It Customer Service
Resume:
I regretted the minute I pressed share :
A Qualitative Study of Regrets on Facebook
Yang Wang Saranga Komanduri Pedro Giovanni Leon
abpoxu@r.postjobfree.com abpoxu@r.postjobfree.com abpoxu@r.postjobfree.com
Gregory Norcie Alessandro Acquisti Lorrie Faith Cranor
abpoxu@r.postjobfree.com abpoxu@r.postjobfree.com abpoxu@r.postjobfree.com
Carnegie Mellon University
Pittsburgh, PA
ABSTRACT of wine and a mug of beer [14]. These incidents demonstrate the
negative impact that a single act can have on an SNS user.
We investigate regrets associated with users posts on a popular
In order to protect users welfare and create a healthy and sus-
social networking site. Our ndings are based on a series of in-
tainable online social environment, it is imperative to understand
terviews, user diaries, and online surveys involving 569 Ameri-
these regrettable actions and, more importantly, to help users avoid
can Facebook users. Their regrets revolved around sensitive top-
them. In the large body of SNS literature, little empirical research
ics, content with strong sentiment, lies, and secrets. Our research
has focused on the negative aspects of SNS usage. We try to ad-
reveals several possible causes of why users make posts that they
dress that gap by examining accounts of regrettable incidents on
later regret: (1) they want to be perceived in favorable ways, (2)
Facebook collected through surveys, interviews, and user diaries.
they do not think about their reason for posting or the consequences
With more than 600 million users, Facebook has become the
of their posts, (3) they misjudge the culture and norms within their
world s largest social networking site (according to Alexa, as of
social circles, (4) they are in a hot state of high emotion when
August 3, 2010, Facebook has the highest traf c among all SNS
posting, or under the in uence of drugs or alcohol, (5) their post-
sites in the US [6]). While well-evolved norms guide socialization
ings are seen by an unintended audience, (6) they do not foresee
and self-disclosure in the of ine world, in the online world it can
how their posts could be perceived by people within their intended
be more dif cult to identify one s audience, control the scope of
audience, and (7) they misunderstand or misuse the Facebook plat-
one s actions, and predict others reactions to them. As a conse-
form. Some reported incidents had serious repercussions, such as
quence, Facebook users might not always anticipate the negative
breaking up relationships or job losses. We discuss methodologi-
consequences of their online activities, and end up engaging in ac-
cal considerations in studying negative experiences associated with
tions that they later regret.
social networking posts, as well as ways of helping users of social
Since they are common experiences that people can recognize
networking sites avoid such regrets.
and describe, we use regrets as an analytic lens to investigate users
negative experience with a social networking site. In all studies pre-
Keywords sented in this paper, we asked our participants about things that they
posted on Facebook and then regretted. Since one of our goals was
Social media, Facebook, regret, privacy
to understand how Facebook users think about regret, we used the
word regret without de ning it, and left the interpretation to our
Categories and Subject Descriptors
participants. In doing so, we sought to give voice to participants
H.5.m [Information Interfaces and Presentation]: Miscellaneous own ways of understanding regrets and related concerns. After an-
alyzing our participants responses, we can summarize regret as a
1. INTRODUCTION feeling of sadness, repentance, or disappointment over one s own
actions and their factual or potential consequences.
As social networking sites (SNSs) gain in popularity, stories of
While regrets in the real world have been studied extensively
regret continue to be reported by news media. In June 2010, a per-
(e.g., see [44] for a meta-analysis), little work has investigated re-
ogie mascot for the Pittsburgh Pirates was red because he posted
grets in online contexts. Our work takes a rst step into examining
disparaging comments about the team on his Facebook page [13].
people s regrets in social media in general, and Facebook in partic-
More recently, a high school teacher was forced to resign because
ular. We identify different kinds of regrets, analyze their causes and
she posted a picture on Facebook in which she was holding a glass
consequences, and examine users existing coping mechanisms.
1.1 Related Work
Copyright is held by the author/owner. Permission to make digital or hard In this section, we review related work on privacy and social
copies of all or part of this work for personal or classroom use is granted networks. We focus on work that identi es risks, studies privacy
without fee.
concerns and practices, and details coping strategies for reducing
Symposium on Usable Privacy and Security (SOUPS) 2011, July 20 22,
privacy risks.
