Computers in Human Behavior 26 (2010) 1289–1295

Contents lists available at ScienceDirect

Computers in Human Behavior

journal homepage: www.elsevier.com/locate/comphumbeh

Social network use and personality

Yair Amichai-Hamburger a,*, Gideon Vinitzky b
a
The Research Center for Internet Psychology (CIP), Sammy Ofer School of Communication, Interdisciplinary Center (IDC), Herzliya, Israel
b
Department of Economics and Business Management, Ariel University Center of Samaria, Israel

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

Article history: Studies have shown a connection between the individual personality of the user and the way he or she
Available online 6 May 2010 behaves on line. Today many millions of people around the world are connected by being members of
various Internet social networks. Ross et al. (2009) studied the connection between the personality of
Keywords: the individual users and their behavior on a social network. They based their study on the self-reports
Internet and personality of users of Facebook, one of the most popular social networks, and measured five personality factors
Facebook using the NEO-PI-R (Costa & McCrae, 1992) questionnaire. They found that while there was a connection
Social network
between the personalities of surfers and their behavior on Facebook, it was not strong. This study is based
on that of Ross et al. (2009), but in our study the self-reports of subjects, were replaced by more objective
criteria, measurements of the user-information upload on Facebook. A strong connection was found
between personality and Facebook behavior. Implications of the results are discussed.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction 1.1. Impression management in social networks

The Internet is a worldwide decentralized network of comput- Impression management (IM) refers to the attempt to control
ers that today has a far-reaching influence and may affect almost information in order to affect others’ opinions of us (Goffman,
all aspects of our existence. In fact, the Internet has integrated so 1959). As Shakespeare eloquently explained: ‘‘All the world is a
well into people’s lives that, for many, it is very difficult to imagine stage, and all the men and women merely players” (1564–1616
how they lived without it. One aspect of our daily existence reprinted 1978, p. 287). Erving Goffman (1959), it seems, would
where the Internet has introduced major changes is our social agree. He writes that the actor, shaped by the environment and tar-
lives (Amichai-Hamburger, Wainapel, & Fox, 2002; Hamburger & get audience, sees the interaction as a performance. The objective
Ben-Artzi, 2000). The Internet is a leading social arena where of the performance is to provide the audience with an impression
people can meet others and interact. Social life on the Internet consistent with the desired goals of the actor. Boyd (2007) sug-
initially comprised social tools, such as chat forums and newsgroups. gested that in the social network arena it seems that the individual
Today it has developed many additional components, such as blogs, ‘‘profile can be seen as a form of digital body where individuals
fantasy environments, and social networks (Amichai-Hamburger, must write themselves into being” (p. 131). Sundén (2003) sug-
2005; Amichai-Hamburger & Barak, 2009). gested that people frequently reproduce themselves on the Inter-
One of the fastest growing and most popular of such sites is the net so that others may see them and communicate with them.
social network. It is defined as a web-based service that allows While social networks request that users construct truthful rep-
individuals to: (1) construct a profile within an organized frame- resentations of themselves, they actually do so with varying de-
work; (2) generate a list of other users with whom they share a grees of accuracy. In fact, Marwick (2005) who studied such
connection; and (3) navigate their own list of connections and view users, basing his findings on three different social networking sites,
those made by others within the system (Boyd & Ellison, 2007). reported that participants had developed complex strategies for
Typically, the social network is utilized to sustain existing offline negotiating the rigidity of the prescribed ‘‘authentic” profile. In
relationships or support offline connections, as opposed to meeting other words, people on Facebook and other social networks do
new people. These relationships may be based on frail ties, but typ- not so much as lie, but rather stretch the truth (sometimes to its
ically there is some common offline connection among individuals outer limits).
(Ellison, Steinfield, & Lampe, 2007). One of the IM tools available on social networks is the accep-
tance, or otherwise, by users of other users wishing to be their
‘‘friends”. Their choice of whom to accept (or reject) reflects a sig-
* Corresponding author. Tel.: +972 9 9527649; fax: +972 9 9527650. nificant part of their ongoing effort to project the ‘‘right” image to
E-mail address: yairah@idc.ac.il (Y. Amichai-Hamburger). the world (Donath & Boyd, 2004). As befits the slogan ‘‘tell me who

