Computers in Human Behavior 28 (2012) 267–274

Contents lists available at SciVerse ScienceDirect

Computers in Human Behavior

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

The influence of personality on Facebook usage, wall postings, and regret

Kelly Moore a,⇑, James C. McElroy b
a
Department of Marketing, 2350 Gerdin Business Building, Iowa State University, Ames, IA 50010-1350, United States
b
Department of Management, 2350 Gerdin Business Building, Iowa State University, Ames, IA 50010-1350, United States

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

Article history: The Five Factor Model of personality has been used extensively in the management and psychology fields
Available online 17 October 2011 to predict attitudes and behaviors. Only recently have researchers begun to examine the role of psycho-
logical factors in influencing an individual’s use of technology platforms, such as Facebook. This study
Keywords: uses both a survey of Facebook users and actual Facebook data to uncover why some individuals are more
Facebook involved in Facebook than others. 219 undergraduate students participated in a survey that assessed
Personality their personality and their reported usage of Facebook. Of these, 143 voluntarily befriended the investi-
Five Factor Model (FFM)
gator, which gave her access to their actual Facebook sites and objective data on their number of friends,
photos, and wall postings. Results showed personality to explain significant amounts of variance over and
above gender and Facebook experience in terms of actual number of Facebook friends, the nature of their
wall postings and on their level of regret for inappropriate Facebook content.
Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction social networks (Amiel & Sargent, 2004; Ross et al., 2009), and their
attitudes toward social networking (Gangadharbatla, 2008). Many
The Internet has opened many new avenues through which of these studies and others (Hamburger & Ben-Artzi, 2000) offer
people can communicate and socialize, with social networking insights into what kind of information people include on Facebook,
sites (SNSs) playing an important part. By the second quarter of their attitudes toward using Facebook, the frequency of their use of
2008, Forrester Research estimated 75% of Internet users were the various features of this medium, as well as on gender differ-
involved in some sort of ‘social media’ (Kaplan & Haenlein, ences (Correa et al., 2010; Hamburger & Ben-Artzi, 2000). However,
2010). Specifically, Facebook users account for about 37.5% of the this research relies almost exclusively on self-reported, rather than
entire US population (Saleem, 2010) and Facebook accounts for actual usage. A notable exception is the work of Amichai-
an astonishing 17.9% of all time spent online (Srinivasan, 2009a). Hamburger and Vinitzky (2010), who encoded individuals’ actual
This growing new trend has prompted researchers to become Facebook page content. Looking at Facebook user pages, they rated
interested in what types of people rely on online social media tools the amounts of basic, personal, educational, and work-related
in their interactions with others (Correa, Hinsley, & de Zúñiga, information about Facebook users. They then examined how the
2010). amounts of each type of information were affected by the person-
Most research regarding Facebook relates to identity presenta- ality of respective users. Our study extends this body of research by
tion and privacy concerns (Ellison, Steinfield, & Lampe, 2007; Ross going beyond the association between personality and self-
et al., 2009). The purpose of this study is to examine the role of per- reported Facebook usage and features (e.g., number of friends
sonality in Facebook usage. In doing so, we look at how personality and photos) to an examination of actual Facebook content. Specif-
affects the digital footprint people leave on this popular social net- ically, we extend the work of Amichai-Hamburger and Vinitzky
work, not only in terms of time spent on Facebook, but also in (2010) on actual Facebook content to include wall posting activity.
terms of its use and content. We hope to address the issue of That is, we consider how personality traits influence the extent to
how personality influences the degree to which individuals use which Facebook users post primarily about themselves or about
this form of social networking and the content they include. others.
Our study has two noteworthy contributions. First, previous Second, questions have arisen about the appropriateness of con-
studies have examined the relationship between the Big Five per- tent being posted on Facebook and other social networking sites
sonality traits and Facebook usage, however much of this has been and the fact that other parties (e.g., universities and employers)
limited to surveys of Facebook users, their motives for turning to may gain access to Facebook information and use it in making deci-
sions that adversely affect the Facebook accountholder (Brady,
⇑ Corresponding author. Tel.: +1 515 294 7860; fax: +1 515 294 7112. 2006). Previous research has looked at user perceptions of the
E-mail address: [email protected] (K. Moore).
appropriateness of information posted on Facebook. For example,

0747-5632/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.chb.2011.09.009
268 K. Moore, J.C. McElroy / Computers in Human Behavior 28 (2012) 267–274

