Content uploaded by Patti M. Valkenburg
Author content
All content in this area was uploaded by Patti M. Valkenburg on Aug 27, 2021
Content may be subject to copyright.
Online Communication and Adolescent
Well-Being: Testing the Stimulation Versus
the Displacement Hypothesis
Patti M. Valkenburg
Jochen Peter
The Amsterdam School of Communications Research (ASCoR)
University of Amsterdam
The aim of this study was to contrast the validity of two opposing explanatory hypothe-
ses about the effect of online communication on adolescents’ well-being. The displace-
ment hypothesis predicts that online communication reduces adolescents’ well-being
because it displaces time spent with existing friends, thereby reducing the quality of
these friendships. In contrast, the stimulation hypothesis states that online communica-
tion stimulates well-being via its positive effect on time spent with existing friends and
the quality of these friendships. We conducted an online survey among 1,210 Dutch
teenagers between 10 and 17 years of age. Using mediation analyses, we found support
for the stimulation hypothesis but not for the displacement hypothesis. We also found
a moderating effect of type of online communication on adolescents’ well-being: Instant
messaging, which was mostly used to communicate with existing friends, positively pre-
dicted well-being via the mediating variables (a) time spent with existing friends and
(b) the quality of these friendships. Chat in a public chatroom, which was relatively
often used to talk with strangers, had no effect on adolescents’ well-being via the medi-
ating variables.
doi:10.1111/j.1083-6101.2007.00368.x
Introduction
Opportunities for adolescents to form and maintain relationships on the Internet
have multiplied in the past few years. Not only has the use of Instant Messaging
(IM) increased tremendously, but Internet-based chatrooms and social networking
sites are also rapidly gaining prominence as venues for the formation and main-
tenance of personal relationships. In recent years, the function of the Internet
has changed considerably for adolescents. Whereas in the 1990s they used the
Internet primarily for entertainment (Valkenburg & Soeters, 2001), at present they
predominantly use it for interpersonal communication (Gross, 2004; Lenhart,
Madden, & Hitlin, 2005).
Journal of Computer-Mediated Communication
Journal of Computer-Mediated Communication 12 (2007) 1169–1182 ª2007 International Communication Association1169
The rapid emergence of the Internet as a communication venue for adolescents
has been accompanied by diametrically opposed views about its social consequences.
Some authors believe that online communication hinders adolescents’ well-being
because it displaces valuable time that could be spent with existing friends (e.g.,
Kraut et al., 1998; Nie, 2001; Nie, Hillygus, & Erbring, 2002). For example, Kraut
et al. (1998) argue that ‘‘by using the Internet, people are substituting poorer quality
social relationships for better relationships, that is, substituting weak ties for strong
ones’’ (p. 1028). Adherents of this displacement hypothesis assume that the Internet
motivates adolescents to form online contacts with strangers rather than to maintain
friendships with their offline peers. Because online contacts are seen as superficial
weak-tie relationships that lack feelings of affection and commitment, the Internet is
believed to reduce the quality of adolescents’ existing friendships and, thereby, their
well-being.
Conversely, other authors suggest that online communication may enhance the
quality of adolescents’ existing friendships and, thus, their well-being. Adherents of
this stimulation hypothesis argue that more recent online communication technol-
ogies, such as IM, encourage communication with existing friends (Bryant, Sanders-
Jackson, & Smallwood, 2006). Much of the time adolescents spend alone with
computers is actually used to keep up existing friendships (Gross, 2004; Subrahmanyam,
Kraut, Greenfield, & Gross, 2000; Valkenburg & Peter, 2007). If adolescents use the
Internet primarily to maintain contacts with their existing friends, the prerequisite
for a displacement effect is not fulfilled. After all, if existing friendships are main-
tained through the Internet, it is implausible that the Internet reduces the quality of
these friendships and, thereby, adolescents’ well-being (Valkenburg & Peter, 2007).
Several studies have investigated the effect of Internet use on the quality of
existing relationships and well-being. Some of these studies used depression or
loneliness measures as indicators of well-being; others employed measures of life-
satisfaction or positive/negative affect. The studies have provided mixed results:
Some have yielded results in agreement with the displacement hypothesis (Kraut
et al., 1998; Morgan & Cotten, 2003, for surfing; Nie, 2001; Nie, Hillygus, & Erbring,
2002; Weiser, 2001). Others have produced results in support of the stimulation
hypothesis. They demonstrated, for example, that Internet use is positively related to
time spent with existing friends (Kraut et al., 2002), to the closeness of existing
friendships (Valkenburg & Peter, 2007), and to well-being (Kraut et al., 2002, study 1;
Morgan & Cotten, 2003, for email and chat; Shaw & Gant, 2002). Finally, several
other studies produced no significant results (Gross, 2004; Kraut et al., 2002, study 2;
Jackson, von Eye, Barbatsis, Biocca, Fitzgerald, & Zhao, 2004; LaRose, Ghuay, &
Bovin, 2002; Mesch, 2001, 2003; Sanders, Field, Diego, & Kaplan, 2000; Waestlund,
Norlander, & Archer, 2001).