2011, Pittsburgh, PA USA
1
1.1.1 Privacy Risks Users display more concern about sharing with their weak-tie
friends than with outsiders or companies. Stutzman and Kramer-
Previous work has identi ed potential privacy risks associated
Duf eld suggest that users adopt friends-only pro les mainly to
with social networks. Gross and Acquisti rst highlighted risks
deal with unintended disclosure to their weak ties rather than out-
such as stalking, identity theft, price discrimination, or blackmail-
siders [48]. Raynes-Goldie found that users cared more about pro-
ing [24]. In addition, Boyd and Ellison identi ed privacy risks
tecting information from members of various social circles, rather
such as damaged reputations, unwanted contacts, surveillance-like
than protecting their information from companies [41]. Besmer
structures due to backtracking functions, harassment, and use of
and Lipford found most photo privacy concerns were about iden-
personal data by third-parties [10].
tity and impression management within the user s social circle.
Skeels and Grudin studied SNS use in the workplace and identi-
These photo-privacy concerns revolve around revelation of incrim-
ed tensions due to the mixing of users personal and professional
inating evidence (e.g., underage drinking), un attering photos, and
circles (for example, the crossing of hierarchy and/or power bound-
unwanted associates (e.g., ex-signi cant others) [8].
aries)[47]. When users actions on SNS make these risks material-
Social in uence seems to play a role in privacy attitudes. In a
ize or lead to negative consequences, they may experience regret.
2008 study of Facebook users at Harvard University, Lewis et al.
found that students are more likely to have private pro les if their
friends and roommates have them [26]. In a 2009 study of privacy
1.1.2 Privacy Concerns and Practices
settings on Flickr, Nov and Wattal found that sense of trust and
A signi cant portion of previous research on risky SNS behavior
sharing norms of a community positively affect community mem-
has focused on users privacy concerns and practices. These studies
bers privacy concerns and information sharing behavior [36].
shed light on the types of SNS activities that may cause negative
outcomes and thus can inform our study of regrets associated with 1.1.3 Coping Strategies to Reduce Risk
SNS activities.
A number of strategies to counter or defuse privacy risks have
Findings from early empirical studies of student Facebook users
been identi ed in the literature. Lampinen et al. found users di-
in the United States suggest behavior inconsistent with stated pri-
viding the platform into separate spaces, using suitable channels of
vacy concerns, excessive sharing of personal data, and rare changes
communication, and performing self-censorship [31]. Similarly,
to default privacy settings. In a 2005 study of Facebook users at
in Lampe et al. s study, some users reported active management of
Carnegie Mellon University, Gross and Acquisti found that the ma-
their pro le, e.g., restricting who can see it and removing sensi-
jority of users shared a large amount of personal data, and only a
tive content. Their interview respondents reported incidients of
very small percentage of users changed their default privacy set-
minor embarrasement but did not report any strong negative con-
tings [24]. In a follow-up 2006 study, Gross and Acquisti found
sequences [30]. Tufekci analyzed college students information
that even users who claimed to be concerned about privacy tended
disclosure behaviors on social networks and found that students
to reveal a great deal of their personal information a discrepancy
manage unwanted audience concerns by adjusting pro le visibility
between stated privacy attitude and actual behavior [5]. Ellison et
and using nicknames but not by restricting the information within
al. found that only 13% of Facebook pro les in the Michigan State
the pro le [51]. In a year-long ethnographic study of Facebook
University network were restricted to friends only in 2007 [18],
users in their 20s, Raynes-Goldie found various strategies includ-
and this was con rmed by a longitudinal study of the same popula-
ing using aliases, deleting wall posts, untagging photos, and cre-
tion from 2006 to 2008 [30].
ating multiple accounts to circumvent Facebook s default privacy
However, these studies only examined American college stu-
settings [41].
dents, and the results might not generalize to other populations.