0747-5632/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.chb.2010.03.018
1290 Y. Amichai-Hamburger, G. Vinitzky / Computers in Human Behavior 26 (2010) 1289–1295

your friends are and I will tell you who you are,” people try to have ies of five dimensional traits (Costa & McCrae, 1992). The first trait,
the right list of friends in order to create the desired image of neuroticism, reflects a person’s tendency to experience psycholog-
themselves. Zinman and Donath (2007) pointed out that this ten- ical distress and high levels of this trait are associated with a sen-
dency sometimes causes people to loosen their criteria as to whom sitivity to danger. Extraversion, the second trait, reflects a tendency
to accept as their ‘‘friend”. to be sociable and able to experience positive emotions. The third
On social networks, as part of their profiles, users frequently factor, openness to experience, represents an individual’s willing-
present varying numbers of photographs. This display is in addi- ness to consider alternative approaches, be intellectually curious,
tion to their cover picture. Zhao, Grasmuck, and Martin (2008) and enjoy artistic pursuits. Agreeableness, the fourth factor, is an-
examined this phenomenon on Facebook, one of the most popular other aspect of interpersonal behavior, reflecting a tendency to be
social networking sites, and found that, on average, Facebook sites trusting, sympathetic, and cooperative. The fifth dimension, con-
display around 88 photos per user. The number of users who dis- scientiousness, reflects the degree to which an individual is orga-
played their profile photos and wall posts was above 90%. While nized, diligent, and scrupulous.
some created walls of privacy for themselves and their friends, The FFM has been applied in a number of recent studies assess-
most users were happy to put on public exhibition a broad range ing the Internet environment. For example, Guadagno, Okdie, and
of photos, including those sent to them by others, mostly of them- Eno (2008) employed it in their study of blogs. They found that
selves in the context of their friends and mostly showing happiness people who are high in openness and high in neuroticism are likely
and enjoyment. It seems clear that the photos displayed on the to be bloggers. Additionally, the neuroticism relationship was
individual profiles constitute an important way to project the im- moderated by gender indicating that women who are high in
age we wish to present to others. neuroticism are more likely to be bloggers as compared with those
low in neuroticism, whereas no differences were found for men.
1.2. Personality and social networks These results indicate that personality factors impact on the likeli-
hood of being a blogger and have implications for understanding
Hamburger and Ben-Artzi (2000) suggested that Internet use is who blogs. Amichai-Hamburger et al. (2008) also employed the
related to personality. They demonstrated that on the Internet ‘‘the Five-Factor Model when they assessed the personality profile of
poor can get richer,” namely, that introverts can compensate them- Wikipedia contributors. Their results revealed significant differ-
selves for the difficulties they experience in offline social interac- ences between Wikipedia members and non-Wikipedia members
tions (see also Amichai-Hamburger et al., 2002). According to with regard to the factors of agreeableness, openness, and consci-
Amichai-Hamburger (2002), personality is a leading factor in entiousness. These were found to be lower for the Wikipedia
understanding why people behave the way they do on the Internet. members as compared with non-Wikipedia members. Ross et al.
Since the net, by its very nature, is powered by human interaction, (2009) suggested that Facebook use is related to personality. Using
it follows that we cannot understand the workings of the Internet a Five-Factor Model personality questionnaire, they examined
without understanding the personalities of those who surf it (see behavior on Facebook as reported by users. Their first five predic-
also Amichai-Hamburger, 2005). This link between personality tions pertained to the relationship between behavior and the user’s
and Internet use has been demonstrated using a number of differ- personality. They predicted that:
ent personality theories, among them those of extroversion and
neuroticism (Hamburger & Ben-Artzi, 2000); need for cognition 1. Due to their greater tendency to be sociable, individuals who
(Amichai-Hamburger, Kinar, & Fine, 2007; Kinar & Amichai-Ham- scored higher on the trait of Extraversion would (a) demon-
burger, 2008); need for closure (Amichai-Hamburger, Fine, & Gold- strate more frequent use of Facebook; (b) make greater use of
stein, 2004) and sensation seeking (Lin & Tsai, 2002). Facebook components for communication: (c) have more ‘‘Face-
The relationship between personality and social networking book friends”; and (d) belong to more Facebook groups.
was previously noted in a study of nostalgic websites carried out 2. It was predicted that individuals who scored higher on the trait
by Amichai-Hamburger, Kaplan, and Dorpatcheon (2008). They of Neuroticism would be more willing to share personally-
found that extroverted participants who frequently surfed nostal- identifying information on Facebook, spend more time on
gic websites, made greater use of the social services found on the Facebook, and be less likely to use private messages, since they
net, as compared with introverted subjects who used a lot nostal- would be seeking to receive social support through Facebook.
gic websites. However, introverts who did not use nostalgic web- 3. People with an ability to engage in caring and meaningful inter-
sites were found to use the social services on the net more than personal offline relationships and who scored higher on the
those extroverts who did not use the nostalgic websites. Amic- trait of agreeableness were expected to have greater numbers
hai-Hamburger and his colleagues suggested that when surfers’ of ‘‘Facebook friends” on their profile.
foremost behavior on the net is surfing nostalgic websites where 4. Those with a tendency to be curious and desirous of exploring
they interact with former offline friends, their pattern of social new activities, who scored higher on the trait of openness to
interaction offline is transferred to their behavior online. In this experience, were expected to be more willing to use Facebook
way, extroverts retain their offline social dominance when they as a communication tool and to use a greater number of compo-
are online. This is consistent with the rich-get-richer theory (Kraut nents, resulting in greater knowledge of Facebook features.
et al., 1998). Conversely, among people who do not surf nostalgic 5. Individuals who scored higher on the trait of Conscientiousness
websites and whose Internet behavior is more explorative, intro- and who placed great importance on fulfilling their obligations
verts are more likely to use the Internet as a compensative envi- and meeting deadlines were expected to demonstrate a more
ronment and this, in turn, may lead them to become more limited use of Facebook activities.
socially dominant on the net as compared with extroverts. This
is consistent with the-poor-get-richer theory (Hamburger & Ben- The results reported by Ross et al. (2009) gave partial support to
Artzi, 2000). It seems logical to assume that the behavior observed the link between individual personality and Facebook behavior.
among users of Facebook, i.e., the transference of old friends from The researchers did not find any significant connection between
their offline world to online, demonstrates the dominance of Facebook behavior and the personality factors of agreeableness
extroverts. and openness (Hypotheses 3 and 4). However, they did find a par-
The Five-Factor Model (FFM) is a broad classification of person- tial link between behavior on Facebook and the traits of extrover-
ality traits. The model separates the human personality into a ser- sion, neuroticism, and conscientiousness.
 Y. Amichai-Hamburger, G. Vinitzky / Computers in Human Behavior 26 (2010) 1289–1295 1291