At least one omission in earlier research may contribute to the inconsistent
findings regarding the Internet-well-being relationship. Most research to date has
been descriptive or exploratory in nature. The studies investigate direct linear rela-
tionships between Internet use and one or more dependent variables, such as social
1170Journal of Computer-Mediated Communication 12 (2007) 1169–1182 ª2007 International Communication Association
involvement, depression, or loneliness (Matei & Ball-Rokeach, 2001). Hardly any
research has been based on a-priori explanatory hypotheses regarding how Internet
use is related to well-being. More importantly, there is no research that contrasts
opposing explanatory hypotheses in the same study. With some exceptions (LaRose
et al., 2001; Morgan & Cotten, 2003; Weiser, 2001), most research has conceptualized
the relationship between Internet use and well-being as a simple stimulus-response
process. Little research has hypothesized possible mediating variables that might
cause a displacement or stimulating effect of Internet use on well-being.
The main aim of this study is to fill the gap in earlier research and pit the
predictions of the displacement hypothesis against those of the stimulation hypoth-
esis. By empirically studying the validity of the processes proposed by the two
hypotheses, we hope to improve theory formation and contribute to a more pro-
found understanding of the social consequences of the Internet. In fact, the two
hypotheses are based on the same two mediators. Both hypotheses state that online
communication affects adolescents’ well-being through its influence on (1) their
time spent with existing friends and (2) the quality of these friendships. However,
the displacement hypothesis assumes a negative effect from online communication
on time spent with existing friends, whereas the stimulation hypothesis predicts
a positive relationship between these two variables. The two opposing hypotheses
are stated below and modeled through paths 1a and 1b in Figure 1:
H1a: Online communication will reduce time spent with existing friends.
H1b: Online communication will enhance time spent with existing friends.
As Figure 1 shows, apart from paths 1a and 1b, the remaining assumptions of
the displacement and stimulation hypotheses are similar. Neither hypothesis
predicts a direct relationship between online communication and well-being. Rather,
both suggest that the influence of online communication on well-being will be
mediated by the quality of friendships. There is general agreement that the quality
of friendships is an important predictor of well-being (Hartup & Stevens, 1997).
Quality friendships can form a powerful buffer against potential stressors in adoles-
cence (Bukowski, 2001; Hartup, 2000), and adolescents with high-quality friendships
Displacement Hypothesis
_
Internet
communication
Time spent
with friends
Quality of
friendships Well-being
1a 2 3
__
Internet
communication
Time spent
with friends
Quality of
friendships Well-being
1b 2 3
+__
Stimulation Hypothesis
Figure 1 The displacement and the stimulation hypothesis modeled
Journal of Computer-Mediated Communication 12 (2007) 1169–1182 ª2007 International Communication Association1171
are often more socially competent and happier than adolescents without such friend-
ships (Hartup & Stevens, 1997). Based on these considerations, we hypothesize that if
online communication influences well-being, it will be through its influence on the
quality of existing friendships. Our second hypothesis, which is modeled via path 3 in
Figure 1, therefore states:
H2: Adolescents’ quality of friendships will positively predict their well-being and act as
a mediator between online communication and well-being.
However, the relationship between online communication and the quality of
friendships may also not be direct. Both the displacement and stimulation hypoth-
eses assume that time spent with existing friends acts as a mediator between online
communication and the quality of friendships. Based on these assumptions, we
hypothesize an indirect relationship between online communication and the quality
of friendships, via the time spent with existing friends (see paths 1a, 1b, and 2 in
Figure 1):
H3: Adolescents’ time spent with friends will predict the quality of their friendships and act as
a mediator between online communication and the quality of friendships.
Type of Online Communication: IM Versus Chat
In earlier Internet effects studies, the independent variable Internet use has often
been treated as a one-dimensional concept. This may be another important reason
why the findings of these studies are so mixed (Baym, Zhang, & Lin, 2004). Many
studies only employed a measure of daily or weekly time spent on the Internet and
did not distinguish between different types of Internet use, such as surfing or online
communication (e.g., Kraut et al., 1998, 2002). Such a simple conceptualization of
the independent variable was already problematic when investigating traditional
broadcast media (Baym, Zhang, & Lin, 2004; Jung, Qio, & Kim, 2001), but it
becomes even more problematic when researching effects of multi-use platforms
such as the Internet (Jung, Qio, & Kim, 2001).