For instance, Joinson conducted a study with primarily non-student 1.2 Research Questions
Facebook users in the U.K. in 2008 and found that the majority of
Most previous work focused on users privacy attitudes and use
the respondents (57.5%) reported having changed the default pri-
of privacy settings. Very little is known about what actually goes
vacy settings [25]. We conducted online surveys and in-person in-
wrong in users SNS activities and what causes regrets to occur.
terviews to examine a broader segment of the population than just
We chose to directly investigate regrets on SNSs and their causes,
students, but did not include participants outside the United States.
with the ultimate goal of designing counter-measures to help users
More recent studies seem to suggest that users are becoming
avoid them. In this paper, we aim to answer the following research
more privacy concerned and more likely to change their privacy
questions:
settings [33, 11]. For instance, according to a 2009 report by the
What posts do users regret sharing on Facebook?
Pew Internet & American Life Project, 71% of SNS users between
the ages of 18 and 29 reported changing their privacy settings [33].
Why do users make regrettable posts?
Demographics seem to affect privacy attitudes and behavior. Fo-
gel and Nehmad found that in general men had less privacy con- What are the consequences of these regrettable posts?
cerns than their female counterparts, and thus tended to disclose
How do users handle regrettable posts?
more personal information such as telephone numbers and physi-
cal addresses on SNSs [20]. Stutzman and Kramer-Duf eld found
How do users currently avoid regrets?
that female users and users who have more Facebook friends are
more likely to have friends-only pro les [48]. In a study of MyS-
2. METHODOLOGY
pace users, Gilbert et al. found that rural users have fewer friends
and comments than urban users. Further, rural users, particularly We rst analyzed reader comments on a New York Times arti-
women, have a higher level of privacy concern and use privacy set- cle about Facebook privacy [2] and developed a survey to probe
tings more often than urban users [21]. boyd and Hargittai also whether the concerns expressed in those comments were typical of
found that individual characteristics such as Internet skill, frequency, American Facebook users. After analyzing the results from that
and type of Facebook use are correlated with making modi cations survey, we conducted semi-structured interviews to ask in-depth
to privacy settings [11]. questions about users experiences on SNSs.
2
While the interviews capture the most memorable experiences sponses. There were 117 male respondents (36.4%) and 204 female
of the interviewees, we also wanted users daily, often mundane respondents (63.6%). The average age of respondents was 31 years
old ( =11.0).
Facebook experiences which they might forget or take for granted.
We also hoped to explore how regrets might affect users subse-
2.2 Interviews
quent behavior on Facebook. For these reasons, we designed a di-
ary study and invited the interviewees to log their daily Facebook
2.2.1 Questions
experiences for a month. These studies raised additional questions
The semi-structured interviews included open-ended questions
about regrets on Facebook, and we conducted another online sur-
about users motivations and use of Facebook, privacy attitudes to-
vey to gain further insights.
ward Facebook usage, experience with Facebook s privacy settings,
While Facebook s user population is quite diverse, the major-
and their own and their friends regrettable experiences on SNSs.
ity of prior research was conducted with college students. Our re-
For instance, one important question was Have you ever posted
search seeks to gain a more comprehensive understanding of the
something on a social network and then regretted doing it? What
SNS user population by studying American Facebook users from a
happened? We include all interview questions in Appendix A.
wider range of ages and occupations. We recruited survey partici-
pants using the Mechanical Turk crowd sourcing site and recruited
2.2.2 Procedure
interviewees from the Pittsburgh Craigslist website. We report on
We used Craigslist2 to recruit Facebook users in Pittsburgh, Penn-
two surveys in this paper, and refer to them as survey1 and sur-
sylvania to come to our lab for for in-person interviews. A pre-
vey2. Our studies were approved as minimal risk studies by our
questionnaire was used to screen and balance our interview par-
university s Institutional Review Board (IRB).
