All about me: Disclosure in online social networking profiles: The case of FACEBOOK

Computers in Human Behavior 26 (2010) 406–418

Contents lists available at ScienceDirect

Computers in Human Behavior

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

All about me: Disclosure in online social networking profiles: The case of FACEBOOK

Amanda Nosko *, Eileen Wood, Seija MolemaDepartment of Psychology, Wilfrid Laurier University, 75 University Avenue, Waterloo, Ont., Canada N2L 3C5

a r t i c l e i n f o

Keywords:FACEBOOK™OnlineSocial communicationSocial networkingDisclosurePrivacyInternetThreat

0747-5632/$ - see front matter � 2009 Elsevier Ltd. Adoi:10.1016/j.chb.2009.11.012

* Corresponding author. Tel.: +1 519 954 4650.E-mail address: [email protected] (A. Nos

a b s t r a c t

The present research examined disclosure in online social networking profiles (i.e., FACEBOOK™). Threestudies were conducted. First, a scoring tool was developed in order to comprehensively assess the con-tent of the personal profiles. Second, grouping categories (default/standard information, sensitive per-sonal information, and potentially stigmatizing information) were developed to examine informationpertinent to identity threat, personal and group threat. Third, a grouping strategy was developed toinclude all information present in FACEBOOK™, but to organize it in a meaningful way as a function ofthe content that was presented. Overall, approximately 25% of all possible information that could poten-tially be disclosed by users was disclosed. Presenting personal information such as gender and age wasrelated to disclosure of other sensitive and highly personal information. Age and relationship status wereimportant factors in determining disclosure. As age increased, the amount of personal information in pro-files decreased. Those seeking a relationship were at greatest risk of threat, and disclosed the greatestamount of highly sensitive and potentially stigmatizing information. These implications of these findingswith respect to social and legal threats, and potential means for identifying users placing themselves atgreatest risk, are discussed.

� 2009 Elsevier Ltd. All rights reserved.

1. Introduction

Based on the surge in online communication, researchers havebegun to explore self-disclosure online (e.g., Christofides, Muise,& Desmarais, 2009; Henderson & Gilding, 2004; Joinson, 2001; Tid-well & Walther, 2002). Relatively little is known about how peopleuse online social networking sites. Recent studies have begun toexamine the use of online technology and the associated attitudesand behaviors that surround online communication (e.g., Ellison,Steinfield, & Lampe, 2007; Walther, Van Der Heide, Kim, Wes-terman, & Tong, 2008; Zhao, Grasmuck, & Martin, 2008). However,research in this area is sparse and even fundamental issues, such asaccurately capturing the kinds of information that are availablethrough online communication profiles or understanding patternsof information disclosure, are not yet fully defined. Before we canbegin to understand why online communication is used or deter-mine the purpose of this communication tool, it is first importantto determine what can be communicated when using this tool.The present study provides a comprehensive examination of theinformation of the potential content available through the FACE-BOOK™ online social networking profiles and, in addition, beginsto explore means for understanding what information is mostlikely to be disclosed and by whom.

ll rights reserved.

ko).

Personal web-pages and online profile networks have emerged atan increasing rate with social networking sites, including FACE-BOOK™, gaining rapid popularity (Yum, 2007). FACEBOOK™ is asocial communication tool designed to allow users to contactand communicate with other FACEBOOK™ users (www.FACE-BOOK™.com). Founded in 2004 and originally designed as a socialnetworking site for students at Harvard University, FACEBOOK™ isnow available to anyone who is older than 13 years of age. Currently,the site has over 61 million active users (Statistics, FACEBOOK™,2008).

The FACEBOOK™ infrastructure is made up of a variety of net-works that are based around companies, schools or geographicalregions. Privacy settings can be used to control access to personalprofiles, such that only designated friends and users within sharednetworks can be provided access to profiles. For those users that donot employ privacy settings, however, profiles are accessible to anyFACEBOOK™ member. Once a member, users can search for friends(these can be actual friends, acquaintances or even strangers) onFACEBOOK™, and add them to their ‘‘friends list”. Users can alsoupload photos, describe interests, work, education history, rela-tionships, personal stories, schedules and more. Users can searchfor friends by typing their names into the FACEBOOK™ searchbar. Users can interact with one another. Each FACEBOOK™ profilehas an email inbox, where other users can contact them and chooseto send them a message. These messages appear on the profilepage. In addition, profiles include recent activity, and a ‘‘wall”where other users can post messages and attach links for other

http://www.facebook.com



http://dx.doi.org/10.1016/j.chb.2009.11.012

mailto:[email protected]

http://www.sciencedirect.com/science/journal/07475632

http://www.elsevier.com/locate/comphumbeh

A. Nosko et al. / Computers in Human Behavior 26 (2010) 406–418 407

sites, videos, or photos. Groups can be created on FACEBOOK™, andusers may be invited to join. Overall, the system provides manymeans for communicating with others.

Given the vast array of information that can be shared and thenumber of users, concerns regarding security and privacy issuesare a recurring issue (Acoca, 2008). Some concerns involve poten-tial threats to personal safety from the abundance of informationthat is assumed to be available and accessible about an individualon their online profile. Specifically, there are concerns regardingidentity theft if users provide too much information (e.g., birthdate, address, phone, full name etc.). In addition, there are concernsfor personal safety for vulnerable users who could be stalked, orotherwise threatened. A less commonly considered threat is thepossibility of social risk as a function of self-identification withminority or stigmatized groups. Although some of these concernshave surfaced in the popular media (e.g., news.cnet.com), there islittle empirical investigation documenting how much and whatkind of information is present in personal FACEBOOK™ profilesto determine the potential for threats of any type, nor is thereany information regarding how users differ in the information dis-closed in their profiles to provide clues as to who is most likely tobe at risk.

