfaction of their psycholog-
ical needs through the Internet (β = .88, p b .01).
4. Discussion
The current study represents a practical clinical trial of treating
adolescents' Internet addiction using the multi-family group approach.
Based on prior family group intervention practices in treating adoles-
cent psychopathology (Chien & Chan, 2013; McDonell & Dyck, 2004;
Zhong et al., 2011), the current study is the first to our knowledge to
apply the approach of MFGT in treating adolescents' Internet addiction.
The adolescents' Internet addiction rate dropped from 100% at the base-
line assessment to 4.8% at the end of the intervention and remained at
11.1% at the three-month follow-up assessment. Time spent on Internet
in the intervention group also significantly declined throughout the
intervention until the three-month follow-up. Analyses of the value
changes in measured variables indicated which factors were associated
with the decrease of adolescent Internet addiction. Improved parent–
adolescent communication and need satisfaction in real life were asso-
ciated with decrease in Internet addiction. If adolescents in the inter-
vention group perceived an improvement in their communication and
relationships with parents, learned alternative ways to fulfill their
needs and felt less reliance on the Internet, this might promote their
motivation to sustainably change their behavior. The results are consis-
tent with prior evidence suggesting the role of feeling supported and
trusted in improving the effectiveness of family group intervention
(Dickerson & Crase, 2005).
These findings corroborate the idea that intervention programs for
adolescents need to get parents actively involved and include them as
part of the solution. The family system approach shifts the emphasis
on individual family members to the entire family as a unit and the dy-
namic interactions between family members (Dickerson & Crase, 2005).
Further, multi-family group is very helpful in the lasting the effective of
intervention. In the multi-family group, each family represents a sub-
system with a shared history and current life situation which makes
for an enriching and complex process. Then, the multi-group serves
both the family system and individual as an arena for cross transfer-
ences and as a reality tester (Leichter & Schulman, 1974). Moreover,
the participation of other family members in the intervention can create
a more supportive environment in which the participants' behavioral
changes are valued, encouraged, and maintained even after the inter-
vention ends (McDonell & Dyck, 2004).
The present study takes a further step in unraveling the underlying
mechanism through which MFGT took effects. Parent–adolescent inter-
actions and relationship quality and adolescents' increasing satisfaction
of psychological needs in real life partially accounted for the effective-
ness of the intervention in reducing adolescents' Internet addiction,
further supporting prior findings (Liu et al., 2012, 2013; Olson et al.,
1974). Consistent with previous studies, perceived positive interaction
with parents protected adolescents from Internet addiction (Liu et al.,
2013; Van den Eijnden et al., 2010). Positive communication facilitated
emotional connections between family members and helped them to
understand and clarify their needs (White, 2000). The alternative
ways and skills of need satisfaction that adolescents and parents learned
from the sessions would also expand the degree of communication and
interaction within the family. All of these changes could be helpful to
keeping children away from Internet addiction. However, the decrease
in advantage of Internet in satisfying need did not predict change in
Internet use behaviors until the three-month follow-up assessment.
 Q.-X. Liu et al. / Addictive Behaviors 42 (2015) 1–8
This delayed effect has been reported before in interventions targeted to
treating adolescents' affective disorders (Goldberg-Arnold, Fristad, &
Gavazzi, 1999). It suggests that improvement in adolescents' satisfac-
tion of psychological needs through real life interactions may take
time to occur. The positive and effective interaction skills and patterns
between adolescents and parents need to be practiced before adoles-
cents gradually feel comfortable and natural enough to get used to
them. It also may take time for adolescents to change and adapt their
need satisfaction habits, which may be a potential advantage of MFGT
as it may have lasting intervention effects. Once a benign interaction
and relationship pattern is established, the family obtains a built-in
force to sustain the intervention effects (McDonell & Dyck, 2004).
Some factors limit the conclusions that can be drawn from these
data. First, the study did not take other competing intervention models
into consideration. It has been recommended in intervention studies to
compare across multiple intervention paradigms to better evaluate the
effectiveness of certain intervention approach. For example, a family
therapy approach or group intervention approach could be incorporat-
ed in a comparison group to further evaluate the effectiveness of the
MFGT paradigm. Second, the gender ratio of parents was not balanced,
as most of the participating parents were mothers. This might be an ob-
stacle in clarifying the unique role of the mother–adolescent relation-
ship or father–adolescent relationship in preventing Internet addiction
among adolescents, as the impact of their relationship with their father
and mother differed with adolescent genders on Internet addiction (Liu
et al., 2013). It would be clearer if further clinical studies endeavored to
recruit parents of both genders. Third, based on a quasi-experimental
design, participants were not randomly assigned to an intervention or
control group. However, the baseline comparison between the groups
did not show a significant difference and indicate that the results of
the study are reliable. Fourth, the data were mainly from questionnaires,
and social desirability might influence adolescents' reports of their
Internet use behavior. Future studies should assess Internet addiction
through psychological interviews. Additionally, the present study only
measured general Internet use and addiction, limiting the generaliza-
tion of results to specific Internet addiction. It would be helpful to im-
prove intervention if further studies paid attention to the treatment of
subtypes of Internet addiction.