ticipants across age, gender, occupation, and frequency of Face-
2.1 Survey 1 book usage. 301 people completed the questionnaire by January
21, 2011. Most of them were college students and the majority
On May 6, 2010, the New York Times Bits Blog posted an arti-
of them were women. We selected 19 users from this pool to in-
cle titled, Ask Facebook Your Privacy Questions. This blog entry
terview (10 males and 9 females). Their ages ranged from 18 to
invited readers to submit their questions to Elliot Schrage, Face-
56 ( =33.0, =13.0). The interviewees came from diverse occu-
book s Vice President for Public Policy, in the blog entry s com-
pations: student, attorney, social worker, artist, telemarketer, man-
ments section [2]. We analyzed the responses from 268 users and
ager, nancial service staff, small business owner, teacher, career
identi ed three main themes: (1) people feel strongly that certain
coach, and unemployed. Of the 19 participants, 8 of them visited
types of information are private, (2) people are frustrated with Face-
Facebook multiple times per day, 8 of them visited Facebook about
book s privacy controls, and (3) people have strong views against
once per day, and the remaining 3 users visited Facebook less than
Facebook s opt-out model regarding sharing users information with
once per week.
third-parties.
We conducted the interviews from August 2010 until January
Based on these ndings, we designed a survey to further investi-
2011. Each interview took approximately 1 to 1.5 hours. Inter-
gate users privacy-related attitudes and behaviors on Facebook.
viewees were asked to log into their Facebook accounts to have
2.1.1 Questions a conversation with the researchers about their Facebook experi-
The online survey was designed to gain a better understanding of ences. From time to time, the researchers took screen shots of in-
Facebook users privacy related experiences and behavior on Face- terviewees Facebook pages when participants consented. Each in-
book. In this paper we only focus on the responses to the question: terviewee was paid $20 as compensation. Interviews were audio
Have you ever posted something on a social network and then re- recorded and then transcribed. One author coded the interviewee
gretted doing it? If so, what happened? data and categorized it post-hoc into a list of common themes.
2.3 Diary Study
2.1.2 Procedure
We recruited adult Facebook users in the United States using
2.3.1 Questions
Amazon s Mechanical Turk1 (MTurk), a crowd sourcing service
The diary study consisted of questions that collected data about
that is gaining popularity for use in HCI research [28]. We pre-
the user s daily experience on Facebook. The diary asked if the
sented our survey as a task on MTurk and asked MTurk users to
user had accepted or rejected any friend requests, what activties the
follow a link to our survey on the SurveyGizmo commercial web
user conducted, if changes had been made to pro les or privacy
survey service. MTurk assigns each of its users an anonymous ID
settings, if the user had any regrets that day, and other questions
and we made sure that each MTurk ID only answered our survey
about positive or negative experiences on Facebook. The regret
once. We paid each participant $0.50 for completing the survey.
question we asked was Have you posted something on Facebook
Since the survey would take about 10-15 minutes to nish, our
and then regretted doing it? Why and what happened? We include
compensation rate was about $2-3 per hour, which is on par with
the list of questions in Appendix A.
the normal hourly pay on MTurk.
As suggested in the literature [28, 16], we used a combination of
2.3.2 Procedure
measures to help determine whether our participants from MTurk
Participants from the interview study were invited to participate
were taking their tasks seriously. We paid attention to unusually
in the diary study at the end of their interviews. Twelve intervie-
short completion time (two standard deviations from the mean), in-
wees actually participated for at least one day. We asked the par-
consistent answers (we asked the same underlying question with
ticipants to answer the same set of questions in a web form every
slightly different wordings), and verbatim repetition or nonsense
day for a month. Participants who lled out the form for 22 or
free-response answers. If we found two suspect answers, we then
more days received $15 as compensation. Two hundred and seven-
excluded the participant s results. After manually checking the an-
teen days of diary logs were entered by Sep. 15, 2010. As with the
swers and ltering out suspect participants, we had 321 valid re-
1 2
http://mturk.amazon.com http://pittsburgh.craigslist.org/
3
interview data, one author coded the diary entries and categorized To protect the privacy of our research participants and to dif-
them by common themes. ferentiate between studies, we use anonymous identi ers. The 11
participants in the interview and diary studies are denoted with P#.