It is quite possible that daily time spent on the Internet does not affect one’s well-
being, whereas certain types of Internet use do have such an effect. In this study, we
focus on the type of Internet use that is theoretically most likely to influence well-
being and the quality of existing friendships: online communication. We believe that
if the Internet influences well-being, it will be through its potential to alter the nature
of social interaction through the use of online communication technologies. In this
study, well-being is defined as happiness or a positive evaluation of one’s life in
general (Diener, 1984; Diener, Suh, Luca, & Smith, 1999).
Online communication in itself is a multidimensional concept. We focus on two
types of communication that are often used by adolescents: IM and chat in public
chatrooms (Lenhart, Madden, & Hitlin, 2005; Valkenburg & Peter, 2007). Both types
of online communication are synchronous and often used for private communica-
tion. However, they differ in several respects. First, whereas chat in a public chat-
room is often based on anonymous communication between unacquainted partners,
1172Journal of Computer-Mediated Communication 12 (2007) 1169–1182 ª2007 International Communication Association
IM mostly involves non-anonymous communication between acquainted partners
(Bryant et al., 2006; Valkenburg & Peter, 2007). Second, whereas chat is more often
used to form relationships, IM is typically used to maintain relationships (Grinter &
Palen, 2004). Although there is no previous research on the social consequences of
IM versus chat, it is entirely possible that these two types of online communication
differ in their potential to influence the quality of existing friendships and well-being.
For example, IM, as used to maintain friendships, may contribute positively to the
quality of existing relationships and well-being, whereas chat in a public chatroom
may have the opposite or no effect. The second aim of our study is to investigate the
differential effects of IM versus chat on well-being and the two mediating variables.
Because previous research does not allow us to formulate a hypothesis regarding
these differential effects of different types of online communication, our research
question asks:
RQ1: How do the causal predictions of the displacement and stimulation hypothesis differ for
IM and chat in a public chatroom?
Method
Sample
In December 2005, an online survey was conducted among 1,210 Dutch adolescents
between 10 and 17 years of age (53% girls, 47% boys). Sampling and fieldwork were
done by Qrius, a market research company in Amsterdam, the Netherlands.
Respondents were recruited from an existing online panel managed by Qrius. The
sample was representative of Dutch children and adolescents who use the Internet in
terms of age, gender, and education. Prior to the implementation of the survey,
institutional approval, parental consent, and adolescents’ informed consent were
obtained. Adolescents were notified that the study would be about Internet and
well-being and that they could stop participation at any time they wished. We took
the following measures to improve the confidentiality, anonymity, and privacy of the
response process (Mustanski, 2001): On the introduction screen of the online ques-
tionnaire, we emphasized that the answers would be analyzed only by us, the principal
investigators. Moreover, we ensured the respondents that their answers would remain
anonymous. Finally, respondents were asked to make sure that they filled in the
questionnaire in privacy. Completing the questionnaire took about 15–20 minutes.
We preferred an online interviewing mode to more traditional modes of inter-
viewing, such as face-to-face or telephone interviews. There is consistent research
evidence that both adolescents and adults report sensitive behaviors more easily in
computer-mediated interviewing modes than in non-computer-mediated modes,
whereas for non-sensitive behaviors no differences in interviewing modes have been
reported (e.g., Beebe, Harrison, McRae, Anderson, & Fulkerson, 1998; Brener et al.,
2006; Tourangeau & Smith, 1996). Therefore, the response patterns in our study may
have benefited from our choice of a computer-mediated interviewing mode as far as
more intimate issues, such as the quality of friendships and well-being, are concerned.
Journal of Computer-Mediated Communication 12 (2007) 1169–1182 ª2007 International Communication Association1173
Measures
IM Use
We measured adolescents’ IM use with four questions: (a) ‘‘On weekdays (Monday to
and including Friday), how many days do you usually use IM?’’ (b) ‘‘On the weekdays
(Monday to and including Friday) that you use IM, how long do you then usually use
it?’’ (c) ‘‘During weekends (Saturday and Sunday), how many days do you usually use
IM?’’ The response options were: (1) Only on Saturday; (2) Only on Sunday; (3) On
both days; and (4) I do not use IM on the weekends. If respondents selected response
options 1 to 3 in the question on IM weekend use, they were asked the following
question for Saturday and/or Sunday: (d) ‘‘On a Saturday (a Sunday), how long do
you usually use IM?’’ Respondents’ IM use per week was calculated by multiplying
the number of days per week that they used IM (range 0 through 7) by the number
of minutes they used it on each day. This operationalization of weekly time spent
with a medium has been proven valid for children older than 9 (Van der Voort &
Vooijs, 1990). The mean time spent with IM per week was 15 hours and 15 minutes
(SD = 21 hours and 10 minutes).