Given the different characteristics of adolescents and their families,
multiple dimensions of multi-family group therapy might be more
complex and require exploration in future studies. The existence of
equifinality in adolescents' Internet addiction behaviors also indicates
that multiple factors, including personal, interpersonal and environ-
mental factors, could lead to the emergence of Internet addiction
among adolescents. For example, individual differences among children
in basic need satisfaction in daily real life were associated with the way
in which they engage with the Internet (Shen et al., 2013). Therefore,
multi-family group interventions that tailor to different etiologies of
Internet addiction might improve its effectiveness.
In the addiction field, prevention is more important than intervention.
Future studies might explore the possibility and feasibility of an integrat-
ed preventive intervention framework including school, community and
family, based on evidence from other related adolescent health-related
domain (Dodge & Godwin, 2013; Lovato et al., 2013). In consideration
of the particularity of Internet addiction behaviors, namely that Internet
serves as a necessary part of life and Internet use cannot be completely
cut off but only be guided to a rational and controllable level, it is very im-
portant to discern how to guide youth use of the Internet appropriately
and optimize the benefits of the Internet for adolescents and children.
Role of funding sources
This study was funded by Project of Social Sciences for Young Scholars from Ministry
of Education in China (Project No. 12YJC190023), China National Science Foundation (Pro-
ject No. 31170990), the Program for Changjiang Scholars, and the Fundamental Research
Funds of Central China Normal University (CCNU13A05046). The funders had no role in
study design, data collection and analysis, decision to publish, or preparation of the
manuscript.
7
Contributors
Dr. Xiaoyi Fang generated the idea for the study and designed the study and interven-
tion program. Dr. Qinxue Liu conducted all the interventions and statistical analyses as
well as wrote the full manuscript. Ni Yan contributed the English Writing, methods and
discussion. Zongkui Zhou contributed the modification of discussion. Xiaojiao Yuan,
Jing Lan and Chaoying Liu participated in the intervention delivery and manuscript
preparation. All authors contributed to and approved the final manuscript.
Conflict of interest
All authors declare they have no conflicts of interest.
Acknowledgments
This research was supported by Project of Social Sciences for
Young Scholars from the Ministry of Education in China (Project No.
12YJC190023), China National Science Foundation (Project No.
31170990), the Program for Changjiang Scholars, and the Fundamental
Research Funds of Central China Normal University (CCNU13A05046).
We thank Nan Zhou and Hong-Jian Cao for their contributions to the
intervention. We also thank all the participants for their efforts on the
study.
References
Barnes, H.L., & Olson, D.H. (1985). Parent–adolescent communication and circumplex
model. Child Development, 56, 438–447.
Block, J.J. (2008). Issues for DSM-V: Internet addiction. The American Journal of Psychiatry,
165, 306–307.
Brezing, C., Derevensky, J.L., & Potenza, M.N. (2010). Non-substance-addictive behaviors
in youth: Pathological gambling and problematic internet use. Child and Adolescent
Psychiatric Clinics of North America, 19(3), 625–637.
Buchnan, C.M., Maccoby, E.E., & Dornbush, S.M. (1991). Caught between parents:
Adolescents' experience in divorced homes. Child Development, 62, 1008–1029.
Cai, Y.Y., Cui, L.J., & Li, X. (2007). A research on the psychological needs of teenagers'
online game behaviors. Psychological Science, 30(1), 169–172 (in Chinese).
Caplan, S.E. (2002). Problematic internet use and psychosocial wellbeing: Development of
a theory-based cognitive-behavioral measurement instrument. Computers in Human
Behavior, 18, 553–575.
Chien, W.T., & Chan, S.W.C. (2013). The effectiveness of mutual support group interven-
tion for Chinese families of people with schizophrenia: A randomized controlled
trial with 24-month follow-up. International Journal of Nursing Studies, 50,
1326–1340.