It is important to point out that, although there was a strong ba- ered virtually all major items that can be found in a Facebook ac-
sis for Ross et al.’s (2009) predictions, we suggest that a more effec- count (see Appendix 1). Using this coding scheme the contents of
tive research method would be to examine the way people build the resulting Facebook accounts were numerically recorded by a
their profile on Facebook instead of relying on self-report question- specially trained graduate research assistant. In addition, the sum
naires. This would have provided a more objective criterion than of all information uploads regarding each dimension of the Face-
asking users about their individual behavior on Facebook, since book profile (i.e., basic information, personal information, contact
these self-reports are likely to be influenced by social desirability. information, education and work) was calculated. However, since
We would, therefore, suggest replicating Ross et al.’s assessments only four participants provided information regarding contact
using the Facebook profiles which are less prone to bias. information, this dimension was removed from the analysis.
We concurred with all of Ross et al.’s (2009) predictions, except
their final one. In the case of prediction number 5, in contrast to 3. Results
Ross et al.’s prediction, we suggest that, in fact, individuals who
scored higher on the trait of conscientiousness will reveal a larger In order to investigate personality features, groups were created
number of friends. This is because conscientiousness involves by dividing each personality domain into equal thirds. The NEO-PIR
high-target orientation which, in the context of the Facebook envi- domain scores cutoffs for each personality domain are listed in
ronment, involves striving for a greater number of friends. Table 1. Only the upper and lower thirds were analyzed and all
scores were within three standard deviations from the mean. Such
2. Method a test allows for greater statistical sensitivity in searching for group
differences, especially given our sample size (see Ross et al., 2009).
2.1. Participants Hypothesis 1 suggests that individuals in the highly extroverted
group will demonstrate a higher number of friends and would be-
Data was collected from 237 students at an Israeli university. All long to more Facebook groups than individuals in the least extro-
were undergraduate students enrolled in the Department of Eco- verted group. In order to examine the effect of extroverted
nomics and Business Management. All participants received credit personality on Facebook use, two analysis of covariance (ANCOVA)
in their introductory courses for participating in the study. Partic- were conducted. For both analyses the independent variables were
ipants comprised 101 males and 136 females with an average age extroverted personality while for the first analysis the number of
of 22. Research suggests that university students form suitable friends was entered as a dependent variable and for the second
samples for studies involving the Internet because they tend to analysis the extent of participation in Facebook groups was en-
be frequent users, including regular use of a variety of CMC func- tered as a dependent variable.
tions (Pornsakulvanich, Haridakis, & Rubin, 2008). In Israel about In both tests gender was entered as a covariant. The data indi-
two million people use Facebook which is about third of the pop- cated that individuals in the highly extroverted group have a sig-
ulation. The assessments typically lasted 30–45 min. nificantly higher number of friends (Mean = 150.96, SD = 116.39)
than individuals in the least extroverted group (Mean = 103.18,
SD = 85.71), F(1,134) = 6.23, p < 0.001. However no significant dif-
2.2. Procedure and measurements ference was found with regard to participation in Facebook groups.
Contrary to the intuitive assumption, individuals in the highly
The study consisted of two sequential phases held over a two extroverted group demonstrated a lower use of personal informa-
month period. All participants were requested to complete the tion (Mean = 0.14, SD = 0.38) than individuals in the least extro-
NEO-PI-R, a self-report measure, in order to assess personality verted group (Mean = 0.47, SD = 1.39), (F(2,150) = 2.72, p = 0.1).
along the domains of the Five-Factor Model. The NEO-PI-R (Costa This finding partially supported hypothesis 1.
& McCrae, 1992) is a widely used self-report measure of the Five- Hypothesis 2 suggests that those with a higher level of neurot-
Factor Model of personality. It is broken down into five domains, icism would be more willing to share personally-identifying infor-
or factors, each of which has six facet scores (each facet comprising mation on Facebook, and be less likely to use private messages.
eight items). The NEO-PI-R contains 240 items which are scored on Information upload regarding to personally-identifying informa-
a five-point Likert scale from ‘‘strongly agree” 1 to ‘‘strongly dis- tion and its relevant fields, such as mailing address or phone num-
agree” 5. Reliability for the five factors is acceptable (from a = .78 ber, was found to be very rare. We therefore examined this
to a = .91). Hypothesis by observing the use of self-pictures as an indicator
In the next phases, user-information upload on Facebook was of personally-identifying information sharing and the feature of
measured and encoded. Based on the Facebook profile setup, four posting other pictures as an indicator of willingness to upload pri-
dimensions of Facebook users were created, namely: Basic informa- vate information. Specifically, two chi square tests were conducted.
tion, personal information, contact information and education, and In the first a correlation between neurotic personalities and post-
work information. Basic information included information which ing the picture of the person holding the account was analyzed
users upload to Facebook concerning gender, birth dates, home- and in the second the correlation between neurotic personalities
town, home neighborhood, family members, family status (single, and use of the feature of posting other pictures was analyzed.
engaged, married, etc.), the type of relationship the user is seeking The results supported the hypothesis. Individuals in the highly
(friendship, dating, random play, etc.), political views, and religious
views. Personal information included activities, interests, favorite
music, favorite TV shows, favorite movies, favorite books, favorite Table 1
quotations, and about me. Contact information included email, IM Cutoff for NEO-PI-R domain scores measure.