Peluchette and Karl (2008) found that 20% of Facebook users extraversion are expected to engage in high amounts of social
reported posting information that they would not want their interaction and approach others more easily (Wehrli, 2008). It is
employers to see, with males being more likely than females to post the least debatable personality trait as it relates to Facebook usage
questionable comments or pictures on Facebook. Disclosing too because it has consistently shown strong, although sometimes
much of an individual’s personal life is very easy to do on a medium contradictory, effects in prior studies. Research has offered two
like Facebook, and can very quickly lead to regret, which in its most competing explanations for the relationship between extraversion
extreme form can lead to ‘‘Facebook suicide’’ or the closing of a user and Facebook usage: social compensation and the ‘‘rich-get-richer’’
account (Justice, 2007). While some research exists on the possible (Ong et al., 2010). According to the social compensation explana-
causes of why people make postings on Facebook they later regret tion, introverts would have the most to gain from the use of social
(Wang et al., 2011), no research exists, however, documenting the networks like Facebook because such indirect communication
degree to which individuals vary in their sense of concern or regret allows them to compensate for their lack of interpersonal skills,
over such Facebook activity. To begin to fill this gap in the literature, while the ‘‘rich-get-richer’’ proposition argues that extraverts ben-
we look at the effects of personality on Facebook users’ sense of efit more since Facebook simply provides another platform for
regret regarding their use of Facebook. them to communicate with friends and contacts made off-line.
While some research has demonstrated that extraverts spend less
time in chat rooms (McElroy et al., 2007), presumably because they
2. Personality and Facebook
prefer face-to-face communication, most research lends support to
the ‘‘rich-get-richer’’ argument. However, the extent of the role of
Personality psychologists have reached a consensus that the
extraversion is dependent on the nature of Facebook usage.
domain of personality can best be described by the Big Five dimen-
Wehrli (2008) and Correa et al. (2010), for example, found
sions of the Five Factor Model (FFM) (Devaraj, Easley, & Crant, 2008).
extraversion to be positively related to the use of social networks.
Barrick, Mount, and Judge (2001) described FFM as the most useful
Extraverts spend more time on social network sites (Wilson,
taxonomy in personality research, while Costa and McCrae (1992)
Fornasier, & White, 2010) and they report having a larger Facebook
consider it the most comprehensive and parsimonious model of per-
imprint in terms of belonging to more Facebook groups (Ross et al.
sonality. The FFM has received considerable empirical support and is
(2009) and more Facebook friends (Ong et al., 2010) than less
now considered the standard personality trait measure (Wehrli,
extraverted users. However, when examining actual Facebook
2008). The five personality factors; extraversion, agreeableness,
activity, Amichai-Hamburger and Vinitzky (2010) found just the
conscientiousness, emotional stability, and openness to new experi-
opposite; that extraversion was not related to the number of Face-
ences; relate to people’s behavior in a wide variety of contexts (Weh-
book groups to which one belonged but that it was positively
rli, 2008). Personality was chosen over other individual differences
related to the actual number of Facebook friends one had. More
such as cognitive style (e.g., Meyers-Briggs Type Indicator) due to
importantly, extraversion does seem to be related to the nature
recent evidence suggesting the Big Five personality factors predict
of Facebook usage and content. Correa et al. (2010) found that
Internet use better than cognitive style (Davis, Bagozzi, & Warshaw,
extraverts reported making more contact with their Facebook
1989; Devaraj et al., 2008; McElroy, Hendrickson, Townsend, &
friends, were more likely to broadcast their activities and events
DeMarie, 2007). Moreover, Amichai-Hamburger (2002) has made
on Facebook, and posted more pictures on Facebook. Moreover,
the case that personality is a major predictor of internet behavior,
Bibby (2008) found that extraverts engaged in more self-disclosure
while others have linked personality to on-line activities such as
through self-generated Facebook content. On the other hand,
blogging (Guadagno, Okdie, & Eno, 2008) and contributing to Wiki-
Amichai-Hamburger and Vinitzky (2010) reported extraverts
pedia (Amachai-Hamburger, Kaplan, & Dorpatcheon, 2008).
engaging in less divulgence of personal information on their Face-
Previous researchers have looked at the role of personality traits
book profiles. These findings suggest that extraversion is more clo-
as they relate to the Internet, in general (Amiel & Sargent, 2004;
sely related to personal disclosure of one’s current activities and
Devaraj et al., 2008; Engelberg & Sjöberg, 2004; McElroy et al.,
thoughts as opposed to established interests, favorites (e.g., songs,
2007; Swickert, Hittner, Harris, & Herring, 2002) and to social med-
movies, etc.), and relationship status, all of which are more likely
ia, such as Facebook, in particular (Amichai-Hamburger & Vinitzky,
already known to friends of extraverts. This speculation is consis-
2010; Amichai-Hamburger, Wainapel, & Fox, 2002; Ross et al.,
tent with the view of Amiel and Sargent (2004), who argued that
2009; Wehrli, 2008). With respect to the Internet, Amiel and
extraverts would see social networks as places to share informa-
Sargent (2004) explored the relationship between personality
tion and opinions rather than as a substitute for real interaction.
and internet usage motives. They found those scoring high in neu-
Finally, because extraverts use social networks for self-disclosure,
roticism (low emotional stability) reported using the Internet to
they are less likely to regret doing so than are introverts. Therefore,
feel a sense of ‘‘belonging’’ and to be informed, while extraverts
we propose:
made more instrumental and goal-oriented use of Internet services
(Amiel & Sargent, 2004). McElroy et al. (2007) found personality to
Hypothesis 1. Extraversion will be related to Facebook usage such
be a much better predictor of internet use than cognitive style,
that more extroverted users will report spending more time on
while Devaraj and associates’ (2008) results showed a moderating
Facebook, use it more frequently, have more Facebook friends,
role for personality on the relationship between technology useful-
more frequently post information on their walls, post more photos,
ness and intention to use and between subjective norms and inten-
and engage in more self-generated wall postings, and report less
tion to use (with the exception of openness).
regret over what they post. Because other-generated wall postings
The research involving Facebook has found some dimensions of
do not involve self-disclosure, we propose no relationship between
personality to be better predictors of Facebook usage than others.
extraversion and wall postings about others.
Moreover, personality has been shown to influence some aspects
of Facebook usage but not others, as noted below.
2.2. Agreeableness and Facebook
2.1. Extraversion and Facebook
Agreeable persons represent the tendency to be sympathetic,
Extraversion refers to the extent to which individuals are social, courteous, flexible, kind, trusting and forgiving. Individuals high
cheerful, optimistic, active and talkative. Individuals high in in agreeableness have been known to avoid conflict, but are
 K. Moore, J.C. McElroy / Computers in Human Behavior 28 (2012) 267–274 269