1.1. Identity theft and threats to personal security

Identity theft and personal security are ever-present concernsassociated with information disclosed online (e.g., Fogel & Neh-mad, 2009; LaRose & Rifon, 2006; Lee, Im, & Taylor, 2008). Inapparent opposition to the numerous warnings issued by lawenforcement and public awareness groups regarding the need tobe cautious in disclosing personal information (e.g., Child OnlineProtection (COP); McCandlish, 2002; Willard, 2006), one of the pri-mary goals of social networking sites is to encourage disclosure ofpersonal information with others online. This personal informationcan include full names, addresses, birth day and year, contact infor-mation, and photos. Even a select few pieces of personal informa-tion, however, has the potential to provide identity thieves withthe means to acquire ‘‘identity-based” information such as socialinsurance numbers, credit cards, drivers licenses, etc. With thesepieces of information, even more critical legal documents can beprocured, such as passports (Sullivan, 2008). Apart from concernsregarding the protection of identity, disclosure of personal infor-mation (even if limited) can be sufficient, when combined withother Internet based tools such as reverse directory checks, to se-cure home phone numbers, full addresses, age and gender andother information that could leave a person vulnerable (Messmer,2007).

Given the widespread use and potential dangers associatedwith online communication, a thorough understanding of the fea-tures of networking sites, and how people are using these sites iscritical in developing ways to educate users about how to protecttheir information and themselves. The overarching goal of thepresent research was to gain a better understanding of what canbe found in online social networking profiles, specifically, FACE-BOOK™. Apart from collecting data on the kinds of informationusers were choosing to include (and exclude) in their personal pro-files, the study examines the impact of individual characteristics onthe type of information that is likely to be present in an online pro-file (i.e., information that is self-disclosed as a function of charac-teristics including age, gender and relationship status).

2. Study 1

The primary purpose of Study 1 was to develop a scoring instru-ment to summarize what information could be disclosed on FACE-

BOOK™ profiles. To do this, a comprehensive coding tool wasrequired. From this tool, we then examined the frequency for eachpiece of information to determine what is and is not likely to befound in online profiles.

2.1. Method

2.1.1. Participant samplingA sample of 400 randomly selected, accessible, personal profiles

from 8 Canadian FACEBOOK™ networks was collected. The net-works included 4 community networks (Toronto, Vancouver, Char-lottetown and Kitchener) and 4 university networks (MountAllison University, Memorial University of Newfoundland, Univer-sity of Toronto and Wilfrid Laurier University). Networks varied insize, ranging from small to large. Small networks contained up to14,000 profiles (39.2% of the sample), medium networks contained36,000–150,000 profiles (37.2% of the sample), and large networkscontained 720,000 profiles or more (23.6% of the sample). Specifi-cally, 12.3% of the sample came from a network exceeding a mil-lion people, 11.3% from a network of approximately 720,000people, 15.2% from a network of approximately 150,000 people,12.2% from a network of approximately 56,000 people, 9.8% froma network of approximately 36,000 people, 15.9% from a networkof approximately 14,000 people, and 23.3% from a network ofapproximately 5000 people.

Of the 400 personal profiles, 328 indicated gender (116 females,155 males). Additionally, 301 profiles disclosed age. Among the fe-males, ages ranged from 19 to 47 years (M = 22.90 years,SD = 3.91 years). Males ranged in age from 17 to 61 years(M = 23.90 years, SD = 5.03 years). The mean age of males and fe-males who reported both their age and gender did not differ,t(269) = 1.73, p = .09.

2.1.2. ProcedureA scoring tool was developed in order to assess the content of the

personal profiles. Construction of the scoring tool was a multi-stepprocess. First, a blank template of a FACEBOOK™ profile was exam-ined to identify potential pieces of information that could be in-cluded in a profile. Two raters independently created checklistsfrom this blank profile template. Following these independent con-tent analyses, an aggregated 90 item checklist was constructedwhich was comprised of all items from both raters. Six independentcoders then used the aggregated coding checklist to code actual on-line FACEBOOK™ profiles. Each coder noted any items missing fromthe checklist that would be required to code their profiles. Codersmet and discussed content that was not captured by the aggregatedcoding checklist. Following this discussion a final checklist com-prised of 97 dichotomously scored items (i.e., whether the pieceof information was present or absent) and 3 identification items(i.e., username link, the network searched, and the size of the over-all network) was constructed (see Appendix A for the completechecklist and descriptions of items).

Once the coding checklist was established, six raters coded the400 participant profiles. Each coder coded 50–100 profiles. Net-works searched were counterbalanced across coders. Randomselection of profiles was achieved using a randomization featureof the FACEBOOK™ program where 10 profiles are generated atrandom. These 10 profiles would include profiles that were andwere not publicly visible. Coders scanned the 10 profiles and se-lected the first one from the list of 10 that was publicly visible.The search for profiles was continued until all raters coded theirfull allotment of accessible profiles. In order to ensure consistencybetween coders, reliability was conducted on 40 profiles, repre-senting 10% of the data. Percentage agreement was 99%, indicatinghigh inter-coder reliability. Differences were resolved throughdiscussion.