(Immediate message) screen name, mobile phone, land phone, ad- Low cutoff High cutoff
dress, city/town, neighborhood, zip code, and website. Education NEO-PI-R
and work information included information about university, focus, Neuroticism 88 104
degree, high school, employer, position description, city/town, and Extroversion 108 126
time period. Openness 108 123
Agreeableness 98 116
The encoding of information upload on Facebook was based on
Conscientiousness 106 129
the method devised by Zhao et al. (2008). The coding scheme cov-
1292 Y. Amichai-Hamburger, G. Vinitzky / Computers in Human Behavior 26 (2010) 1289–1295

neurotic group were found to prefer posting their photos on their features. In order to examine the effect of openness to experience
Facebook profile more than individuals in the least neurotic group on Facebook use, an analysis of covariance (ANCOVA) was con-
v2(1) = 4.30, p < 0.05. Individuals in the highly neurotic group were ducted. The independent variable was personality characterized
less inclined to use the feature of upload other pictures than indi- by openness to experience, and the dependent variable was the
viduals in the least neurotic group v2(1) = 2.42, p < 0.1. number of features used from the personal information section of
In addition an approximate regression analysis revealed a U Facebook. Gender was entered as a covariant. The hypothesis was
shape correlation between Neurotic personality and the amount found to be supported by the data. Specifically, individuals who
of basic information reported on Facebook (F(2,234) = 2.33, scored higher on the trait of openness to experience used more fea-
p < 0.1, Exponential term: t = 1.96, p < 0.05). This result indicates tures from the personal information section (Mean = 0.48,
that people with a low or high level of neuroticism prefer to share SD = 1.51) than individuals who scored lower on the trait of open-
more basic information than people with a moderate level of neu- ness to experience (Mean = 0.22, SD = 1.01) F(2,148) = 2.85,
roticism. This may suggest the conflict encountered by people with p < 0.05).
neurotic personalities with regard to the amount of the informa- Hypothesis 5 suggests that individuals who scored higher on
tion sharing needed. the trait of conscientiousness would demonstrate a higher number
Hypothesis 3 suggests that individuals who scored higher on of friends. In order to examine the effect of conscientiousness on
the trait of agreeableness would have more ‘‘Facebook friends” Facebook use, an analysis of covariance (ANCOVA) was conducted.
added to their profile. In order to examine the effect of agreeable- The independent variable was conscientious-type personality and
ness on Facebook use, an analysis of covariance (ANCOVA) was the dependent variable was the number of friends. Gender was en-
conducted. The independent variable was agreeableness and num- tered as a covariant. The hypothesis was supported by the results.
ber of friends was the dependent variable. Gender was entered as a Individuals who scored higher on the trait of conscientiousness
covariant. The data did not support hypothesis 3. Interestingly, an were found to have a higher number of friends (Mean = 147.8
interaction effect on uploading pictures was found between agree- SD = 112.6) than individuals who scored lower on the trait of con-
able personality and gender. Specifically, to examine the effects of scientiousness (Mean = 112.72 SD = 82.49) F(2,136) = 5.04, p < 0.05.
agreeableness and gender on the number of pictures uploaded to In addition, individuals who scored higher on the trait of conscien-
Facebook, a 2 (high agreeableness vs. low agreeableness)  2 (male tiousness were found to demonstrate less use of the picture upload
vs. female) two-way ANOVA was conducted. Means and SDs for the feature than individuals who scored lower on the trait of conscien-
feature of picture use, according to agreeableness and gender in tiousness v2(1) = 2.75, p < 0.09.
Facebook, are presented in Table 2. A significant interaction was To increase the degree of freedom in our analysis and to better
found between agreeableness and gender, F(1,147) = 4.04, understand how personality influences on Facebook use, a multi-