presumed to not reject an offer of friendship (Wehrli, 2008). as distrustfulness, sadness, anxiety, embarrassment, and difficulty
Generally speaking, agreeableness is said to favorably influence so- managing stress. Neuroticism has previously been assumed to be
cial interactions and their perceived quality (Wehrli, 2008). How- negatively associated with social relationships (Wehrli, 2008) but
ever, agreeable people might not necessarily be driven to positively related to the amount of time they spend online (McEl-
establish an on-line connection (Swickert et al., 2002). Compared roy et al., 2007).
to extraversion, much less research has been conducted relative Early studies found individuals high on neuroticism were hea-
to agreeableness and Facebook usage. Ross et al. (2009) and Amic- vier Internet users than extraverts (Amichai-Hamburger et al.,
hai-Hamburger and Vinitzky (2010), for example, found no rela- 2002; Correa et al., 2010). Recently, Wehrli (2008) found a positive
tionship between agreeableness and Facebook usage, while relationship between neuroticism and social media usage, offering
Swickert et al. (2002) found a significant relationship between the explanation that individuals low in emotional stability tend to
agreeableness and the reported use of recreational Facebook activ- spend more time on social networking sites because they may try
ities, such as instant messaging and online games. In related to make themselves look as attractive as possible.
research, Amachai-Hamburger, Kaplan, and Dorpatcheon (2008) People who exhibit neurotic tendencies like to use chat rooms
found that individuals low in agreeableness were more likely to (Hamburger & Ben-Artzi, 2000) and instant messaging (Ehrenberg,
be Wikipedia members, presumably because they could correct Juckes, White, & Walsh, 2008). Ross and associates (2009) found
others without having to be sympathetic or courteous. This sug- people high on the trait of neuroticism reported the Wall as their
gests that agreeableness may be negatively related to frequency favorite Facebook component. Amichai-Hamburger and Vinitzky
of use, wall postings about others and to regret. That is, people (2010) found that highly neurotic people were more likely to post
who are low in the trait of agreeableness may be more likely to private information and post fewer photos on their Facebook pro-
use Facebook, post comments on their own Facebook wall about file than those in the less neurotic group. Butt and Phillips (2008)
others, but express less regret in doing so. Consequently, while suggest that this is related to a need on the part of more neurotic
there is little basis for hypothesizing a relationship between agree- people for control over information. A related explanation is that
ableness and the size of one’s Facebook imprint (e.g., number of participation in online communication gives neurotic personalities
friends, photos and amount of time spent on Facebook) we do offer longer to contemplate what they are going to say as opposed to
the following hypothesis: face-to-face communication (Correa et al., 2010; Ehrenberg et al.,
2008; Ross et al., 2009). Given neurotic personalities are anxious
Hypothesis 2 . Agreeableness will be related to Facebook usage and nervous by nature, they would be more likely to be upset with
such that more agreeable users will use Facebook less frequently, and regret posting anything of a questionable nature.
make fewer wall postings about others and express more regret
over their Facebook activity. Hypothesis 4 . Emotional stability will be negatively related to
Facebook usage such that people higher in emotional stability will
spend less time on Facebook, use it less frequently, have fewer
2.3. Conscientiousness and Facebook friends and express less regret, but positively related to Facebook
content; i.e., people higher in emotional stability will post more
Conscientiousness represents the tendency to be reliable, pictures and a greater number of self-generated wall postings.
responsible, organized and self-disciplined. Conscientious individ-
uals have a high intrinsic motivation to achieve and are usually try-
2.5. Openness to experience and Facebook
ing to improve their level of job performance. The rationale is that
if a highly conscientious person believes Facebook will not drive
Openness to experience represents an individual’s curiosity,
efficiency or production, they will have decreased behavioral
open-mindedness, and their willingness to explore new ideas.
intentions towards it (Devaraj et al., 2008). In other words, consci-
Openness reflects an individual’s imagination and originality. Little
entious individuals will not invest a large amount of resources in
research has been conducted on the relationship between open-
Facebook because they prefer to stick to their main goals by avoid-
ness and Facebook. Tangentially, McElroy et al. (2007) found that
ing distraction (Wehrli, 2008). Previous research indicating that
openness was a significant predictor of general Internet use, while
conscientious people spend more time online engaged in academic
Guadagno et al. (2008) found people high in openness to be more
pursuits than in leisure activities (McElroy et al., 2007) supports
likely to blog. Ross et al. (2009) found that individuals high on the
this notion, as does the finding of Wilson et al. (2010) that highly
trait of openness to experience were more willing to consider alter-
conscientious people spend less time on social networking sites.
native methods of communication, which is important in Facebook
However, such support is not unanimous (Ross et al., 2009). On
use. Correa et al. (2010) also found a positive relationship between
the other hand, being more responsible, highly conscientious peo-
openness and social media use. People higher in openness engaged
ple will more likely express regret over posting inappropriate
in increased online sociability through Facebook, a finding they
material. For the aforementioned reasons, we expect a person
attributed to the novel nature of this form of social communication
who is high in conscientiousness to use Facebook less than other
technology. Amichai-Hamburger and Vinitzky (2010) found that
individuals, but express higher levels of regret.
more open individuals revealed more personal information about
themselves on their Facebook profile. This suggests that individu-
Hypothesis 3 . Conscientiousness will be negatively related to
als with high scores in openness to experience will be more likely
Facebook usage such that people higher in conscientiousness will
to use and keep up with Facebook. Given their curious nature, peo-
spend less time on Facebook, use it less frequently, have fewer
ple high in openness to new experiences are less likely to regret
friends, post fewer pictures, make fewer wall postings, and express
their Facebook experiences.
more regret.