http://www.news.cnet.com

408 A. Nosko et al. / Computers in Human Behavior 26 (2010) 406–418

2.2. Results

Descriptive summaries were calculated for each item on thechecklist that was available among the accessible profiles (seeTable 1 for a summary). Overall, the 15 most consistently dis-closed/provided pieces of information (available on 63% or moreof the profiles) described personally identifying information (i.e.,birth date, gender, profile pictures, photo albums, tagged photosand general photos of the user) as well social connections (i.e.,groups joined, and friends viewable). In addition, education infor-mation (college/university attended) and regular update informa-tion (status, wall and mini-feed) were included. Finally, playfulcommunications such as acceptance of pokes, messages, and gifts,and applications were frequently provided. In contrast, the 15 leastfrequently included items (9% or less) described key personal infor-mation (zip/postal code), phone numbers (both land line and mo-bile), home address, city or town, website and former name. Inaddition, there was also limited amount of information providedregarding some aspects of educational experience (i.e., school mail-box, courses, degree, awards, and room). Finally, optional ‘‘wallfeatures” (i.e., Super Wall and Advanced Wall) market place listingsand Events as described in the mini-feed were also infrequently in-cluded. Of the 97 items that could be dichotomously scored, 26items were disclosed by at least 50% of the sample (see Table 1for summary).

2.3. Discussion

Two important outcomes were apparent from the descriptivesummary of FACEBOOK™ profile content. First, despite the poten-tial for significant disclosure through these online social network-ing profiles, on average, people were choosing to displayapproximately 25% of possible information for other users to view.This clearly indicates either a reticence to invest heavily in devel-oping online profiles, or active decisions to limit disclosure. Inter-estingly, the pieces of information that were disclosed wereneither consistently ‘‘safe” nor ‘‘unsafe”. Specifically, for both themost prevalent and least prevalent pieces of information included,there were highly personal pieces of information. That is, in somecases profiles contained the most salient pieces of information re-quired for identity theft and personal safety threats, includingidentifying pictures, birth date, and regular mini-feed updates.However, equally critical pieces of information such as land andmobile phone numbers and zip/postal code were not readily appar-ent. Although it makes sense that features such as walls might beless frequently present because these features require active crea-tion and generation to establish and maintain, it is surprising to seeinformation that is prompted being selectively included/disclosed.Therefore, users are demonstrating some discretion regardingwhat kinds of revealing information they are willing to share.

In an attempt to understand who is likely to include revealinginformation and what types of information are likely to be pro-duced together, two studies were conducted.

3. Study 2

With a rapid rise in popularity and use, online social networkingsites have introduced new and potentially harmful ways for indi-viduals to access personal information. Identity theft and personalsecurity threats are ever present online, and thus necessitates re-search examining specific kinds of information disclosed onlinethat may pose a particular risk to the user (Acoca, 2008). Two kindsof threats are possible. One deals with identity theft, which occurswhen personal information is used to commit a crime or theft (e.g.,impersonation) without the discloser’s consent and/or knowledge.

Full names, addresses, phone numbers and birth dates are allpieces of information that can be potentially harmful to the dis-closer if used the wrong way. Other potentially harmful informa-tion (e.g., social insurance numbers, credit card information) canbe accessed through indirect means using these core pieces ofinformation. Another kind of threat is social threat or potentialdanger to the self and/or group(s) that an individual belongs to.Information such as contact information, sexual orientation, groupmembership, religious affiliation, and political affiliation are all de-tails that can potentially be used to harm and stigmatize an indi-vidual. In this second study, we explored means for examiningidentity threat, personal and group threats.

3.1. Method

3.1.1. Materials and procedureThree disclosure categories were developed. The first category

reflected personal identity information, the second involved sensi-tive personal information, and the third involved potentially stig-matizing information. Thematic analysis was conducted toconstruct each of the three categories. Refer to Table 2 for categorydescriptives.

3.1.2. Personal identity information (default/standard information)The first category involved revealing basic personal identifying

information, or what was deemed default/standard information.This information was defined as details people might disclose inbanks, schools, jobs etc., but that could be used in potentiallythreatening ways. To develop this category a three stage processwas used. First, two researchers conducted a content analysis ofthe 97 items in the full checklist. The number of items was first re-duced based on frequency. Items that were very infrequently dis-closed (less than 5% of users disclosed the information) wereomitted. Items were further reduced based on whether the itemdid or did not reveal default or standard personal information thatis typically required to identify a person. Items that failed to meetthis requirement were omitted. Next, 18 research assistants withexperience working in psychology labs examining technologyand technology applications in daily life, were recruited to partic-ipate in a pilot study to construct the final category of standard/de-fault information. First, these research assistants were asked abouttheir familiarity with FACEBOOK™ through two questions. Specif-ically they were asked to use a 7-point Likert-type scale (1 = not atall familiar with FACEBOOK™, 7 = extremely familiar withFACEBOOK™),

‘‘Please rate on the 7 point scale below how familiar you feelyou are with Facebook”.

Among these research assistants the level of familiarity withFACEBOOK™ was relatively high (M = 5.67, SD = 1.82), however,scores ranged from 1 to 7.

The research assistants were also asked about their log inbehavior. Specifically, they were asked,

‘‘On average, how often do you log into Facebook in a one-weekperiod?”

On average, the majority of research assistants logged intoFACEBOOK™ multiple times per week (M = 19.17, SD = 22.92),although the range from 0 to 85 times per week was variable, withtwo participants never signing in. The majority of participants(50%) indicated that they logged in between 12 and 15 times perweek.

The research assistants were asked to use the checklist to re-spond to one question. Research assistants were asked to indicate(with a checkmark) which items they felt were standard or default

Table 1Frequencies for all dichotomously scored items in the FACEBOOK™ checklist.