Chat Use
We measured respondents’ chat use in the same way as their IM use. Using the same
four questions, we asked the respondents to evaluate how much time per week they
used chat in public chatrooms. The mean time spent with chat per week was 1 hour
and 23 minutes (SD = 7 hours and 30 minutes).
Time Spent with Friends
Time spent with existing friends was measured with three items that were adopted
from the companionship subscale of Buhrmester’s (1990) Network of Relationship
Inventory. We first asked respondents to think of the friends they know from their
offline environment, such as from school and the neighborhood. Then we asked
them three questions: (a) ‘‘How often do you meet with one or more of these
friends?,’’ (b) ‘‘How often do you and these friends go to places and do things
together?,’’ and (c) ‘‘How often do you go out and have fun with one or more of
these friends?’’ Response options ranged from 1 (never)to9(several times a day). The
three items loaded on one factor, which explained 69% of the variance (Cronbach’s
alpha = .76; M= 5.78; SD = 1.65).
Quality of Friendships
The quality of existing friendships was measured with the relationship satisfaction
(three items), approval (three items), and support (three items) subscales of
Buhrmester’s (1990) Network of Relationship Inventory. We asked respondents to
think of the friends they know from their offline environment, such as from school
and the neighborhood. Example items were: (1) ‘‘How often are you happy with your
relationship with these friends?’’ (satisfaction), ‘‘How often do these friends praise
you for the kind of person you are?’’ (approval), and (3) ‘‘How often do you turn to
1174Journal of Computer-Mediated Communication 12 (2007) 1169–1182 ª2007 International Communication Association
these friends for support with personal problems?’’ Response options ranged from 1
(never)to5(always). The nine items were averaged to form a quality of friendship
scale (Cronbach’s alpha = .93; M= 3.44; SD = 0.72).
Well-Being
We used the five-item satisfaction with life scale developed by Diener, Emmons,
Larsen, and Griffin (1985). Examples of items of this scale are ‘‘I am satisfied with my
life’’ and ‘‘In most ways my life is close to my ideal.’’ Response categories ranged from
1(agree entirely)to5(disagree entirely) and were reversely coded. Cronbach’s alpha
for the scale was .88, which is comparable to the alpha of .87 reported by Diener et al.
(1985).
Results
Time Spent with IM and Chat
Respondents spent significantly more time per day on IM than on chat. Specifically,
they spent on average two hours and 11 minutes per day on IM and on average
12 minutes per day on chat. This greater amount of time spent on IM suggests that if
any effect of the Internet is to be expected, it will occur through the use of IM.
However, to verify this claim, we test the separate effects of IM and chat in the
subsequent analyses.
Online Communication with Existing Friends
We also investigated the assumption in this and earlier studies that IM is most often
used to communicate with existing friends, whereas chat in a public chatroom is
more often used to communicate with strangers. This assumption was supported.
Ninety-one percent of the respondents indicated that they ‘‘often’’ to ‘‘always’’ used
IM to communicate with existing friends. Thirty-seven percent of the respondents
indicated that they ‘‘often’’ to ‘‘always’’ used chat to communicate with existing
friends.
Pitting the Displacement Hypothesis against the Stimulation Hypothesis
Following the displacement and stimulation hypothesis, we did not assume a direct
relationship between online communication and well-being. Rather, we expected
that the direct relationship between online communication would be mediated by
the time spent with existing friends and the quality of these friendships (see Fig-
ure 1). Table 1 presents the zero-order correlations between the independent vari-
ables (IM and chat use), the first mediating variable (time spent with friends), the
second mediating variable (quality of friendships), and the dependent variable (well-
being). In line with our expectations, neither IM nor chat use was directly related to
well-being. However, the results in Table 1 do suggest a mediated positive effect of
IM use and, to a lesser extent, a positive mediated effect of chat use on well-being
through the time spent with friends and the quality of friendships.
Journal of Computer-Mediated Communication 12 (2007) 1169–1182 ª2007 International Communication Association1175
We used a formal mediation analysis to test our hypotheses. In recent years,
several approaches to examining indirect or mediated effects have been discussed
(for a review, see MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002). The most
widely used approach is the causal steps approach developed by Judd and Kenny
(1981) and Baron and Kenny (1986). This approach assumes that in order to test
mediation, the independent, dependent, and mediator (or intervening) variables
must all be correlated with each other. The causal steps approach has recently been
criticized, first because it does not provide a statistical test of the size of the indirect
effects, and second because the requirement that there must be a significant direct
association between the independent and dependent variable is considered too
restrictive (MacKinnon, Krull, & Lockwood, 2000; MacKinnon, Lockwood, et al.,
2002; Shrout & Bolger, 2002).