Hypothesis 5 . Openness to new experience will be positively

2.4. Emotional stability and Facebook related to Facebook usage such that people higher in openness will
spend more time on Facebook, use it more frequently, and initiate
The opposite of emotional stability is neuroticism, which refers more self-generated wall postings, than people low in openness,
to the extent to which individuals display negative attributes such but will be negatively related to regret.
270 K. Moore, J.C. McElroy / Computers in Human Behavior 28 (2012) 267–274

3. Method use of the ‘‘poking’’ function which conveys only an interest in

communicating later, not actual communication per se. The three
3.1. Sample items used asked respondents to indicate how frequently they
commented on others’ photos, posted on others’ walls, and
A sample of 219 undergraduate students at a large Midwestern checked their own walls, with response options ranging from never
university participated in the study. The students were offered to multiple times per day. Cronbach’s alpha for this scale was .84.
extra course credit to participate. Of the 219 students in the sam- Finally, regret was measured using a five item scale, which asked
ple, 204 (93%) indicated they had a Facebook account. Of the 204 respondents how often they posted material or comments that
students who had a Facebook account, 127 (63%) were male and they regretted later, postings or comments that they would not
77 (37%) were female. want their employer to see, or posted material that they would
not want their parents to see. A 4-point response format ranging
from never to frequently was used. Cronbach’s alpha for this
3.2. Procedure
scale was .85.
The second part of the study enabled investigators access to the
The study consisted of two parts. The first part entailed an elec-
actual Facebook pages of 143 respondents. The analysis of actual
tronic, web-based survey given in a campus computer lab, which
Facebook pages is similar to the methodology employed by
was completed by all 204 participants. Upon completion of the sur-
Amichai-Hamburger and Vinitzky (2010), although we look at dif-
vey, participants were invited to login to Facebook and send a
ferent types of information. Amichai-Hamburger and Vinitzky
friend invitation to one of the investigators. Those who voluntarily
(2010) examined what they refer to as user information upload;
sent a friendship request to the investigator did so knowing that
a profile of user basic information (e.g., user demographics), per-
this provided the investigator access to the respondent’s actual
sonal information (e.g., activities, interests, and favorite things),
Facebook page. One hundred forty three of the participants volun-
contact information (e.g., home address, e-mail) and educational/
tarily provided the investigator with access to their actual Face-
work information (e.g., school, company). We opted to look, not
book sites. To determine whether any differences existed
at information about the Facebook user, but what they do and
between those who participated in the second phase of the study
put on Facebook. Specifically, visiting respondent Facebook sites
and those who opted out, F-tests were conducted between the
enabled us to collect data on the actual number of friends and pho-
143 who provided the investigator with access to their Facebook
tos for each respondent and to evaluate their wall postings. Num-
pages and the 61 who did not. No significant differences were
ber of Facebook friends is an automated count on the website
found between the two groups on any of the five personality fac-
appearing in the users’ ‘‘friends’’ section. Number of photos was
tors or gender. The only significant difference was that more expe-
measured by the investigator counting the actual number of pho-
rienced Facebook users were less likely to grant the investigator
tos posted. Wall postings were read and categorized into self-
access to their sites than were less experienced users (F = 18.80;
focused (posting about oneself) versus postings about others over
p 6 .01).
a 5 day period by two experimenters. There were 184 total wall
posts over the 5-day period. Of the 184 wall posts, raters agreed
3.3. Measures on the categorization of 175 wall posts. Inter-rater agreement
was 95%, with a kappa statistic of .90, which according to Landis
The survey portion of the study allowed us to gather data on and Koch (1977) indicates nearly perfect agreement. The nine dis-
user personality and Facebook usage. The Big-Five personality fac- crepancies were discussed between raters and placed in an agreed
tors were measured using Goldberg et al. (2006) 50-item Interna- upon category.
tional Personality Item Pool (IPIP). A number of other measures Finally we collected data on two additional variables, gender
of the Five Factor Model of personality have been developed, and experience with Facebook. Research showing gender differ-
including the NEO Personality Inventory (NEO-PI-R) and the NEO ences in perceptions of appropriateness (Peluchette & Karl,
Five-Factor Inventory (NEO-FFI) (Costa & McCrae, 1992), the Cali- 2008), coupled with the fact that the longer one has been on Face-
fornia Psychological Inventory (Gough, 1987) and the Hogan Per- book the more likely they are to regret something, led us to control
sonality Inventory (Hogan & Hogan, 1992), all of which are for these two individual differences. Gender was entered as a dum-
proprietary. While most research on personality and Facebook my variable, with 1 = male and 2 = female. Experience with Face-
has relied on the NEO PI-R or NEO-FFI instrument (e.g., book was measured using a single item, 6-point response scale
Amichai-Hamburger & Vinitzky, 2010; Ross et al., 2009), the IPIP asking respondents to indicate how long they have had a Facebook
is very user friendly (i.e., nonproprietary and much shorter) and account, with responses ranging from 6 months to 3 or more years
research has shown strong evidence of convergent and discrimi- in 6 month increments.
nant validity and interchangeability with the NEO-FFI (Lim & Ploy-
hart, 2006). Scale reliabilities for the five personality dimensions of
the IPIP were acceptable with Cronbach’s alpha values of .90 for 4. Results
extraversion, .81 for agreeableness, .82 for conscientiousness, .83
for emotional stability and .79 for openness to experience. Means and standard deviations along with a correlation matrix
The Facebook usage measures included respondent assessments are shown in Table 1. Hierarchical regression was used to test the
of time spent on Facebook, frequency of use, and regret. Time spent effect of personality on Facebook usage and content. Gender and
on Facebook was measured by a single item asking respondents to length of experience using Facebook were entered in step 1 as con-
indicate the average amount of time spent per day on Facebook. A trol variables with their respective beta values and significant lev-
5-point response scale was used with options ranging from less els shown as Model 1 in Table 2. Following this, the Five Factor
than ½ h to over 2 h using ½ h increments. Frequency of use was Model of personality was entered in step 2. Their betas and signif-
measured using a three item 10-point scale, adapted from Ross icance levels are reported as Model 2 in Table 2.
et al. (2009). Ross et al. measured the frequency of basic use func- As shown in Table 2, gender had significant effects on Facebook
tions and included both public (e.g., posting photos) and private usage and content, both independently and in the presence of the
(private Facebook messages). For our study, we deleted the item personality factors. A significant positive relationship was found
dealing with the frequency of sending private messages and the between gender and a number of variables of interest. Specifically,
 K. Moore, J.C. McElroy / Computers in Human Behavior 28 (2012) 267–274 271