Item Sub categories within dichotomous variables Frequency (%) N

Can you poke them? 99.30 397Can you message them? 99 396Mini-feed 97.50 390Profile picture 91.80 367

(Profile picture) self 73.80 295(Profile picture) friends 22 88(Profile picture) relationship partner 11.30 45(Profile picture) activity 7.30 29(Profile picture) random picture 7 28(Profile picture) family 4.50 18(Profile picture) animal(s) 3.80 15(Profile picture) school 0.50 2(Profile picture) work 0.30 1

Birth day 88 352Birth day 83.80 335Birth year 73.30 293

Friends viewable 87.80 351College/university 83.30 333Wall 83 332Gender 82 328

(Gender) male 38.80 155(Gender) female 29 116

Applications 81.50 326Groups (joined at least one group) 79.30 317Photos of. . . 77.50 310Tagged photos 75.80 303Can you send them a gift? 71.30 285Photo album(s) 70 280Relationship status 63 252

(Relationship status) in a relationship 30.30 121(Relationship status) single 23.80 95(Relationship status) married 4.80 19(Relationship status) engaged 3 12(Relationship status)it’s complicated 1.30 5(Relationship status) in an open relationship 0 0

High school 62.80 251Concentration 62.30 249Graduation year 56 224Self-selected photos 55.30 221Interested in (sexual orientation) 54.50 218

(Interested in) women 33 132(Interested in) men 22.50 90

Favorite music 51.30 205Favorite movies 50.30 201Gifts 47.80 191Favorite quotes 47.30 189Interests 46.80 187Favorite TV shows 46.80 187Hometown 45.40 182Email address 43.30 173Favorite books 40.80 163Activities 37 148Employer 35.3 141Political views 35 140Job listed (have listed at least one job) 32.50 130Religious views 32 128About me 30.80 123Position 30.50 122Status 25.80 103Posted items 25.30 101Time period 24.30 97Notes 23.50 94Personal pages 23.50 94Looking for (relationship preference) 19.50 78

(Looking for) friendship 19.80 78(Looking for)relationship 5 20(Looking for)networking 4.80 19(Looking for) dating 4 16(Looking for)whatever I can get 2.50 10(Looking for) random play 2.30 9

Job description 17.80 71Tagged videos 13 52Fun wall 9.80 39Residence 9.30 37Degree 9 36

(continued on next page)


Table 2Range of scores, minimum scores, maximum scores, mean scores, and standard deviations for default/standard information, sensitive personal information and potentiallystigmatizing information (N = 400).

Scale N Range Min Max Mean SD

Default/standard 400 7 0 7 2.94 1.24Default/standard (minus age information) 400 5 0 5 1.37 .83Sensitive personal 400 11 3 14 9.65 2.27Potentially stigmatizing 400 14 0 14 7.07 3.81

Table 1 (continued)

Item Sub categories within dichotomous variables Frequency (%) N

(Degree) masters 4.50 18(Degree) bachelors 3.30 13(Degree) doctorate 2.30 9(Degree) diploma 1.30 5(Degree) post-doctorate 0 0(Degree) alumni 0 0

Website 8 32City/town 7.80 31Former name 5 20Mobile phone 5 20Events (as indicated in their mini-feed) 4.80 19Home address 3.50 14Room 3.30 13Advanced wall 3 12Super wall 3 12Market place listings 3 12Land phone 2.30 9Courses 2.30 9School mailbox 1.30 5Zip 0.80 3Awards 0 0


requirements for the construction of a personal profile or informa-tion typically required on personal profile websites. Specificallyparticipants were asked,

‘‘We are trying to determine what pieces of information peopleconsider to be ‘‘standard” or typical information requested onpersonal profile sites. We would like you to look through the listbelow to identify the elements that you believe are customarilyrequired in online profiles?”

Items from the checklist that were positively identified by morethan 50% of the assistants were considered for inclusion in thescale. Five items were identified as default or standard by morethan 50% of the participants. The items and percentage endorse-ment included: gender (83.3%), birth day (61.1%), birth year(72.2%), email (61.1%), and profile picture (55.6%).

Third, a local police department and police college were con-tacted, and through consultation 3 additional items (i.e., street ad-dress, city/town, postal code) were added to the default/standardinformation scale. These items identified personal information thatcould be used in potentially harmful ways.

Finally, this yielded an 8 item default/standard informationscale, comprised of street address, city/town, postal code, gender,birth day, birth year, profile picture and email (possible scores ran-ged from 0 to 8). Higher scores indicated higher levels of self-dis-closure for default items.

3.1.3. Sensitive personal informationThe second category reflected personally revealing/sensitive

information. This information was defined as details that couldbe used to locate or identify an individual, and could be used tothreaten or harm another. To develop this category, two research-ers conducted a thematic analysis of the checklist items to deter-mine which variables should belong in the sensitive informationcategory based on the nature of each variable and the potentialdanger in disclosing each item. After discussion, the following were

placed in the sensitive information category: email, employer, jobposition, status, mini-feed, regular wall, profile picture, photo al-bums, self-selected photos, tagged photos, message, poke, send agift, and friends viewable (possible scores ranged from 0 to 14).Higher scores indicated higher levels of self-disclosure.