p < 0.05; (see Fig. 1). A t-test for simple effects revealed that, variable regression were performed with the five personality do-
among females, individuals belonging to the low agreeableness mains as predictors of Facebook use. Specifically, two hierarchical
group have fewer pictures compared with individuals belonging multiple regressions were used to test the extent to which person-
to the high agreeableness group, t(86) = 2.63, p < 0.01. However ality factors moderate the motivations associated with Facebook
among males no effect was found. The analysis revealed no main use. The following predictors were entered in the two given steps:
effect in terms of agreeableness. No differences in the number of (1) covariant (gender); (2) personality factors. The entry order of
pictures uploaded was found between participants with lower the variables permits an examination as to whether the variables
and higher agreeableness, and no main effect was found in terms of interest account for any additional variance in the criterion var-
of gender. No difference was found between males and females iable that is not explained by previously entered predictors. The re-
in the number of pictures uploaded. sults of the hierarchical regression analyses are shown in Table 3.
However, interestingly, individuals who scored higher on the Hierarchical multiple regression results indicated that the number
trait of agreeableness were found to have used less page features of friends on Facebook may only be predicted by extroverted-type
(Mean = 2.15, SD = 7.50) than individuals who scored lower on personality b = .24, t(201) = 2.27, p < .01. Extroverted personality
the trait of agreeableness (Mean = 0.52, SD = 1.39) F(2,111) = 3.95, explained a significant proportion of variance in depression scores
p < 0.05. as well, R2 = .07, F(6,201) = 2.64, p < .01. In addition the amount of
In addition, an approximate regression analysis revealed a U personal information reported by individuals on Facebook was
shape correlation between agreeable-type personality and the found to be predicted by gender, b = .18, t(225) = 2.74, p < .01.,
number of pictures uploaded to Facebook (F(2,234) = 4.65, openness b = .18, t(225) = 2.45, p < .01, and negatively by extrover-
p < 0.01, Exponential term: t = 1.97, p < 0.05). This result indicates sion b = .21, t(225) = 2.48, p < .01. These personality factors
that people with low and high levels of agreeableness are inclined were found to explain a significant proportion of variance in
to upload more pictures than people with a moderate level of depression scores, R2 = .04, F(6,225) = 2.64, p < .01.
agreeableness.
Hypothesis 4 suggests that individuals who scored higher on 4. Discussion
the trait of openness to experience were expected to be more will-
ing to use Facebook as a communication tool and to use a greater Following the study by Ross et al. (2009), the present study
number of features, resulting in greater knowledge of Facebook sought to analyze personality and motivations associated with

Table 2
Means and SDs for use of the picture upload feature, according to agreeableness and gender.

Low/high agreeableness – personality trait Facebook

Male Female Total
SD M SD M SD M
Low agreeableness 55.9 26.13 4.02 1.17 42.33 14.32
High agreeableness 31.3 19.83 54.33 25.4 48.17 23.66
Total 47.8 23.73 43.74 15.76 45.49 19.09
 Y. Amichai-Hamburger, G. Vinitzky / Computers in Human Behavior 26 (2010) 1289–1295 1293

Fig. 1. Means of picture upload among high and low agreeableness in male and female groups.

Table 3
Summary of hierarchical regression analyses for variables predicting number of friends and uploading of personal information.

Number of friends Personal information

D t-Value R2adj (R2cha) FCha b t-Value R2adj (R2cha) FCha
Control variables
Gender 0.00 0.04 0.18* 2.74
D Step 1 0.00 0.04 0.03** 7.52
Main effects
Conscientiousness 0.11 1.32 0.04 0.51
Agreeableness 0.07 0.98 0.05 0.68
Openness 0.08 0.10 0.18* 2.45
Extroversion 0.24* 2.27 0.21* 2.48
Neuroticism 0.10 1.18 0.02 0.23
D Step 2 0.07* 2.86 0.05* 1.97

For number of friends: after step 1 (1205) after step 2 (6200). For number of pictures and personal information: after step 1 (1230) after step 2 (6225).
*
p = 0.05.
**
p = 0.01.