The problems inherent in the causal steps approach are solved in the intervening
variable approach proposed by MacKinnon and his colleagues (MacKinnon, Krull,
et al., 2000; MacKinnon, Lockwood, et al., 2002), which was used in the present
study. The first step in this approach is to run a regression analysis with the inde-
pendent variable predicting the mediator. The second step is to estimate the effect of
the mediator on the dependent variable, after controlling for the independent vari-
able. However, because we hypothesized that two (rather than one) intervening
variables would mediate the effect of online communication on well-being, we used
a four-step procedure to test for mediation.
In the first step, the independent variable (online communication) predicted the
first intervening variable (time spent with friends). In the second step, the first
intervening variable (time spent with friends) predicted the second intervening vari-
able (quality of friendships), while controlling for the independent variable (online
communication). In the third step, the first intervening variable (time spent with
friends) predicted the second intervening variable (quality of friendships), and in the
fourth and final step, the second intervening variable (quality of friendships) pre-
dicted well-being, while controlling for the first intervening variable (time spent with
friends). The results of these four regression analyses are presented in Table 2.
Table 1 Zero-order correlations among all variables in the study
Variables Time spent
with IM
Time spent
with chat
Time spent
with friends
Quality
of friendships
Well-being
Time spent with chat 33***
Time spent with friends .16*** .07*
Quality of friendships 2.01 2.07* .30***
Well-being 2.03 .01 .18*** .20***
Age .20*** .03 2.03 .06* 2.12*
Gender (male) .05 .02 .00 2.17*** .05
Note: *p,.05; **p,.01; p,.001
1176Journal of Computer-Mediated Communication 12 (2007) 1169–1182 ª2007 International Communication Association
As the first mediation analysis in Table 2 shows, time spent with IM was
positively related (b= .15, p,.001) to the time spent with existing friends,
a result which supports the stimulation hypothesis and our H1b (see Figure 1).
The opposite displacement hypothesis expressedinH1a,whichpredictedanega-
tive path between these two variables, was not supported. As Table 1 shows, the
regression analysis showed that time spent with chat was not significantly related
to time spent with friends (b= .02, n.s.). This implies that the first condition for
mediation was not met in the case of time spent with chat. In other words, the
causal predictions of the two hypotheses (H1a and H1b) only applied to IM, but
not to chat. Therefore, the subsequent mediation analyses were only conducted for
time spent with IM.
Our second hypothesis stated that the quality of friendships would positively
predict well-being and act as mediator between time spent with friends and well-
being (path 3 in Figure 1). This hypothesis was supported. As the second mediation
analysis in Table 2 shows, the quality of friendships significantly predicted well-being
(b= .16, p,.001), even when the first mediating variable (time spent with friends)
was controlled. The fact that time spent with friends remained a significant predictor
(b= .13, p,.001) of well-being when the quality of friendship was controlled
indicates that the mediation of quality of friendship was only partial. Finally, in
support of our third hypothesis (path 2 in Figure 1), time spent with friends acted
as a full mediator between time spent with IM and the quality of friendships (see the
significant bof .32 for time spent with friends versus the nonsignificant bof 2.05 for
time spent with IM).
Table 2 Mediation analyses
BSEb
First mediation analysis
DV: Time spent with friends
IV: IM frequency .0014 .0003 .15*
IV: chat frequency .0005 .0008 .02
DV: Quality of friendships
IV: IM frequency 2.0002 .0001 2.05
MV: Time spent with friends .1376 .0121 .32*
Second mediation analysis
DV: Quality of friendships
IV: Time spent with friends .1339 .0119 .30*
DV: Well-being
IV: Time spent with friends .0646 .0146 .13*
MV: Quality of friendships .1858 .0335 .16*
Notes: DV = Dependent variable; IV = Independent variable; MV = Mediating variable
*p,.001; The 4 decimals for the Bs are presented to enable the reader to recalculate the
Sobel test.
Journal of Computer-Mediated Communication 12 (2007) 1169–1182 ª2007 International Communication Association1177
We tested the significance of the indirect effects by means of a formula developed
by Sobel (1982). If the Sobel test leads to the critical z-value of 1.96, the mediator
carries the influence of the independent variable to the dependent variable. An online
version of this test is available at http://www.unc.edu/~preacher/sobel/sobel.htm
(Preacher & Leonardelli, 2005). The z-value for the first mediation analysis was
5.03, p= .001; the z-value for the second mediation analysis was 4.98, p=.001.