3.1.4. Potentially stigmatizing informationThe third category reflected sensitive personal information that

could result in stigmatization within society. To develop this cate-gory two researchers conducted a thematic analysis of the items todetermine which variables should belong in the stigmatizing (aswell as sensitive) category based on the nature of each variableand the potential for stigmatization of the user by viewers. Afterdiscussion, the final scale was comprised of religious views, polit-ical views, birth year, sexual orientation, photos, friends viewable,interests, activities, favorite music, favorite movies, favorite TVshows, favorite books, favorite quotes, and about me (possiblescores ranged from 0 to 14). Higher scores indicated higher levelsof self-disclosure.

3.2. Results

Analyses explored whether any differences in self-disclosureemerged for profile information. Analyses consisted of 15 indepen-dent t-tests with 5 t-tests for each of the grouping categories (de-fault/standard information, sensitive personal information, andpotentially stigmatizing information). Comparisons assessed po-tential differences as a function of network type (university vs.community), gender revealed (indicated/not indicated), gender(male vs. female), relationship status revealed (indicated/not indi-cated), and age revealed (indicated/not indicated). The probabilityfor each comparison was corrected to p < .003 to accommodate thenumber of tests conducted. Examination of variables dealing withrelationships and with age was carried out using analysis of

Fig. 1. Presence of default/standard information, sensitive personal information,and potentially stigmatizing information as a function of relationship status (single,in a relationship and status not indicated/missing).


variance (ANOVAs) and regression analyses. Refer to Table 3 for asummary of analyses.

3.2.1. Comparisons among network, gender revealed, genderidentified, relationship status revealed, and age revealed

Overall, within each set of 5 t-tests, the same three comparisonswere statistically significant. Specifically, users who provided infor-mation about their gender (present or absent), relationship status,and age disclosed more default/standard information, more sensi-tive personal information, and more potentially stigmatizing infor-mation in their online profiles than their peers who did not disclosetheir gender, relationship status or age (smallest t(1, 398) = �2.81,p = .005). See Table 3 for a complete summary.

3.2.2. Relationships status (single, in a relationship, status notindicated/missing)

For each of the three grouping categories (i.e., default/standardinformation, sensitive personal information, and potentially stig-matizing information), one analysis of variance (ANOVA) was con-

Table 3Means, standard deviations and t-test statistics for default/standard information,sensitive personal information and potentially stigmatizing information.

Grouping Strategy/Factor Mean (SD) t-Test

Default/standard informationNetwork type n.s.

UniversityCommunity

Gender revealed t(1, 398) = �9.92, p < .001*

Indicated 3.14 (1.11)Not indicated 1.41 (1.13)

Gender n.s.MaleFemale

Relationship status revealed t(1, 398) = �4.94, p < .001*


Age revealed t(1, 398) = �3.81, p < .001*

Indicated 1.12 (.88)Not indicated 1.47 (.79)

Sensitive personal informationNetwork type n.s.

UniversityCommunity






Age revealed t(1, 398) = �3.05, p = .002*


Potentially stigmatizing informationNetwork type n.s.

UniversityCommunity






Age revealed t(1, 398) = �4.36, p < .001*


* Significant at the .003 level.

ducted to compare the amount of disclosure of default/standardinformation, sensitive personal information, and potentially stig-matizing information as a function of one of three relationship sta-tus possibilities (single, in a relationship, missing)1. Relationshipstatus served as the between subjects factor.

Results for the default/standard and sensitive personal informa-tion categories provided similar outcomes. Specifically, there was asignificant main effect of relationship status condition (F(2, 397) =15.37, p < .001 and F(2, 397) = 12.85, p < .001, for default/standardinformation and sensitive personal information, respectively).Tukey-b post hoc comparisons revealed that users who indicatedtheir relationship status as either single or in a relationship dis-closed significantly more default/standard information (Msin-gle = 3.34 and Mrelationship = 3.11) and sensitive personalinformation (Msingle = 10.12 and Mrelationship = 10.08) than userswho did not indicate their relationship status (Mmissing = 2.54 andMmissing = 8.94, for default/standard information and sensitivepersonal information, respectively).

Although the analysis for stigmatizing also yielded a significantmain effect of relationship status condition (F(2, 397) = 14.46,p < .001), the Tukey-b post hoc comparisons revealed that all threerelationship statuses significantly differed from one another. Singleusers (M = 8.61) disclosed the highest amount of this grouping ofinformation in their profiles, followed by users in a relationship(M = 7.16), and finally, users who did not indicate their relationshipstatus disclosed the least amount of stigmatized information(M = 6.03). See Fig. 1.

3.2.3. AgeThree linear regressions were conducted to explore whether age

predicted disclosure of information for each of the three categories(default/standard information, sensitive personal information, andpotentially stigmatizing information). The default/standard infor-mation was altered for this analysis through the removal of birthday and birth year in order to prevent redundancy in this measure.Age was entered as the predictor variable, and the new defaultinformation aggregate scale was entered as the dependent vari-able. The overall model was significant in all three cases, R2 = .06,F(1, 289) = 18.53, p < .001, R2 = .08, F(1, 289) = 11.31, p < .001 andR2 = .08, F(1, 289) = 25.74, p < .001, for default/standard informa-tion, sensitive personal information, and potentially stigmatizinginformation, respectively. In all three cases, as age increased, the

1 Only two statuses (i.e., single and in a relationship) and status missing (did noindicate their relationship status) were used in these analyses due to very small celnumbers (largest n < 19) for the remaining four statuses (i.e., married, engaged, it’scomplicated, in an open relationship).

tl


amount of default information presented in personal profilesdecreased (b = �.25, p < .001, b = �.19, p = .001, and b = �.29,p < .001, for default/standard information, sensitive personal infor-mation, and potentially stigmatizing information, respectively).