Table 1
Descriptive statistics and correlation matrix.a

Variables Mean S.D. 1 2 3 4 5 6 7 8 9 10 11 12 13 14

b
1. Gender 1.38 .49 –
2. Facebook experience 5.62 .97 .17 –
3. Extraversion 3.42 .73 .24 .18 –
4. Agreeableness 3.84 .51 .38 .17 .44 –
5. Conscientiousness 3.55 .58 .21 .02 .17 .27 –
6. Emotional stability 3.33 .63 .09 .01 .18 .16 .02 –
7. Openness 3.57 .53 .04 .11 .38 .31 .19 .18 –
8. Time spent 2.33 1.32 .28 .17 .15 .09 .02 .20 .02 –
9. Frequency of use 3.90 1.76 .40 .30 .30 .20 .02 .16 .06 .49 –
10. Actual # friends 423.93 334.16 .25 .27 .36 .23 .10 .03 .05 .23 .41 –
11. Number of photos 286.25 536.92 .35 .17 .22 .22 .08 .01 .09 .06 .28 .50 _
12. Self-postings .46 1.26 .11 .04 .11 .21 .03 .02 .13 .06 .18 .07 .11 –
13. Other-postings .49 1.29 .03 .01 .12 .15 .18 .12 .05 .02 .15 .12 .10 .63 –
14. Regret 3.04 .65 .07 .15 .18 .06 .17 .20 .09 .26 .28 .14 .11 .03 .10 –
a
Correlations greater than ±.14 are statistically significant at p 6 .05.
b
1 = male, 2 = female.

women reported spending more time on Facebook, had a greater to time spent on Facebook, number of photos, or the number of
number of Facebook friends, posted more photos and more post- wall postings (either about self or others).
ings about themselves than did males. The significant negative Lack of research on the role of agreeableness and Facebook jus-
relationship between gender and frequency of use suggests that tified only a couple of predictions in Hypothesis 2. The only finding
although women spend more time overall on Facebook, they visit supportive of Hypothesis 2 was the positive relationship between
their Facebook site less frequently than men do.1 agreeableness and regret. More agreeable people expressed greater
Facebook experience was also significantly related to several levels of regret about inappropriate content they may have posted
variables of interest in the study. The more experience Facebook on Facebook. The predictions that agreeableness would be nega-
users had, the less frequently they visited Facebook and the more tively related to frequency of use and number of wall postings
Facebook friends they had. The more experienced Facebook users about others received no support. Surprisingly, people higher in
were also likely to spend more time on Facebook, post more pho- agreeableness did make a greater number of postings about them-
tos, but have fewer posting about themselves on their Facebook selves than did less agreeable people.
walls. These latter findings, however, are superseded by the pres- Hypothesis 3 predicted that more conscientious people would
ence of the personality factors, which suggests that personality is be less likely to use Facebook, post fewer photos and wall postings,
a more important determinant of these aspects of Facebook use ostensibly because it distracts them from goal related activity.
and content than is experience. However, because of their responsible nature, a positive relation-
As shown in Table 2, the addition of the five personality factors ship between conscientiousness and regret was hypothesized.
significantly added to the variance explained in time spent using Results shown in Table 2 offer some support for the connection
Facebook, actual number of Facebook friends, the amount of self- between conscientiousness and Facebook content. People high in
generated postings and postings about others and user perceptions conscientiousness made significantly fewer wall postings, about
of regret over inappropriate Facebook content. The statistically sig- either self or others, and expressed more regret than did less con-
nificant amounts of additional variance explained by personality scientious users. Conscientiousness was not related to time spent,
over and above that explained by gender and experience on Face- frequency of use, number of friends or number of photos. Thus
book ranged from 6% for time spent on Facebook to 41% for post- Hypothesis 3 received only partial support.
ings about others. Personality did not significantly add to the Hypothesis 4 received little support. Emotional stability was
model’s prediction of frequency of Facebook use or the actual num- negatively related to time spent on Facebook. That is, more neu-
ber of photos people post on Facebook. rotic users spend more time on Facebook than do those higher in
Extraversion was predicted in Hypothesis 1 to be positively emotional stability. None of the other predictions proved signifi-
related to time spent on Facebook, frequency of use, number of cant. Emotional stability was not significantly related to actual
friends and photos and the number of postings about oneself and number of friends or photos, or to the number of wall postings
negatively related to regret. The results shown in Table 2 show that about either self or other. Unexpectedly, emotional stability was
more extraverted people have more Facebook friends and that they positively related to both how frequently they use Facebook to
report less regret over Facebook content than less extraverted indi- keep up with others and regret.
viduals. These findings lend some support for Hypothesis 1. Unex- Finally, openness proved to have no significant effect on either
pectedly, more extraverted individuals reported significantly less Facebook usage or content, as shown by the lack of significant find-
frequent use of Facebook for keeping up with others than intro- ings in Table 2. Consequently, Hypothesis 5 received no support.
verts; a finding that offers some support for the ‘‘social compensa-
tion’’ explanation of the relationship between extraversion and
5. Discussion
social network site use. Extraversion was not significantly related
Previous research has suggested that personality may not be as
1
As noted, gender was significantly related to most of the dependent variables influential a factor in the use of social networks as previously
both independently and in the presence of personality. At the suggestion of one of the thought (Ross et al., 2009). Our study, on the other hand, demon-
reviewers, we ran additional regression analyses entering in interaction effects for strated that personality accounted for significant amounts of vari-
gender and each of the five personality factors as additional predictors of each of the ance over and above that explained by gender and Facebook
seven dependent variables. Because only a couple of the 35 tests of significance
proved significant, a result that could be accounted for by chance, we conclude that
experience. While it only explained 6% of the variance in self-
gender does predict some aspects of Facebook usage and content but has little reported time spent on Facebook, personality was found here to
interaction with personality in doing so. explain 14% of the variance in regret, 16% of the variance in actual
272 K. Moore, J.C. McElroy / Computers in Human Behavior 28 (2012) 267–274