2.4 Survey 2 For instance, we use P1 to represent the rst interviewee (and diary
participant). Survey respondents are not identi ed by number. In-
Based on results from the preceding online survey and subse-
stead, we specify which survey the data was from when we report
quent interviews, we designed a survey to focus on speci c aspects
it, e.g., a survey1 respondent said . . . .
of regrets. In survey1, the interview study, and the diary study, we
did not focus solely on users who had regrets on Facebook. For
3.1 What Do People Regret Posting?
this survey, however, we asked people to take our survey only if
In this section, we focus on participants responses to questions
they had posted something on Facebook and later regretted it.
of the form: Have you posted something on Facebook and then
2.4.1 Questions regretted doing it? If so, what happened?
The second survey contained 34 questions. We began by ask-
3.1.1 Sensitive Content
ing survey participants Have you ever regretted posting some-
Our participants reported several types of sensitive content that
thing (status updates, pictures, likes, comments, locations, etc) on
they regretted posting. We loosely categorize that content here. In
Facebook? For example, have you ever posted something that you
some cases, e.g., illegal drug use, merely posting this content is
felt bad about later or wished you hadn t posted? We then asked
enough to cause regret. In other cases, sensitive content can be part
how many times they regretted posting on Facebook in the last 12
of a deeper cause of regret. For example, we nd that profanity can
months. In order to help participants recall speci c details about
sometimes be offensive on its own or it can be used to insult others.
their regrets, we asked them to think about the one posting that they
regret the most and then answer the following questions with re-
Alcohol and Illegal Drug Use
spect to that post. We then asked the participants several multiple-
Many participants regretted posts about drinking. One survey2
choice and open-ended questions to learn about their post, speci -
respondent said, I posted photos from a party that got a bit out of
cally: why the post was made, what happened after the post, when
hand, and the photos were not very attering. What bothered me
the regret occurred, the reason(s) they regretted the post, how much
was that I realized I posted them and my pro le was public and
they regretted it, and what they did in response to the regret. We
other people could see them. He then explained why he posted
also asked about the participant s mood when he or she posted the
them: . . . out of habit; after an event with friends most of us post
regrettable content (e.g., very happy or sad) and whether they were
the photos. This quote suggests that the culture and norms of a
under the in uence of drugs or alcohol. The list of survey questions
person s social circle play a role in one s decision to post. In this
can be found in Appendix A.
case, most of the participant s friends post event photos.
2.4.2 Procedure If such posts are the norm, why did this participant regret it?
He said, I realized they weren t something I wanted other peo-
As with survey1 (see Section 2.1), we hosted this survey on Sur-
ple to see that didn t know me, because they d get the wrong idea.
veyGizmo and recruited adult Facebook users in the United States
This highlights the issues of unintended audience (in this case, peo-
who had regrets on Facebook. The survey was deployed on MTurk
ple who did not know him) and impression management. He felt
for about one week in early March 2011. We paid each participant
uncomfortable because these photos might lead to a particular im-
$0.50. We followed the same methodology to detect suspicious
pression that violates how he wants himself to be perceived by oth-
responses as in survey1 (see Section 2.1.2). After ltering out sus-
ers. He also said, one person asked me to remove the tag of their
picious respondents and those who did not report any regrets on
photo. These posts can also violate others self-representations.
Facebook, we had 492 valid responses. There were 216 male re-
Some regrettable posts mentioned illegal drugs. One survey2
spondents (43.9%) and 276 female respondents (56.1%). The aver-
respondent said, I regretted posting a picture of me smoking mar-
age age of respondents was 28 years old ( =8.6). Compared with
ijuana at a party. People in my family seen it and other people I
the sample of survey1, the sample of this survey was younger and
didn t want seeing it. He posted it because I thought it was cool
more male.
at the time. I had an I didn t care attitude. He regretted posting be-
cause it embarrassed others: Certain people around me give me a
3. RESULTS sense of disapproval when I was around them. My mom for exam-
ple told me it was embarrassing for her. Sometimes just a mention
The results that we report below include data from the interviews
of drugs can cause trouble. One survey1 respondent said, I posted
and user diaries as well as answers to several regret-related open-
a music video of a song called I m just a girl that you lost to co-
ended questions in survey1 and survey2. As with the interview
caine and my parents were concerned I was using drugs, but they
data, we coded the free responses from the two surveys and catego-
never listened to the lyrics of the song, which were actually about
rized them post-hoc to produce a list of common themes.
a girl leaving a boy who was addicted to drugs.