2 Cronbach alphas for scales consisting of only 3 items were low. Alpha levels were.48 for contact information, and .57 for photo album and profile picture information.Inter-item correlations were carried out for scales with 2 items. Item correlationswere .37 for view information, .34 for wall information, and .34 for relationshipinformation.

3.3. Discussion

In summary, all three disclosure categories were relevant, inthat all were sources for disclosure and potentially sources forthreat. The results revealed a similar trend for all three groupingcategories (i.e., default/standard, sensitive personal information,and potentially stigmatizing information). FACEBOOK™ userswho were acknowledged to have included information about theirgender, relationship status, and age, disclosed more information inall three disclosure categories than people who did not indicatethis information at all. Further, users who indicated they were sin-gle had the highest number of stigmatizing items, followed byusers who were in a relationship, and finally, users who did notspecify a status had the lowest number of stigmatizing items. Thistrend was not present for default/standard information or sensitivepersonal information. Instead, users who were single or in arelationship did not differ on their disclosure of default/standardinformation or sensitive personal information, but users who didnot indicate a status had less of this information in their onlineFACEBOOK™ profiles.

Individuals at greatest risk for threat, therefore, are those whoare seeking a relationship. It could be that those people who aresearching for romantic relationships, are using the online mediumas a way to self-present or to advertise themselves to potential dat-ing partners. In fact, research has shown that a large majority ofpeople looking to date use the Internet as a means for finding apartner (Madden & Lenhart, 2006). Online daters not only disclosemore intimate details, but disclose these details at a faster pacethan do offline daters (Rosen, Cheever, Cummings, & Felt, 2008).While disclosing details online may be necessary to attract a po-tential mate, there is clearly greater potential for harm.

Interestingly, the type of network and whether the profile wasgenerated by a male or female did not impact on whether or notinformation appeared in a FACEBOOK™ profile. That is, the com-munity and university samples, and males and females did not dif-fer in the amount of information disclosed in their online profiles.While females are generally high disclosers when compared tomales in more traditional social interactions (Dindia & Allen,1992), perhaps males feel more comfortable disclosing in an onlinesetting where there is less pressure to conform or adhere to strictsocial rules and male stereotypes for disclosure.

Lastly, for those users who included information about theirage, as age increased, disclosure of all three types of information(default/standard, sensitive personal information, and potentiallystigmatizing information) decreased. Older individuals’ may bemore cautious when disclosing details about themselves. For theyounger generation, disclosing personal information across a vari-ety of domains (e.g., school, online) is part of everyday life, andmay have begun to reduce the gap between private and publicselves. Thus, younger people may be less cautious when disclosinghighly personal details and feel more comfortable with online dis-closure (Goodstein, 2007). Alternatively, it is also possible that asage increases, disclosing certain types of information may not beseen as appropriate. For example, in Western society it is socialcustom that one does not ask older people for socially revealinginformation as a mark of respect (Mann, 2007). Yet, another possi-ble explanation is that social communication that is asynchronousyet personal, such as online social networks, may not provide theforum for communication that is comfortable for older users.While older people may be open to technology, they use is less

(Bucur, Renold, & Henke, 1999), and thus may prefer differentmediums for social connections.

In summary, although Study 1 indicated that there were incon-sistencies such that some pieces of personal and revealing informa-tion were disclosed and others were not, the results from Study 2provided insight indicating that some pieces of personal informa-tion may be more likely to be grouped together as well as providingan initial understanding of potential individual differences that pre-dict who is most likely to disclose what information. Study 3 furtherexplored variables that explain what is disclosed and by whom.

4. Study 3

One concern in Study 2 is that only limited information wasexamined. Study 3 provided an examination of all of the contentavailable through FACEBOOK™ in order to better understandhow to conceptualize the information that can be provided andwho is likely to disclose it.

4.1. Method and results

The sample included the 400 profiles specified in Study 1 andemployed the same checklist.

A fourth grouping strategy was conducted to include all infor-mation present in FACEBOOK™ but to organize it in a meaningfulway. The process involved several steps. First, information wassorted as a function of the layout of the FACEBOOK™ profile tem-plate. However, the template organization tended to include infor-mation that was not obviously related thematically. Tworesearchers carried out discussion about how to categorize infor-mation based on thematic coherence. Through discussion, the fol-lowing 10 categories were identified: personal information,relationship information, age information, contact information,education information, work information, view information, gen-eral picture information, message and poke acceptance (whetheror not users allowed other users to send them private messagesand to poke them), and wall and update information (i.e., regularwall, advanced wall, super wall, fun wall, mini-feed, and status).

A principal components factor analysis with a Varimax rotationwas conducted to confirm these categories and the relationshipsbetween these variables. The factor analysis yielded 11 factorswith Eigenvalues greater than 1.00, which accounted for a totalof 66.74% of the variance. Overall, the factors corresponded withthe thematic analysis of the two researchers with the exceptionthat one category identified by the researchers was divided intotwo categories through the factor analysis. Initial Eigenvalues ran-ged from 1.04 to 6.60, and indicated that the factors explained16.93%, 9.62%, 8.20%, 6.13%, 4.80%, 4.54%, 4.31%, 3.50%, 3.08%,2.96%, and 2.68% of the variance, respectively. The factor loadingmatrix is presented in Table 4. The following eleven aggregatescales were created based on the factor loadings and were usedin the subsequent analyses: personal information, photo and up-date information, work information, education information, mes-sage and poke acceptance (whether users allow for receipt ofprivate messages and nudges from other users), photo album andprofile picture information, age information, contact information,view information, other wall presence, and relationship informa-tion. Higher scores indicated higher levels of self-disclosure forthese items. Reliability was conducted for scales that had morethan 3 items using Cronbach’s alpha2. The alphas ranged from

Table 4Factor loadings based on a principal components analysis with Varimax rotation for 39 items.