Table 2
Regression results for the effects of personality on Facebook usage, wall postings and regret.

Time spent Freq. of use Actual # friends # Photos Self-postings Other-postings Regret
Model Model Model 1 Model 2 Model 1 Model 2 Model 1 Model 2 Model Model Model Model Model Model
1 2 1 2 1 2 1 2
Control variables
Gender .61** .58** 1.41*** 1.26*** 148.28*** 91.39* 527.22*** 521.86*** 1.32* 1.38* .60 .65 .06 .12
Facebook .19* .18 .37** .31** 114.91*** 94.72*** 92.79*** 71.13 .84* .51 .03 .01 .09 .07
experience
Personality
Extraversion .27 .55** 169.11*** 131.11 .12 .52 .22**
Agreeableness .18 .03 27.23 9.99 1.50* 1.34 .25*
Conscientiousness .23 .27 7.39 60.71 1.16* 1.52** .24**
Emotional .45** .43* 1.20 6.85 .11 .45 .22**
stability
Openness .04 .13 23.78 100.25 .65 .45 .17
R2 .08 .15 .22 .29 .24 .40 .21 .27 .20 .44 .02 .43 .02 .17
Adjusted R2 .07 .11 .22 .26 .23 .37 .20 23 .16 .31 .03 .30 .01 .13
F 8.45*** 4.36*** 27.36*** 10.69*** 23.75*** 13.72*** 17.71*** 6.21* 4.31* 3.31* .36 3.44** 2.16 5.15***
AF 2.58* 3.34 7.60*** 1.30 2.53* 4.59** 6.22***
AR2 .07 .07 .16 .06 .24 .41 .15
*
p 6 .05.
**
p 6 .01.
***
p 6 .001.

number of Facebook friends, 24% of the variance in the number of interactions when engaging in recreational Facebook activities
self-postings and 41% of the variance in the number of postings (wall posting in the case of this study). As expected, we found that
about others. These effect sizes are substantial in light of previous highly conscientious people use the Facebook wall function signif-
research suggesting that one of the problems in personality icantly less than other individuals and that individuals high in
research is small effect sizes (Ross et al., 2009). The discrepancy agreeableness are more likely to post wall content about them-
in the magnitude of the role of personality may be a function of selves. We found no support for Ross et al.’s (2009) findings that
what is being measured. That is, personality appears to have a those high in emotional stability have a higher preference for post-
much larger role in predicting actual Facebook usage and content, ing photos, that those low in emotional stability have a higher pref-
as attested to here and in the work of Amichai-Hamburger and erence for wall posting, or that those high in openness send more
Vinitzky (2010), than in predicting self-reported Facebook usage, messages to others (other-directed wall postings).
as was the case in Ross et al.’s (2009) work. The message here is This research also suggests that both gender and experience are
clear. To accurately assess the role of personality, one should use important predictors of Facebook usage and content, but not
actual Facebook data where possible and rely on survey data for regret, and should, at minimum, be controlled for in future
information that cannot be obtained objectively, such as feelings research. Experience becomes less important as a predictor of
of regret or one’s motivation for using particular Facebook features. Facebook use and content in light of personality differences; a fact
Future research could test this by making direct comparisons that makes sense, since if personality is a predictor of Facebook,
between the effects of personality on self-reported versus actual then certain personalities are more likely to acquire such experi-
measures of number of friends, photos, time spent on Facebook, ence. Gender, on the other hand, remained a significant predictor
etc. even when personality is factored in. While gender did not interact
The purpose of this study was to investigate the role of person- with personality factors in this sample, it warrants additional study
ality on Facebook use and content. In terms of Facebook usage, less as a variable of interest in research on Facebook.
emotionally stable (neurotic) individuals report spending more
time on Facebook, while more emotionally stable and more intro- 5.1. Limitations and future research
verted users report more frequently going to Facebook to keep up
with friends. All of the personality factors are related to regret, As with previous research, our findings were not always in line
with the exception of openness to new experiences, with more with our predictions. Our predictions were based on previous
agreeable, more conscientious, more emotionally stable and less research, some of which is equivocal and some of which is based
extraverted users reporting greater levels of regret for inappropri- on preferences or self-reports versus others on actual Facebook
ate content. data. A number of methodological issues require caution in inter-
With respect to Facebook content, this study confirms the work preting the results of this study and in comparing our results with
of Amichai-Hamburger and Vinitzky (2010) showing that extra- previous research. For example, a direct comparison of our results
verts have a significantly wider social network than introverts. In with previous research must be tempered by the fact that we used
fact, in our study, the 10% of our respondents scoring the highest a different measure of personality (IPIP), in spite of evidence of its
in extraversion had, on average, 484 more friends than the 10% interchangeability with the NEO-FFI (Lim & Ployhart, 2006). In
scoring the lowest in extraversion. We also lend some support to addition, the absence of a standardized measure of regret in the
the notion that introverts use social networks to compensate for context of Facebook usage, resulted in the use of a self-developed
a lack of interpersonal communication (Ong et al., 2010). We found five item measure. While all five items loaded on a single factor,
no evidence, however, to support Bibby’s (2008) claim that extra- more work needs to be done on the development of a regret instru-
verts self-disclose through self-generated content such as wall ment if this line of inquiry is to continue. Finally, sample size in
postings. This finding supports the positive social interaction experimental research, particularly that involving personality, is
hypothesis offered by Swickert and colleagues (2002) that suggests an issue. The loss of 61 participants between the two parts of this
that highly agreeable individuals may experience more positive study, in spite of a lack of substantial differences between those
 K. Moore, J.C. McElroy / Computers in Human Behavior 28 (2012) 267–274 273