These significant z-values indicate that both the time spent with friends and the
quality of friendships are valid underlying mechanisms through which the effect of
IM on well-being can be explained.
Discussion
The aim of this study was to test the validity of two opposing explanatory hypotheses
on the effect of online communication on well-being: the displacement hypothesis
and the stimulation hypothesis. Both hypotheses assume that online communication
affects adolescents’ well-being through its influence on their time spent with existing
friends and the quality of those friendships. However, the displacement hypothesis
assumes a negative effect from online communication to time spent with existing
friends, whereas the stimulation hypothesis predicts a positive relationship between
these variables.
We used formal mediation analyses to test the validity of the two mediating
variables. Our results were more in line with the stimulation hypothesis than with
the displacement hypothesis. We found that time spent with IM was positively
related to the time spent with existing friends. In addition, the quality of friendships
positively predicted well-being and acted as a first mediator between time spent with
IM and well-being. Finally, we found that time spent with friends mediated the effect
of time spent with IM on the quality of friendships.
However, the positive effects of our study held only for the time spent with IM
and not for time spent with chat in a public chatroom. IM and chat seem to have very
different functions for adolescents. In line with earlier studies, we found that the
majority of adolescents use IM to talk with their existing friends. Chat in a public
chatroom is less often used by adolescents. However, when utilized, adolescents
primarily seem to chat with strangers. It is important for future research to differ-
entiate between the uses of online communication technologies, because there is
a risk of finding misleading null-effects when these different uses are unknowingly
combined in a survey.
Overall, our study suggests that Internet communication is positively related to
the time spent with friends and the quality of existing adolescent friendships, and,
via this route, to their well-being. These positive effects may be attributed to two
important structural characteristics of online communication: its controllability
and its reduced cues. Several studies have shown that these characteristics of online
communication may encourage intimate self-disclosure (e.g., Joinson, 2001; Leung,
2002; McKenna, Green, & Gleason, 2002; Tidwell & Walther, 2002; Valkenburg &
1178Journal of Computer-Mediated Communication 12 (2007) 1169–1182 ª2007 International Communication Association
Peter, 2007), especially when adolescents perceive these characteristics of Internet
communication as important (Schouten, Valkenburg, & Peter, in press; Valkenburg
& Peter, 2007). Because intimate self-disclosure is an important predictor of reciprocal
liking, caring, and trust (Collins & Miller, 1994), Internet-enhanced intimate self-
disclosure may be responsible for a potential increase in the quality of adolescents’
friendships.
Our results have several implications for future research. Because the stimulation
hypothesis seems to be the best working hypothesis, follow-up research can be
specifically designed to explore this hypothesis further and pursue a next step in
theory formation: Attempting to answer why online communication may stimulate
the quality of existing friendships. Although there is growing evidence for the posi-
tive effect of online communication on intimate self-disclosure, to date no research
has demonstrated whether this potential mediator may account for a stimulation
effect on the quality of existing friendships.
Although this study pitted two causal effects hypotheses against one another, we
acknowledge that the assumptions of these hypotheses were tested with cross-
sectional data. Although our study was theory driven, a reverse explanation for
our findings may also be plausible. That is, how people choose to use online com-
munication may be influenced by the quality of their existing relationships and/or by
their trait sociability (see also Baym, Zhang, & Lin, 2004). There is a pressing need for
causal-correlational research to investigate the longitudinal relationships between
online communication and the quality of adolescent existing relationships. Not only
are longitudinal designs better able to distinguish causation from covariance, but
they are also pre-eminently suitable for exploring the validity of the underlying
mechanisms by which Internet communication influences adolescents’ social
relationships.
Acknowledgment
We would like to thank the Netherlands Organisation for Scientific Research [NWO]
for providing support for this study.
References
Baron, R., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology,51(6), 1173–1182.
Baym, N. K., Zhang, Y. B., & Lin, M. C. (2004). Social interactions across media: Interpersonal
communication on the Internet, telephone and face-to-face. New Media & Society,6(3),
299–318.
Beebe, T. J., Harrison, P. A., McRae, J. A., Anderson, R. E., & Fulkerson, J. A. (1998).
An evaluation of computer-assisted self-interviews in a school setting. Public Opinion
Quarterly,62(4), 623–632.
Journal of Computer-Mediated Communication 12 (2007) 1169–1182 ª2007 International Communication Association1179
Brener, N. D., Eaton, D. K., Kann, L., Grunbaum, J. A., Gross, L. A., Kyle, T. M, et al. (2006).