Variable 1 2 3 4 5 6 7 8 9 10 11

Sexual orientation .167 .152 .105 .106 .004 .096 .114 .054 .110 �.073 .729Looking for .155 .020 .161 �.135 �.036 �.015 .081 �.027 .103 .230 .700Birth day �.034 .095 .015 .096 .031 .049 .885 .039 .074 .013 .133Birth year .028 .062 .010 .121 .006 �.034 .891 �.046 .053 �.048 .029Political views .161 �.014 .143 .006 .015 .061 .150 .046 .689 �.092 .227Religious views .265 .052 .133 �.018 .017 .045 .020 .160 .664 .008 .014Address �.014 �.001 .056 .061 .064 .004 �.026 .798 .028 .020 �.016City/town �.003 .111 �.009 .121 �.324 �.003 .057 .527 .258 �.032 �.222Zip or postal code .046 .011 �.090 .018 .028 .033 �.004 .825 .030 �.017 .130Activities .694 .055 .060 .073 .017 .066 �.024 .043 .053 .037 .112Interests .806 .083 .060 .061 �.025 .017 .076 �.009 .081 .056 .035Favorite music .837 .105 .057 .050 .046 .012 .014 �.023 .038 .017 .107Favorite TV shows .832 .078 .088 .000 �.030 .024 .042 �.037 �.093 .023 .165Favorite movies .864 .123 .056 .040 �.019 �.008 .063 �.021 �.026 .002 .041Favorite books .788 .020 .058 .059 .029 .054 .032 .030 .069 �.007 �.031Favorite quotes .656 �.021 .059 .000 �.029 .107 �.062 �.003 .156 �.091 �.045About me .660 .002 .084 .039 �.077 .085 �.117 .056 .150 .021 .018College/university .001 .013 .103 .764 .035 �.169 .105 .086 �.175 �.096 .004Concentration .103 �.018 .080 .781 �.034 .041 .057 .050 �.064 �.129 .159Graduation year .086 .058 .157 .712 .063 .098 �.001 �.018 .150 �.030 �.164High school .086 �.005 .119 .689 �.021 .213 .078 .052 .093 .040 �.016Employer .078 .048 .891 .121 �.019 �.035 .060 �.011 .031 .036 .046Job position .084 .032 .904 .145 .003 �.029 .042 �.014 .042 .054 .076Job description .166 .046 .765 .023 .032 .015 .006 �.007 .072 .052 .022Job city/town .034 �.014 .486 .091 �.133 .050 �.162 .003 �.014 �.148 .265Job time period .106 �.051 .828 .105 �.059 .077 .039 �.003 .099 .105 �.011Photos of. . . .089 .848 .052 .015 �.036 .222 .072 .053 �.078 �.044 �.023Tagged photos .066 .849 .069 .073 .024 .256 .071 .053 �.066 �.050 �.016Self-selected photos .149 .539 .188 �.042 .016 .411 .141 .110 �.347 �.120 .044Friends viewable .077 .704 �.022 �.022 .161 .060 .000 �.052 .074 .106 .109Wall .051 .802 .003 �.101 .053 �.132 .012 .043 .052 .097 �.006Groups .096 .553 �.111 .213 .127 .183 .015 �.062 .193 .146 .157Photo albums .085 .172 .171 �.053 .046 .697 .225 .071 �.259 �.054 �.010Send message �.055 .150 �.060 .017 .929 .062 .063 .011 .030 .007 �.064Poke �.024 .121 �.054 .038 .940 .068 �.015 �.006 .009 .000 .008Profile picture .117 .259 �.059 .226 .103 .646 �.118 �.039 .186 .022 .019Applications .129 .172 �.054 .080 .031 .659 �.085 .002 .198 .166 .088Fun wall .050 .052 .083 �.061 �.008 .016 �.022 .016 �.054 .777 .004Super wall �.021 .075 .035 �.094 .018 .065 �.011 �.023 .000 .767 .077


moderate to high: .91 for personal information (8 items), .90 forwork information (4 items), .76 for education information (4 items),and .84 for photo and update information (5 items). Refer to Table 5for a list of items in each scale and means.

Following the analyses in Study 2, t-tests were conducted firstto explore possible differences as a function of network type (uni-versity vs. community), gender revealed (indicated/not indicated),gender (male vs. female), relationship status revealed (indicated/not indicated), and age revealed (indicated/not indicated) for eachof the 11 variables. Given the number of t-tests conducted, the al-pha level for each test was set to .005. These t-tests were followedby the analysis of variance (ANOVA) looking within relationship

Table 5Variables, mean scores, and standard deviations for factor analyzed topic scales (N = 400).

Scale Variables included in scale

Personal information Activities, interests, favorite music, favoriteinformation ‘‘about me”

Photo and update information Photos of. . ., friends viewable, tagged photWork information Employer, position, description, and time pEducation information College/university, concentration, graduatioMessage and poke information Send a private message, and poke (nudge)Photo album and profile picture

informationProfile picture, photo albums, and applicat

Age information Birth day and birth yearContact information Address, city/town, and zip/postal codeView information Political views, and religious viewsOther wall presence Fun wall and super wallRelationship information Looking for (relationship, friendship, netwo

status, and finally, with the linear regression using age. Refer toTable 6 for a summary of the t-test statistics.