Facebook use. However, unlike the results of Ross et al. which did pared with extroverts. This may be explained by the fact that
not show a strong link between individual personality and Face- extroverts rely on their social skills and so feel less need to pro-
book use, this study demonstrated that such correlation indeed ex- mote themselves.
ists. The main difference between the two studies was in their Hypothesis 2 suggests, as do Ross et al. (2009), that individuals
methodological approach. While Ross et al.’s study relied on self- who scored higher on the trait of neuroticism would be more will-
reports by participants, this study was based on more objective ing to share personally-identifying information on Facebook, and
criteria. less likely to use private messages. We examined this hypothesis
As in the study by Ross et al. (2009), our hypothesis 1 suggests by observing the use of self-pictures as an indicator of person-
that individuals in the highly extroverted group would show a lar- ally-identifying information sharing and the use of posting other
ger number of friends and would belong to more Facebook groups pictures as an indicator of willingness to upload private informa-
than individuals in the least extroverted group. This hypothesis tion. Our results support hypothesis 2. Specifically, individuals in
was partially supported. Specifically, the data demonstrate that the highly neurotic group were found to be more inclined to post
extroversion indeed has a positive effect on the number of friends, their photos on their Facebook profile than individuals in the less
but no effect was found with regard to the use of Facebook groups. neurotic group. In addition, individuals in the highly neurotic
These findings are contrary to those of Ross et al. which suggest group were less inclined to use the picture upload features than
that extroverted personality has an effect on the use of Facebook individuals in the less neurotic group. These findings are unlike
groups, but not on the number of friends. In addition, the current those of Ross et al. (2009), which demonstrated that individuals
findings indicate that a highly extroverted personality may demon- in the highly neurotic group are less inclined to post their photos
strate lower use of personal information than less extroverted per- on their Facebook profile than individuals in the less neurotic
sonalities. It seems that introverts transfer their pattern of group.
intoverted behavior from the offline into the online world. This is The current study revealed a U-shaped correlation between
reflected in the size of their social network which tends to be smal- neurotic personality and the amount of basic information reported
ler than that of the extraverts. Interestingly, however, introverts on Facebook. This result indicates that people with low or high lev-
place more personal information on their facebook profiles as com- els of neuroticism were inclined to share more basic information
1294 Y. Amichai-Hamburger, G. Vinitzky / Computers in Human Behavior 26 (2010) 1289–1295