China Internet Network Information Center (2013r). Statistical report on the develop-
ment of internet in China. (2013, Jan).
Cho, S.C., Kim, J.W., Kim, B.N., Lee, J.H., & Kim, E.H. (2008). Biogenetic temperament
and character profiles and attention deficit hyperactivity disorder symptoms in
Korean adolescents with problematic internet use. Cyberpsychology & Behavior,
11(6), 735–737.
Conner, K.R., Shea, R.R., McDermott, M.P., Grolling, R., Tocco, R.V., & Baciewicz, G. (1998).
The role of multifamily therapy in promoting retention in treatment of alcohol and
cocaine dependence. The American Journal on Addictions, 7(1), 61–73.
Davis, R.A. (2001). A cognitive-behavioral model of pathological internet use (PIU).
Computers in Human Behavior, 17(2), 187–195.
Deng, L.Y., Fang, X.Y., Wan, J.J., Zhang, J.T., & Xia, C.C. (2012). The relationship of
psychological needs and need gratification with internet addiction among college
students. Journal of Psychological Science, 35(1), 123–128 (in Chinese).
Dickerson, A.D., & Crase, S.J. (2005). Parent–adolescent relationships: The influence of
multi-family therapy group on communication and closeness. The American Journal
of Family Therapy, 33(1), 45–59.
Dodge, K.A., & Godwin, J. (2013). The conduct problems prevention research group.
Social-information-processing patterns mediate the impact of preventive interven-
tion on adolescent antisocial behavior. Psychological Science, 24(4), 456–465.
Gray, M.R., & Steinberg, L. (1999). Unpacking authoritative parenting: Reassessing a mul-
tidimensional construct. Journal of Marriage and Family, 61, 574–587.
Goldberg-Arnold, J.S., Fristad, M.A., & Gavazzi, S.M. (1999). Family psychoeducation: Giv-
ing caregivers what they want and need. Family Relations, 48, 411–417.
Kim, D. H., Jeong, E. J., & Zhong, H. (2010). Preventive role of parents in adolescent prob-
lematic Internet game use in Korea. Korean Journal of Sociology, 44(6), 111–133.
King, D.L., Delfabbro, P.H., Griffiths, M.D., & Gradisar, M. (2011). Assessing clinical trials of
internet addiction treatment: A systematic review and CONSORT evaluation. Clinical
Psychology Review, 31, 1110–1116.
Kratochwill, T.R., McDonald, L., Levin, J.R., Scalia, P.A., & Coover, G. (2009). Families and
schools together: An experimental study of multi-family support groups for children
at risk. Journal of School Psychology, 47(4), 245–265.
Kuss, D.J., Griffiths, M.D., & Binder, J.F. (2013). Internet addiction in students: Prevalence
and risk factors. Computers in Human Behavior, 29, 959–966.
Lau, R.R., Quadrel, M.J., & Hartman, K.A. (1990). Development and change of young adults'
preventive health beliefs and behavior: Influence from parents and peers. Journal of
Health and Social Behavior, 31, 240–259.
Lei, L., & Yang, Y. (2007). The development and validation of adolescent pathological
internet use scale. Acta Psychologica Sinica, 39(4), 688–696 (in Chinese).
8 Q.-X. Liu et al. / Addictive Behaviors 42 (2015) 1–8
Leichter, E., & Schulman, G. (1974). Multi-family group therapy: A multidimensional
approach. Family Process, 13, 95–110.
Liddle, H.A. (2004). Family-based therapies for adolescent alcohol and drug use: Research
contributions and future research needs. Addiction, 99(Suppl. 2), 76–92.
Liu, Q.X., Fang, X.Y., Zhou, Z.K., Zhang, J.T., & Deng, L.Y. (2013). Perceived parent–ad-
olescent relationship, perceived parental online behaviors and pathological internet
use among adolescents: Gender-specific differences. PloS ONE, 8(9), e75642.
Liu, Q.X., Fang, X.Y., Deng, L.Y., & Zhang, J.T. (2012). Parent–adolescent communication,
parental internet use and internet-specific norms and pathological internet use
among Chinese adolescents. Computers in Human Behavior, 28, 1269–1275.
Liu, Q.X., Fang, X.Y., & Zhou, N. (2011). A review of the research on internet addiction
among Chinese adolescents. Journal of South China Normal University (Social Science
edition), 3, 65–70.
Lortie, C.L., & Guitton, M.J. (2013). Internet addition assessment tools: Dimensional struc-
ture and methodological status. Addiction, 108, 1207–1216.