The association of survey setting and mode with self-reported health risk behaviors
among high school students. Public Opinion Quarterly,70(3), 354–374.
Bryant, J. A., Sanders-Jackson, A., & Smallwood, A. M. K. (2006). IMing, text messaging,
and adolescent social networks. Journal of Computer-Mediated Communication,11(2),
article 10. Retrieved July 2, 2007 from http://jcmc.indiana.edu/vol11/issue2/bryant.html
Buhrmester, D. (1990). Intimacy of friendship, interpersonal competence, and adjustment
during preadolescence and adolescence. Child Development,61(4), 1101–1111.
Bukowski, W. M. (2001). Friendships and the worlds of childhood. New Directions for Child
and Adolescent Development,91, 93–105.
Collins, N. L., & Miller, L. C. (1994). Self-disclosure and liking: A meta-analytic review.
Psychological Bulletin,116, 457–475.
Diener, E. (1984). Subjective well-being. Psychological Bulletin,95(3), 542–575.
Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction with life scale.
Journal of Personality Assessment,49, 71–75.
Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being: Three
decades of progress. Psychological Bulletin,125(2), 276–302.
Grinter, R. E., & Palen, L. (2004). Instant messaging in teen life. Proceedings of the 2002 ACM
Conference on Computer Supported Cooperative Work (CSCW ’02). Retrieved July 5, 2004
from http://doi.acm.org/10.1145/587078.587082
Gross, E. F. (2004). Adolescent Internet use: What we expect, what teens report. Journal of
Applied Developmental Psychology,25(6), 633–649.
Hartup, W. W. (2000). The company they keep: Friendships and their developmental
significance. In W. Craig (Ed.), Childhood Social Development: The Essential Readings
(pp. 61–84). Malden, MA: Blackwell.
Hartup, W. W., & Stevens, N. (1997). Friendships and adaptation in the life course.
Psychological Bulletin,121(3), 355–370.
Jackson, L. A., von Eye, A., Barbatsis, G., Biocca, F., Fitzgerald, H. E., & Zhao, Y. (2004). The
impact of Internet use on the other side of the digital divide. Communications of the ACM,
47(7), 43–47.
Joinson, A. N. (2001). Self-disclosure in computer-mediated communication: The
role of self-awareness and visual anonimity. European Journal of Social Psychology,
31, 177–192.
Judd, C. M., & Kenny, D. A. (1981). Process analysis: Estimating mediation in treatment
evaluations. Evaluation Review,5(5), 602–619.
Jung, J. Y., Qiu, L. J., & Kim, Y.-C. (2001). Internet connectedness and inequality.
Communication Research,28(4), 507–535.
Kraut, R., Kiesler, S., Boneva, B., Cummings, J., Helgeson, V., & Crawford, A. (2002). Internet
paradox revisited. Journal of Social Issues,58(1), 49–74.
Kraut, R., Patterson, M., Lundmark, V., Kiesler, S., Mukopadhyay, T., & Scherlis, W. (1998).
Internet paradox: A social technology that reduces social involvement and psychological
well being? American Psychologist,53(9), 1017–1031.
LaRose, S., Ghuay, F., & Boivin, M. (2002). Attachment, social support, and loneliness in
young adulthood: A test of two models. Personality and Social Pscyhology Bulletin,28(5),
684–693.
1180Journal of Computer-Mediated Communication 12 (2007) 1169–1182 ª2007 International Communication Association
Lenhart, A., Madden, M., & Hitlin, P. (2005). Teens and Technology. Washington, DC:
Pew Internet & American Life Project.
Leung, L. (2002). Loneliness, self-disclosure and ICQ (‘‘I Seek You’’) use. CyberPsychology &
Behavior,5(3), 241–251.
MacKinnon, D. P., Krull, J. L., & Lockwood, C. M. (2000). Equivalence of the mediation,
confounding, and suppression effect. Prevention Science,1(4), 173–181.
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002).
A comparison of methods to test mediation and other intervening variable effects.
Psychological Methods,7(1), 83–104.
Matei, S., & Ball-Rokeach, S. J. (2001). Real and virtual social ties: Connections in the
everyday lives of seven ethnic neighborhoods. American Behavioral Scientist,45(3),
550–564.
McKenna, K. Y., Green, A. S., & Gleason, M. E. (2002). Relationship formation on the
internet: What’s the big attraction? Journal of Social Issues,58(1), 9–31.
Mesch, G. (2001). Social relationships and Internet use among adolescents in Israel. Social
Science Quarterly,82(2), 329–340.