than people with moderate levels of neuroticism. This may indicate play on Facebook and how to behave there. How far these social
that the same behavior may stem from different motivations. norms impact on behavior on Facebook may well depend on per-
While the emotionally secure individual focuses on self-actualiza- sonality influences. The participants of our study were students
tion and expresses it by sharing more information with others from from a single university, and may be thought of as a fairly cohesive
a secure base, the neurotic person who also strives to share more group and therefore it is likely that norms played a significant role.
information, is motivated to do so by the need for self assurance. To study this in greater depth, it is important that we learn the spe-
We recommend that this be examined in future research. cific norms of a particular group in terms of their Facebook use and
Hypothesis 3 suggests, as in Ross et al. (2009), that individuals its interaction with their personality. This will enable us to gain an
who scored higher on agreeableness would have a greater number understanding of how this effects their behavior on Facebook.
of ‘‘Facebook friends”. However, unlike Ross et al. (2009), our data To further increase our knowledge as to how personality is rel-
demonstrates that agreeableness is related to features of Facebook evant to behavior on Facebook, it is important to study how people
use. Specifically, an interaction between agreeableness and gender design and change their Facebook profiles over a significant period
was found to have an effect on uploading pictures. Individuals who of time. Such a study for example over the period of a year would
scored higher on the trait of agreeableness used fewer page fea- enable us to increase our understanding of the long term interac-
tures than individuals who scored lower on the trait of agreeable- tion between personality and the Facebook dynamic.
ness. In addition, a U-shaped correlation between agreeableness An understanding of the relevance of personality to Facebook
and the number of pictures uploaded to Facebook, as well as a U- behavior may help to explain why some people are heavy users
shaped correlation between agreeableness and contact informa- of a specific components of Facebook than others are not. Future
tion upload to Facebook, were found. The fact that agreeableness studies may also help us to gain an understanding as to why some
does not correlate with having more friends demonstrates, we be- people get addicted to Facebook and others do not use Facebook at
lieve, the complexity of the motives that lie behind behavior on all.
Facebook. It makes sense that the norm number of friends may
have more of an impact on this specific result than agreeableness.
However, it is also important to stress that, although this result 4.2. Implications
runs contrary to our prediction, it nevertheless demonstrates that
personality is relevant to Facebook behavior. The somewhat puz- The results of this study which elucidate the connection be-
zling agreeableness outcomes should be the subject of future re- tween personality and consumer behavior, may have implications
search in this field. for marketing practitioners. Recent, studies in marketing suggest
Hypothesis 4 suggests, as did Ross et al. (2009), that individuals the effectiveness of utilizing social networking for marketing pur-
who scored higher on the trait of openness to experience were ex- poses (Goldenberg, Han, Lehmann, & Hong, 2009; Kratzer & Lettl,
pected to be more willing to use Facebook as a communication tool 2009). Researchers encourage the identification within social net-
and to use a greater number of features. The hypothesis was indeed works of users who are innovators, lead users, opinion leaders, or
supported by the data. Specifically, individuals who scored higher have a large number of friends. Such people should be a primary
on the trait of openness to experience used more features from segment targeted by marketing practitioners. Their object being
the personal information section than individuals who scored low- the creation of positive vibes passed on through such sites concur-
er on the trait of openness to experience. It seems that people who rent with the launch of a new product. However the identification
are more open are more expressive on their Facebook profile. of consumers, who may be considered ‘‘early adopters” of new
In contrast to Ross et al. (2009), our hypothesis 5 suggests that products, may be challenging.
individuals who scored higher on the trait of conscientiousness Studies indicate that people are transferring their offline
would demonstrate a higher number of friends, while Ross et al. activities and interaction with friends to online environments.
demonstrated that conscientiousness was unrelated to features These studies encourage marketers to look for new ways to learn,
of Facebook use. Results support our hypothesis. Specifically, the characterize, and segment different groups of customers. Using
data demonstrates that individuals who scored higher on the Facebook or other social network applications for this purpose,
trait of conscientiousness have a greater number of friends than may be particularly effective. First, because on Facebook marketing
individuals who scored lower on the trait of conscientiousness. In researchers can observe their potential customers in their own
addition, individuals who scored higher on the trait of conscien- environments, devoid of any the pressures associated with tradi-
tiousness were also found to demonstrate less use of the picture tional marketing research. Secondly, marketers can receive more
upload features than individuals who scored lower on the trait of comprehensive details of the consumers interactions with their
conscientiousness. environment.
Overall, these results demonstrate very clearly that personality Our study provides an instrument for achieving marketers’
is, in fact, clearly linked to Facebook use. We have demonstrated aims. Nov and Ye (2008) suggest that innovators are more open
that each of the personality factors examined was relevant to as- to experience than other customers. Thus by analyzing the profile
pects of Facebook use. This is consistent with other studies that of potential customers on Facebook, marketers can classify which
demonstrated the link between personality and the Internet in customers are higher in openness and concentrate their efforts
other areas, such as blog use (Guadagno et al., 2007); Wikipedia on that segment. Moreover, by using a Facebook environment,
(Amichai-Hamburger, Lamdan, Madiel, & Hayat, 2008); and nostal- marketers can estimate the activity of each customer in the net-
gic websites (Amichai-Hamburger et al., 2008). work (i.e. number of friends, general activity) and so manage their
own efforts more efficiently.
4.1. Limitations and future research This study has shown that personality is very much related to
social network use. This finding has created new questions. These
The results were not all in line with our predictions. It seems center on the interaction between personality and an additional
clear that other factors, besides that of personality, have a bearing variables that affect users’ behavior on social networks. This is in
on Facebook use. Such factors may, for example, include social the nature of pioneer research studies. It is important to further
norms. People function in social groups and organizations that our understanding of the impact of personality on the use of social
have expectations as to how to use Facebook. These, in turn may networks in order to enhance the psychological wellbeing of the
create pressure on the individual as to what is appropriate to dis- surfer.
 Y. Amichai-Hamburger, G. Vinitzky / Computers in Human Behavior 26 (2010) 1289–1295 1295