who participated in part 2 and those who did not, remains a limi- development of a theoretical framework to explain why some
tation. While small effect sizes are common to research on person- people devote varying amounts of time and energy to this phe-
ality (Ross et al., 2009), Amichai-Hamburger and Vinitzky (2010) nomenon. Personality has certainly earned a place in such a frame-
reported effect sizes of .05 (for the effect of personality on personal work. But with the development of additional social network sites
information) and .07 (for number of friends); effect sizes that fall (e.g., Linked-In), such a framework would help identify early
between small and medium (but closer to small), according to adopters, as well as extent of use. Incorporating personality into
Cohen (1992). Small effect sizes require very large sample sizes existing models of technology adoption and use, such as TAMS 2
for research involving the number of independent variables used (Venkatesh & Davis, 2000) and UTAUT (Unified Theory of Accep-
in our study, while medium effect sizes call for a sample size of tance and Use of Technology, Venkatesh, Morris, Davis, & Davis,
around 100 subjects (Cohen, 1992). Consequently, a larger sample 2003), would be a useful first step.
size would have enabled us to be more confident of our results.
An additional limitation is that we looked at the role of the Big
Acknowledgement
Five personality factors on certain aspects of Facebook usage and
content but other variables, both independent and dependent, pro-
The authors would like to thank Dr. Marc Anderson for his com-
vide direction for future research. For example, Ross et al. (2009)
ments on an earlier version of this paper.
has suggested that specific traits or motivations not captured by
the Big Five factors may provide additional insight into Facebook
use and content and we echo that call. To their suggestions of shy- References
ness, narcissism, and desire for communication, we would add self-
Amachai-Hamburger, Y., Kaplan, H., & Dorpatcheon, N. (2008). Click to the past: The
efficacy and need for affiliation, among others. Additional research, impact of extroversion by users of nostalgic website on the use of Internet social
not only on whom but on why people use Facebook is also war- services. Computers in Human Behavior, 24, 1907–1912.
ranted. Moreover, gaining access to actual Facebook sites opens Amichai-Hamburger, Y. (2002). Internet and personality. Computers in Human
Behavior, 18, 1–10.
up additional avenues for future research examining actual Face- Amichai-Hamburger, Y., & Vinitzky, G. (2010). Social network use and personality.
book content such as the nature of photographs posted, the con- Computers in Human Behavior, 26(6), 1289–1295.
tent of postings (e.g., communication, social, work related), Amichai-Hamburger, Y., Wainapel, G., & Fox, S. (2002). ‘‘On the Internet No One
Knows I’m an Introvert’’: Extroversion, neuroticism, and internet interaction.
comments posted by one’s friends, privacy settings, and the use CyberPsychology & Behavior, 5(2), 125–128.
of specific Facebook features. The concept of regret also offers Amiel, T., & Sargent, S. (2004). Individual differences in Internet usage motives.
opportunities for future research. While we were one of the first Computers in Human Behavior, 20(6), 711–726.
Barrick, M., Mount, M., & Judge, T. (2001). Personality and performance at the
to look at the role of personality on regret, we only generally
beginning of the new millennium: What do we know and where do we go next?
tapped into this concept. Future research could extend Wang International Journal of Selection and Assessment, 9(1&2), 9–30.
et al. (2011) research on what type of Facebook content leads to Bibby, P. A. (2008). Dispositional factors in the use of social networking sites:
regret, whether it is the content per se or the consequences associ- Findings and implications for social computing research. Lecture Notes in
Computer Science, 5075, 392–400.
ated with such content that causes regret, and the varying amounts Brady, E. (2006). What you say online could haunt you. USA Today (March 9), 1a.
of regret generated by such content. Butt, S., & Phillips, J. (2008). Personality and self reported mobile phone use.
Additional control variables could also be examined. We only Computers in Human Behavior, 24(2), 346–360.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155–159.
controlled for gender and experience, but a number of other indi- Correa, T., Hinsley, A., & de Zúñiga, H. (2010). Who interacts on the Web?: The
vidual differences such as age, family size (which yields a poten- intersection of users’ personality and social media use. Computers in Human
tially higher number of ‘‘friends’’), and the extent to which a Behavior, 26(2), 247–253.
Costa, P., Jr., & McCrae, R. (1992). NEO personality inventory-revised (NEO-PI-R) and
Facebook user belongs to other social network sites may prove rel- NEO five-factor inventory (NEO-FFI) professional manual. Odessa, FL:
evant. Finally, research on the connection between personality and Psychological Assessment Resources.
Facebook has yet to address the issue of potential intervening vari- Davis, F., Bagozzi, R., & Warshaw, P. (1989). User acceptance of computer
technology: A comparison of two theoretical models. Management Science,
ables. For example, is it possible to become committed to Facebook 35(8), 982–1003.
(over other social network sites) or even addicted to it and, if so, Devaraj, S., Easley, R. F., & Crant, J. M. (2008). Research note – How does personality
how does that affect Facebook use and content? matter? Relating the five-factor model to technology acceptance and use.
Information Systems Research, 19(1), 93.
Ehrenberg, A., Juckes, S., White, K., & Walsh, S. (2008). Personality and self-esteem
5.2. Implications as predictors of young people’s technology use. CyberPsychology & Behavior,
11(6), 739–741.
Facebook has become an important phenomenon to scholars Ellison, N., Steinfield, C., & Lampe, C. (2007). The benefits of Facebook ‘‘friends’’:
Social capital and college students’ use of online social network sites. Journal of
and practitioners alike. The site has become very popular and Computer Mediated Communication-Electronic Edition, 12(4), 1143.
gained an extremely large audience as it surpassed MySpace for Engelberg, E., & Sjöberg, L. (2004). Internet use, social skills, and adjustment.
the SNS with the highest market share (Srinivasan, 2009b). More- CyberPsychology & Behavior, 7(1), 41–47.
Gangadharbatla, H. (2008). Facebook me: Collective self-esteem, need to belong,
over, the implications for businesses are wide-ranging. Many busi- and internet self-efficacy as predictors of the iGeneration’s attitudes toward
nesses are changing the way they conduct their marketing social networking sites. Journal of Interactive Advertising, 8(2), 1–28.
activities through the use of Facebook as an advertising vehicle, Goldberg, L., Johnson, J., Eber, H., Hogan, R., Ashton, M., Cloninger, C., et al. (2006).
The international personality item pool and the future of public-domain
distribution channel, and to foster word-of-mouth referrals. The personality measures. Journal of Research in Personality, 40(1), 84–96.
degree to which we can ferret out the relationship between per- Gough, H. G. (1987). California Psychological Inventory administrator’s guide. Palo
sonality and Facebook offers companies valuable insight into the Alto, CA: Consulting Psychologists Press.
Guadagno, R., Okdie, B., & Eno, C. (2008). Who blogs? Personality predictors of
nature of who is likely to comment on company Facebook sites blogging. Computers in Human Behavior, 24(5), 1993–2004.
about things such as product usage and quality and the extent of Hamburger, Y., & Ben-Artzi, E. (2000). The relationship between extraversion and
potential word of mouth advertising conveyed through personal neuroticism and the different uses of the Internet. Computers in Human
Behavior, 16(4), 441–449.
Facebook pages.
Hogan, R., & Hogan, J. (1992). Manual for the Hogan personality inventory. Tulsa, OK:
Understanding the role played by personality as an individual Hogan Assessment Systems.
difference and its impact on SNS usage will help researchers Justice, E. (2007). Facebook suicide: The end of a virtual life. UK: The Times (Online).
explain how technology usage in general evolves. Research has Kaplan, A., & Haenlein, M. (2010). Users of the world, unite! The challenges and
opportunities of Social Media. Business Horizons, 53(1), 59–68.
now identified a number of relationships between personality Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for
and Facebook usage and content. What is needed now is the categorical data. Biometrics, 33, 159–174.
274 K. Moore, J.C. McElroy / Computers in Human Behavior 28 (2012) 267–274