Our initial study was a three-part study consisting of a survey,
The consequences of these posts can sometimes go beyond a
and interview/diary study. For these initial studies, we recruited
problem of image. One survey2 respondent said It was a photo
Facebook users regardless of whether they had any regrets. Some
that had underage drinkers in it. I thought no one in authority would
of our studies gather data on both the regrets of our study partici-
see it . . . [as a result] some one lost their job from it. Photos in-
pants ( rst-party) and the regrets of friends of our study participants
volving underage drinking were a common source of regret among
(third-party). We had a total of 340 participants from these initial
our participants.
studies including 321 survey respondents, and 19 participants in the
interview/diary study. We found that 66 out of 321 survey respon-
Sex
dents (21%) and 11 out of 19 (58%) interview/diary participants
reported having rst-party regrets. For the remainder of the paper Posting sexual content was another common issue. P7, an artist
we discuss only those participants who reported rst-party regrets. in his forties, told us about a conversation (a status update and then
4
comments) he had with a friend on Facebook: She said something Sometimes profanity causes problems of impression manage-
like whenever the divorce is nalized, I should come out to New ment. One survey2 respondent said, Posting anything with a swear
Mexico and we should have a party. And I said what will the party word in it now that I m friends with my family. I ve done this a cou-
entail. And she said I was thinking of a lot of alcohol. And I said I ple times, and when you do it from your phone, you can t delete
was hoping for [sex]. He then told us at the time he did not think right away! He attributed this to a spur of the moment decision
about it, but in hindsight he would not want his mother or young and explained, It s inappropriate for my family. ok for friends, but
Facebook friends to see it. not family or church friends.
Sometimes, people accidentally post sexual content. One sur-
vey2 respondent said, I accidentally posted a video of my hus- Personal and Family Issues
band and I having sex . . . I didn t mean to post it, I had accidentally Sometimes people share their personal issues to gain support,
clicked on the video of my daughter taking her rst steps and on but it is tricky to balance how much to share and how much to keep
that video and they both uploaded together . . . I didn t know I had private. One survey2 respondent said, The status update included
posted it until the day after, when I logged on again, and saw all a curse and it described a medical condition I was experiencing.
the comments from all of our friends and family, and my husbands I had broken out in hives and I was posting about the discomfort.
coworkers (he s in the army). She regretted posting because it I don t think all my friends needed to know about the itchiness
was a personal video between my husband and I. In this case, the and swelling. It was just complaining to everyone . . . I regretted
posting was an accident, and not a result of failing to foresee con- oversharing.
sequences. Another survey2 respondent said, I posted that I was no longer
single and I was dating this guy in my class . . . I was happy and
Religion and Politics excited about myself . . . People read it and told my parents and they
People can specify their religious or political beliefs in their did not approve. This shows that people sometimes post things
Facebook pro les. However, posts that express these beliefs can when they are in an extremely positive mood that they later regret.
cause debates, offend people, and damage relationships. One sur- On the other hand, sometimes family issues are brought up when
vey2 respondent said, [I posted] my beliefs about religion. Be- in a negative mood. One survey1 respondent wrote, I did post
cause my name was also tied to my business, people who disagreed something about a ght with my husband once and regretted it after
with my beliefs about religion took action against my business he saw it and was offended that I was airing our dirty laundry for
. . . My business was given bad online reviews. Another survey re- everyone to see.
spondent said, I got in a religious debate on Facebook. I did delete
my comments but several people dropped me as their friend. Work and Company
P5 is a volunteer in a local church. When we were going through Our participants also reported regrets caused by posting about
his photos on Facebook, we saw pictures of some people in his their work or company in a negative way. One survey2 respondent
church getting baptized. He told us: . . . I posted some of the pho- said, When I badmouthed my job due to disciplinary I was on for
tos and tag her name, later on either the tags were removed by them- b.s. stuff. My managers are my friends on facebook and ended up
selves, or some of the people just warned against me on this, that ugly at work. He then explained, I was mad . . . I said it out of
is if you want to post them online you d better ask for their permis- anger and not thinking.