Appendix 1. Rules for encoding Facebook pages (abbreviated) Amichai-Hamburger, Y., Wainapel, G., & Fox, S. (2002). On the Internet no one
knows I’m an introvert: Extroversion, neuroticism, and Internet interaction.
Variable Value and value label Cyberpsychology & Behavior, 2, 125–128.
Boyd, D. (2007). Why youth (heart) social network sites?: The role of networked
Number of Friends R’s Friend publics in teenage social life. In D. Buckingham (Ed.), MacArthur Foundation
0 – N/A Series on Digital Learning – Youth, identity, and digital media volume
– # of Friends (pp. 119–142). Cambridge, MA: MIT Press.
Boyd, D. M., & Ellison, N. B. (2007). Social network sites: Definition, history, and
Number of Pictures R’s Pictures scholarship. Journal of Computer-Mediated Communication, 13(1), article 11.
0 – N/A Available from: http://jcmc.indiana.edu/vol13/issue1/boyd.ellison.html.
– # of Pictures Costa, P. T., & McCrae, R. R. (1992). NEO PI-R. Professional manual. Odessa, FL:
Psychological Assessment Resources.
Gender R’s Gender Donath, J., & Boyd, D. (2004). Public displays of connection. BT Technology Journal,
0 – N/A 22(4), 71–82.
1 – Male Ellison, N., Steinfield, C., & Lampe, C. (2007). The benefits of Facebook ‘‘friends”:
Exploring the relationship between college students’ use of online social
2 – Female
networks and social capital. Journal of Computer-Mediated Communication, 12(3),
College R’s College article 1. Available from: http://jcmc.indiana.edu/vol12/issue4/ellison.html.
0 – Not stated Retrieved July 30, 2007.
Goffman, E. (1959). The presentation of self in everyday life. New York: Doubleday.
1 – Name
Goldenberg, J., Han, S., Lehmann, D. R., & Hong, J. W. (2009). The role of hubs in the
Self Picture R’s Self Picture adoption processes. Journal of Marketing, 73(2), 1–13.
0 – N/A Guadagno, R. E., Okdie, B. M., & Eno, C. A. (2008). Who blogs? Personality predictors
1 – Name of blogging. Computers in Human Behavior, 24, 1993–2004.
Hamburger, Y. A., & Ben-Artzi, E. (2000). The relationship between extraversion and
Birthday R’s Birthday neuroticism and the different uses of the Internet. Computers in Human
0 – N/A Behavior, 16, 441–449.
1 – Month, day, year Kratzer, J., & Lettl, C. (2009). Distinctive roles of lead users and opinion leaders in
the social networks of schoolchildren. Journal of Consumer Research, 36(4),
Group R’s Groups joined 646–659.
0 – N/A Kinar, O., & Amichai-Hamburger, Y. (2008). The effects of need for cognition on
– # of Groups joined Internet use revisited. Computers in Human Behavior, 24, 361–371.
Kraut, R., Lundmark, V., Patterson, M., Kiesler, S., Mukopadhyay, T., & Scherlis, W.
Album R’s Albums (1998). Internet paradox: A social technology that reduces social involvement
0 – N/A and psychological well-being? American Psychologist, 53, 1017–1031.
– # of Albums Lin, S. S. J., & Tsai, C. C. (2002). Sensation seeking and Internet dependence
of Taiwanese high school adolescents. Computers in Human Behavior, 18,
411–426.
Marwick, A. (2005). I’m a lot more interesting than a Friendster profile: Identity
References presentation, authenticity, and power in social networking services. Paper
presented at Internet Research 6.0, Chicago, IL.
Amichai-Hamburger, Y. (2002). Internet and personality. Computers in Human Nov, O., & Ye, C. (2008). Personality and technology acceptance: The case for
Behavior, 18, 1–10. personal innovativeness in IT, openness and resistance to change. In Proceedings
Amichai-Hamburger, Y. (2005). Personality and the Internet. In Y. Amichai- of the 41st Hawaii International Conference on System Sciences (HICSS 41). Hawaii,
Hamburger (Ed.), The social net: Human behavior in cyberspace (pp. 27–55). USA: IEEE Press.
New York: Oxford University Press. Pornsakulvanich, V., Haridakis, P. M., & Rubin, A. M. (2008). The influence of
Amichai-Hamburger, Y., & Barak, A. (2009). Internet and well-being. In Y. Amichai- dispositions and Internet-use motivation on online communication satisfaction
Hamburger (Ed.), Technology and Well-being (pp. 34–76). Cambridge University and relationship closeness. Computers in Human Behavior, 24, 2292–2310.
Press. Ross, C., Orr, E. S., Sisic, M., Arseneault, J. M., Simmering, M. J., & Orr, R. R. (2009).
Amichai-Hamburger, Y., Fine, A., & Goldstein, A. (2004). The impact of Internet Personality and motivations associated with Facebook use. Computers in Human
interactivity and need for closure on consumer preference. Computers in Human Behavior, 25, 578–586.
Behavior, 20, 103–117. Shakespeare, W. (1564–1616 reprinted 1978). Eight Comedies. Franklin Center, PA:
Amichai-Hamburger, Y., Kaplan, H., & Dorpatcheon, N. (2008). Click to the past: The The Franklin Library.
impact of extroversion by users of nostalgic website on the use of Internet social Sundén, J. (2003). Material virtualities: Approaching online textual embodiment. New
services. Computers in Human Behavior, 24, 1907–1912. York: Peter Lang.
Amichai-Hamburger, Y., Kinar, O., & Fine, A. (2007). The effects of need for cognition Zhao, S., Grasmuck, S., & Martin, J. (2008). Identity construction on Facebook: Digital
on Internet use. Computers in Human Behavior, 23, 880–891. empowerment in anchored relationships. Computers in Human Behavior, 24,
Amichai-Hamburger, Y., Lamdan, N., Madiel, R., & Hayat, T. (2008). Personality 1816–1836.
characteristics of Wikipedia members. Cyberpsychology & Behavior, 11(6), Zinman, A., & Donath, J. (2007). Is Britney Spears spam? Paper presented at the
679–681. Fourth Conference on Email and Anti-Spam, Mountain View, CA.