Lovato, C., Watts, A., Brown, K.S., Lee, D., Sabiston, C., Nykiforuk, C., et al. (2013). School
and community predictors of smoking: A longitudinal study of Canadian high school.
American Journal of Public Health, 103(2), 362–368.
McDonell, M.G., & Dyck, D.G. (2004). Multiple-family group treatment as an effective in-
tervention for children with psychological disorders. Clinical Psychology Review, 24,
685–706.
Morris, M., & Ogan, C. (1996). The internet as mass medium. Journal of Communication,
46(1), 29–50.
Nieter, L., Thornberry, T., Jr., & Brestan-Knight, E. (2013). The effectiveness of group par-
ent–child interaction therapy with community families. Journal of Child and Family
Studies, 22, 490–501.
Olson, D.H. (2000). Circumplex model of marital and family systems. Journal of Family
Therapy, 22(2), 144–167.
Olson, D.H., Sprenkle, D.H., & Russell, C.S. (1974). Circumplex model of marital and family
systems: I. Cohesion and adaptability dimensions, family types, and clinical
applications. Family Process, 18, 3–28.
Park, S.K., Kim, J.Y., & Cho, C.B. (2008). Prevalence of internet addiction and correlations
with family factors among South Korean adolescents. Adolescence, 172(43), 895–907.
Petersen, K.U., Weymann, N., Schelb, Y., Thiel, R., & Thomasius, R. (2009). Pathological
internet use — Epidemiology, diagnostics, co-occurring disorders and treatment.
Fortschritte Der Neurologie Psychiatrie, 77, 263–271.
Peukert, P., Sieslack, S., Barth, G., & Batra, A. (2010). Internet- and computer game addic-
tion: Phenomenology, comorbidity, etiology, diagnostics and therapeutic implica-
tions for the addictives and their relatives. Psychiatrische Praxis, 37, 219–224.
Shen, C.X., Liu, R.D., & Wang, D. (2013). Why are children attracted to the internet? The
role of need satisfaction perceived online and perceived in daily real life. Computers
in Human Behavior, 29, 185–192.
Suler, R. (1999). To get what you need: Healthy and pathological internet use.
Cyberpsychology & Behavior, 5(2), 385–393.
Van den Eijinden, R., Meerkerk, G.J., Vermulst, A.A., Spijkerman, R., & Engles, R. (2008).
Online communication, compulsive internet use, and psychosocial well-being
among adolescents: A longitudinal study. Developmental Psychology, 44(3), 655–665.
Van den Eijnden, R.J.J.M., Spijkerman, R., Vermulst, A.A., van Rooij, T.J., & Engels, R.C.M.E.
(2010). Compulsive internet use among adolescents: Bidirectional parent–child
relationships. Journal of Abnormal Child Psychology, 38(1), 77–89.
Wan, J.J., Zhang, J.T., Liu, Q.X., Deng, L.Y., & Fang, X.Y. (2010). Development of college stu-
dents’ psychological need internet gratification questionnaire. Studies of Psychology
and Behavior, 8(2), 118–125 (in Chinese).
Weinstein, A., & Lejoyeux, M. (2010). Internet addiction or excessive internet use.
The American Journal of Drug and Alcohol Abuse, 36, 277–283.
White, F.A. (2000). Relationship of family socialization processes to adolescent moral
thought. The Journal of Social Psychology, 140(1), 75–91.
Winkler, A., Dörsin, B., Rief, W., Shen, Y., & Glombiewski, J.A. (2013). Treatment of internet
addiction: A meta-analysis. Clinicial Psychological Review, 33, 317–329.
Young, K.S., Pistner, M., O'Mara, J., & Buchanan, J. (2000). Cyber-disorders: The mental
health concern for the new millennium. Cyberpsychology & Behavior, 3(5), 475–479.
Young, K.S., & Abreu, C. (2011). Internet addiction. A handbook and guide to evaluation and
treatment. Hoboken, NJ: John Wiley & Sons.
Zhong, X., Zu, S., Sha, S., Tao, R., Zhao, C., & Yang, F. (2011). The effect of a family-based
intervention model on internet-addicted Chinese adolescents. Social Behavior and
Personality, 39(8), 1021–1034.
Zubrick, S.R., Ward, K.A., Silburn, S.R., Lawrence, D., Williams, A.A., Blair, E., et al. (2005).
Prevention of child behavior problems through universal implementation of a
group behavioral family intervention. Prevention Science, 6(4), 287–303.