Mesch, G. (2003). The family and the Internet: The Israeli case. Social Science Quarterly,84(4),
1083–1050.
Morgan, C., & Cotten, S. R. (2003). The relationship between Internet activities and
depressive symptoms in a sample of college freshmen. CyberPsychology & Behavior,
6(2), 133–141.
Mustanski, B. S. (2001). Getting wired: Exploiting the Internet for the collection of sexually
valid data. The Journal of Sex Research,38(4), 292–301.
Nie, N. H. (2001). Sociability, interpersonal relations and the Internet: Reconciling conflicting
findings. American Behavioral Scientist,45(3), 420–435.
Nie, N. H., Hillygus, D. S., & Erbring, L. (2002). Internet use, interpersonal relations, and
sociability: A time diary study. In B. Wellman & C. Haythornthwaite (Eds.), The Internet
in Everyday Life (pp. 215–243). Oxford: Blackwell.
Preacher, K. J., & Leonardelli, G. J. (2005). Calculation for the Sobel test: An interactive
calculation tool for mediation tests. Retrieved March 1, 2006 from http://www.unc.edu/
~preacher/sobel/sobel.htm
Sanders, C. E., Field, T. M., Diego, M., & Kaplan, M. (2000). The relationship of Internet use
to depression and social isolation. Adolescence,35(138), 237–242.
Schouten, A. P., Valkenburg, P. M., & Peter, J. (in press). Precursors and underlying processes
of adolescents’ online self-disclosure: Developing and testing an ‘‘Internet-attribute-perception’’
model. Media Psychology.
Shaw, L. H., & Gant, L. M. (2002). In defense of the Internet: The relationship between
Internet communication and depression, loneliness, self-esteem, and perceived social
support. CyberPsychology & Behavior,5(2), 157–170.
Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonexperimental studies:
New procedures and recommendations. Psychological Methods,7(4), 422–445.
Sobel, M. E. (1982). Asymptotic intervals for indirect effects in structural equations models.
In S. Leinhart (Ed.), Sociological Methodology (pp. 290–312). San Francisco: Jossey-Bass.
Subrahmanyam, K., Kraut, R. E., Greenfield, P. M., & Gross, E. (2000). The impact of home
computer use on children’s activities and development. Future of Children,10(2),
123–144.
Journal of Computer-Mediated Communication 12 (2007) 1169–1182 ª2007 International Communication Association1181
Tidwell, L. C., & Walther, J. B. (2002). Computer-mediated communication effects on
disclosure, impressions, and interpersonal evaluations. Human Communication Research,
28(3), 317–348.
Tourangeau, R., & Smith, T. W. (1996). Asking sensitive questions: The impact of data
collection mode, question format, and question context. Public Opinion Quarterly,60(2),
275–304.
Valkenburg, P. M., & Peter, J. (2007). Preadolescents’ and adolescents’ online communication
and their closeness to friends. Developmental Psychology,43(2).
Valkenburg, P. M., & Soeters, K. (2001). Children’s positive and negative experiences with the
Internet. Communication Research,28(5), 653–676.
Van der Voort, T. H., & Vooijs, M. W. (1990). Validity of children’s direct estimates of time
spent television viewing. Journal of Broadcasting and Electronic Media,34(1), 93–99.
Waestlund, E., Norlander, T., & Archer, T. (2001). Internet blues revisited: Replication and
extension of an Internet paradox study. CyberPsychology & Behavior,4(3), 385–391.
Weiser, E. B. (2001). The functions of Internet use and their psychological consequences.
Cyberpsychology & Behavior,4(6), 723–744.
About the Authors
Patti M. Valkenburg is a professor in the Amsterdam School of Communications
Research (ASCoR) at the University of Amsterdam. Her research interests include
the effects of media on children and adolescents. She is chair of CAM, the research
center of Children, Adolescents, and the Media; see http://www.cam-ascor.nl
Address: Amsterdam School of Communication Research, ASCoR, University of
Amsterdam, Kloveniersburgwal 48, 1012 CX Amsterdam, The Netherlands
Jochen Peter is an associate professor in the Amsterdam School of Communications
Research (ASCoR) at the University of Amsterdam. His research focuses on the
consequences of adolescents’ Internet use for their sexual socialization and psycho-
social development; see http://www.cam-ascor.nl
Address: Amsterdam School of Communication Research, ASCoR, University of
Amsterdam, Kloveniersburgwal 48, 1012 CX Amsterdam. Kloveniersburgwal 48,
1012 CX Amsterdam, The Netherlands
1182Journal of Computer-Mediated Communication 12 (2007) 1169–1182 ª2007 International Communication Association