Lim, B.-C., & Ployhart, R. E. (2006). Assessing the convergent and discriminant Swickert, R., Hittner, J., Harris, J., & Herring, J. (2002). Relationships among Internet
validity of Goldberg’s International Personality Item Pool. Organizational use, personality, and social support. Computers in Human Behavior, 18(4),
Research Methods, 9(1), 29–54. 437–451.
McElroy, J., Hendrickson, A., Townsend, A., & DeMarie, S. (2007). Dispositional Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the technology
factors in Internet use: Personality versus cognitive style. MIS Quarterly, 31(4), acceptance model: Four longitudinal field studies. Management Science, 46(2),
809–820. 186–204.
Ong, E. Y. L., Ang, R. P., Ho, J. C. M., Lim, J. C. Y., Goh, D. H., & Lee, C. S. (2010). Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of
Narcissism, extraversion and adolescents’ self-presentation on Facebook. information technology: Toward a unified view. MIS Quarterly, 27(3),
Personality and Individual Differences, 50(2), 180–185. 425–478.
Peluchette, J., & Karl, K. (2008). Social networking profiles: An examination of Wang, Y., Komanduri, S., Leon, P.G., Norcie, G., Acquisti, A., & Cranor, L.F. (2011). I
student attitudes regarding use and appropriateness of content. regretted the minute I pressed share: A qualitative study of regrets on Facebook.
CyberPsychology & Behavior, 11(1), 95–97. In Symposium on Usuable Privacy and Security (SOUPS), Pittsburgh, PA (July 20–
Ross, C., Orr, E. S., Sisic, M., Arseneault, J. M., Simmering, M. G., & Orr, R. R. (2009). 22).
Personality and motivations associated with Facebook use. Computers in Human Wehrli, S. (2008). Personality on social network sites: An application of the five
Behavior, 25(2), 578–586. factor model. Eth zurich sociology working papers, ETH Zurich. Predictors, 0.37-
Saleem, M. (2010). By the numbers: Facebook vs the United States. <http:// 33.54.
mashable.com/2010/04/05/facebook-us-infographic/> Accessed 05.04.10. Wilson, K., Fornasier, S., & White, K. M. (2010). Psychological predictors of young
Srinivasan, A. (2009a). Ten interesting facts about Facebook. <http:// adults’ use of social networking sites. Cyberpsychology, Behavor, and Social
techcrunchies.com/ten-interesting-facts-about-facebook/> Accessed 04.12.09. Networking, 13(2), 173–177.
Srinivasan, A. (2009b). Market share of top five social networks in USA. <http://
techcrunchies.com/market-share-of-top-five-social-networks-in-usa/>
Accessed 04.12.09.