sions. Later, P5 noted in his diary that originally I set the privacy
3.1.2 Content with Strong Sentiment
settings on photos as friends of friends, but I think it doesn t make
sense, so I switched it to friends . Participants reported that they regretted posting strongly nega-
Another reason why people post about religion or politics is be- tive or offensive comments as well as in engaging in arguments on
cause they want to share their opinions. But sharing one s religious Facebook.
or political belief can be perceived as pressuring others to have the
same belief. For instance, one survey2 respondent said, I agreed Negative or Offensive Comments
with a political statement made by a friend and reposted it on my People often post negative content because they are in a bad
own status . . . because at the time I agreed with it. Even though I mood, and we heard many accounts of regret due to angry posts.
agreed with it, I partly regretted it because making statements about One survey2 respondent said, posted a negative comment to a man
religious or political things are a ne line. I have my beliefs but I I care about . . . emotions high with frustrations lashing out at him
would never want my friends or family to think I was trying to force when I should instead be more in control . . . I regret hurting him
my beliefs on them. I was afraid some of them might think that. especially in writing when I can t change it later. No back button
or undo. It hurts to hurt him so I regret doing it.
Profanity and Obscenity As mentioned previously, bashing one s company or employer
Postings with profanity or obscenity can be a cause of regret. can also be troublesome. Another survey2 participant wrote, I
One survey2 respondent said, I said something along the lines of posted negative comments on the fan page of the company I work
Hey Bob at ST, stop treating us women like trash . . . fuck you! for thinking it would be anonymous, but they looked at my page
The profanity is often a result of the users mood at the time when and saw the franchise owner in my friends list and called him and
they posted the content. In this case, the respondent explained, I told him. I got in huge trouble, fortunately I kept my job and was
posted it because I was very angry. He is a customer at my place of able to smooth things over. I was making comments that attacked
business and hates women . . . I was only venting my frustration. the way the company hurts its dealers nancially as well as our abil-
Another survey2 respondent reported, It was a status update that ity to give adequate customer service. I had also posted a picture of
said that I hated someone I used to love very much in the past. It a friend/customer in our store on my pro le, my wife posted HA-
said word by word, I fuckin hate you! You will never be loved HAHAHAHAHAHAHAHA on it and they assumed I was making
again you anorexic piece of shit! We can sense his emotion as he fun of customers. To try and get them to improve the way they
explained why he posted it: I posted it because I was upset at [her] treat us and to improve our ability to help our customers. I was
because she broke up with me. put under a microscope for months, and almost lost my job, and
5
could have lost the job of my boss for him. This example illus- and skip school. I was trying to be cool . . . One of my teachers and
trates how interaction in the online social environment can cause my family members saw it. I got kicked out of the house because I
unforeseen problems. Here, the wife s comment about their friend was already on thin ice.
was misinterpreted and made the situation even worse.
It s Funny
Arguments Trying to be funny is another source of regret when what was
One survey2 participant wrote, The post I regret the most was thought to be funny turns out to be offensive. One survey2 respon-
a ght I was having with my 25 year old niece. We had a dis- dent wrote, My post was about the Border Patrol not doing their
agreement in her status, and we went back and forth for a long time job. I was trying to make an interesting event sound funny. One
arguing our points. Although we both took it too far, I feel I could of my friend s husbands is an agent and [my friend] was very of-
have stopped it up front. I was angry enough to want the world to fended. Another survey2 respondent said, It was a picture of a
see that what she was saying didn t make any sense. Afterward, I friend. The person was posing for a picture at a restaurant. The
just regretted it all. I regretted posting it because it didn t end up person did not look the best and later was upset that I had posted
making any difference. My niece still believes she was right, and it because she said it made her look fat. I thought it was a funny
many other ghts broke out because of it. No one contacted me picture and would get some laughs. I hurt my friend s feelings after
about the post because there was no swearing in it. I ended up tak- she told me how she felt.
ing my niece off my friend list though. At the time of posting, the
participant wanted the argument to be out in public, but she later Venting Frustration
regretted the sentiment. Users in a highly emotional state often vent their feelings on
Facebook. A survey2 respondent wrote, I posted something about
3.1.3 Lies and Secrets my feelings about an argument I had with a friend. I didn t mention
Telling lies and revealing the lies of others are another source of her by name but it was fairly obvious to those who knew about the
regret. In some cases, such as when posts reveal illegal activities argument who I was referring to. I felt the need to vent and get the
such as underage drinking, the consequences are serious. In many situation off of my chest. Also, I m sure a small part of me wanted
other cases the consequences are less severe, leading to misunder- her to read it and feel bad. Like the argument mentioned in Sec-
standing and the need for dif cult explanations. tion 3.1.2, users want to express their frustration in a public forum,
One survey1 respondent said, As an April Fool s joke, I changed though they sometimes regret doing so.
my relationship status to Married . My much younger cousin didn t
get the joke and told my aunt I d eloped. A small misunderstand- Good Intentions
ing, but sort of awkward to explain to my parents. A survey2 Sometimes regrettable posts are made with the best of intentions.
respondent said, I uploaded various photos I took while at a party One survey2 respondent said, I posted something about a friend
with friends. These photos varied from group pictures, drunk pic- who had gained a lot of weight recently. I hadn t seen her in a long
tures, and pictures of drinking games. I posted it to share the pic- time and I just thought my friend was pregnant at the time I posted
tures I took that night with the people in them and with friends that it. I was congratulating her on her upcoming pregnancy. So I asked
weren t able to attend. The photos I uploaded got a friend in trouble if she was pregnant and she told me no, she had gained a lot of
by catching him in a lie. He promised someone that he wouldn t weight. I felt horrible.
drink that night, but a few photos show him with a beer in his hand. Another survey2 respondent wanted to provide useful informa-
Although he never told me to not upload them, I felt bad that I was tion but then was misunderstood. He said. [I] made a location
the reason he got caught. I found out later that through a mutual check in at a club with some friends . . . to let a friend we were
friend that the photos caused our friend trouble. I then went to ask waiting for know we arrived. The boyfriend of one of my friends I
him about it and apologized. was with thought she was cheating on him with me and they started
to argue. He called me and started to yell that I was stealing his
3.2 Why Do People Make Regrettable Posts? girl. He then broke up with my friend, his girlfriend.
In this section, we consider the reasons why Facebook users
make regrettable posts. We rst describe the intended purposes of I Didn t Think about It
the posts, and then we explore why they turned out to be problem- When posting on Facebook becomes habitual, people rarely think
atic and led to regret. about why they post things. The following survey2 respondent s
story is telling: . . . I was so addicted to facebook! It s like an in-
3.2.1 Intended Purposes voluntary action. You feel something and you express that in face-
In many instances, users report that they had no speci c purpose book.
for posting. In others, they explain the reason behind their posts in Some users also did not think about the potential consequences
order to explain their regrets. We categorize and explain commonly of their postings. One survey2 respondent reported posting a photo
reported reasons here. of his underage friend getting drunk and tagging him in it: I didn t
think his parents would see it, and I didn t think about any of the
It s Cool consequences at the time. Another respondent said, I regret post-
Some people reported wanting to be perceived as interesting or ing a joke a friend told me, that could be thought of offensive to
unique. However, when the content or behavior described in the women. I posted it because I didn t think about who would read it,
post was controversial, this caused regret. One survey2 repondent and I thought it was funny. I think I offended many of my friends.
said, I posted a photo of me smoking hooka and got in trouble
with it from my employer . . . at the time I thought it was cool. I
lost my job because of it. My boss talked to me about it and told
me they did not want that image in the company. Another survey2
respondent wrote, I said that I was going to pretend that I was sick
6
3.2.2 Hot States book: My wife didn t pay spousal support . . . she posted on her
Facebook that she got a job from somewhere. I took a screen shot
Users often regret things they posted while in a highly emotional
of that
1.1 Related Work
Copyright is held by the author/owner. Permission to make digital or
Contact this candidate