Type of network. One out of 11 t-tests was significant. Results re-vealed that users who belonged to a community network weremore likely to include their political and religious views in theirpersonal profiles than were their university network counterparts.

Gender revealed (indicated/not indicated). Five out of the 11 t-tests were significant. Results revealed that users who indicatedtheir gender, also had higher levels of disclosure for: personalinformation, photo and update information, education informa-tion, photo album and profile picture information, and ageinformation.

Mean (SD)

TV shows, favorite movies, favorite books, favorite quotes, and 3.51 (3.07)

os, self-selected photos, groups, and wall 4.50 (1.85)eriod 1.80 (1.54)n year, and high school 2.64 (1.41)

1.98 (.18)ions 2.43 (.84)

1.57 (.76).12 (.41).67 (.78).13 (.39)

rking, random play etc), and sexual orientation .74 (.73)

Table 6Means, standard deviations, and t-test statistics for FACEBOOK™ topic scales as afunction of network type, gender revealed (indicated/not indicated), gender (male/female), relationship status revealed (indicated/not indicated), and age revealed(indicated/not indicated).

Factor Mean (SD) t-Test

Network typeView information t(1, 398) = 2.84, p = .005*

University .56 (.75)Community .78 (.80)

Gender revealedPersonal information t(1, 398) = �3.75, p < .001*


Photo and update information t(1, 398) = �6.19, p < .001*


Education information t(1, 398) = �6.61, p < .001*


Photo album andprofilepicture information

t(1, 398) = �6.46, p < .001*

Indicated 2.53 (.72)Not indicated 1.72 (1.26)

Age information t(1, 398) = �5.61, p < .001*


GenderView information t(1, 352) = 5.58, p < .001*

Male .93 (.80)Female .48 (.70)

Relationship status revealedPersonal information t(1, 398) = �2.79, p = .005*

Indicated 3.83 (3.01)Not Indicated 2.95 (3.10)

Photo and updateinformation t(1, 398) = �4.08, p < .001*


Photo album andprofilepicture information

t(1, 398) = �3.86, p < .001*

Indicated 2.57 (.67)Not indicated 2.20 (1.03)

Age information t(1, 398) = �3.77, p < .001*


View information t(1, 398) = �3.37, p = .001*

Indicated .77 (.80)

Not indicated .50 (.73)

Age revealedEducation information t(1, 398) = 3.22, p = .001*


* Significant at the .005 level.

Table 7Means of each of the three relationship statuses for photo and update information,photo album and profile picture information, age information, and work information.

Scale Single In a relationship Status missing

Photo and update information 4.98 4.81 4.12Photo album and profile

picture information2.47 2.65 2.19

Age information 1.76 1.66 1.35Work information 1.38 1.21 .76


Male vs. female. One out of the 11 t-tests was significant. Resultsrevealed that males expressed more information about their polit-ical and religious views than did females.

Relationship status revealed (indicated/not indicated). Five of the11 t-tests were significant. Overall, disclosing one’s relationshipstatus was related to higher levels of disclosure of various topics,including: personal information, photo and update information,photo album and profile picture information, age information;and view information.

Age revealed (indicated/not indicated). One of the 10 t-tests wassignificant. Users who disclosed their age also disclosed more edu-cation information.

4.1.1. Relationship statusA series of one-way analyses of variance (ANOVAs) were con-

ducted with the relationship status condition (single, in a relation-ship, missing) as the between subjects factor and the each of the 11aggregate scales as the dependent variables. Results revealed a sig-

nificant main effect of relationship status condition for 7 of the 11scales including: personal information (F(2, 397) = 7.20, p = .001);photo and update information (F(2, 397) = 8.45, p < .001); photo al-bum and profile picture information (F(2, 397) = 12.40, p < .001);age information (F(2, 397) = 10.94, p < .001); work information(F(2, 397) = 5.80, p = .003); view information (F(2, 397) = 5.88,p = .003); relationship information (F(2, 397) = 18.91, p < .001).

For personal information, Tukey-b post hoc comparisons re-vealed that single users (M = 4.51) indicated significantly morepersonal information than did users who were either in a relation-ship (M = 3.36), or who did not disclose their relationship status(M = 3.04).

According to Tukey-b post hoc comparisons, similar patternswere found for photo and update information, photo album andprofile picture information, age information, and work informa-tion. In general, users who either indicated they were single or ina relationship, disclosed significantly more information related tothese topics than did users who did not disclose their relationshipstatus. Refer to Table 7 for means. For information pertaining toreligious and political views, Tukey-b post hocs showed that singleusers (M = .85) had significantly more information present in theirprofiles than did users who did not indicate their relationship sta-tus (M = .52). Lastly, for relationship information, all three groupsdiffered significantly from one another. Users who were single(M = 1.06) displayed the most relationship information, followedby users who were in a relationship (M = .78), and finally, userswho did not disclose their relationship status were the least likelyto display relationship information in their online profiles(M = .50). See Fig. 2.

4.1.2. AgeTen linear regressions were conducted to explore how age as a

continuous factor was related to disclosure of information online.Given the number of t-tests conducted, the alpha level for eachtest was set to .005. Age was entered as the predictor variable,and each of the aggregated topic scales were entered as thedependent variables. For 5 of the 10 scales, age significantly pre-dicted disclosure of topic. The overall models were significant forthe following scales: personal information (R2 = .04,F(1, 289) = 14.11, p < .001, b = �.22, p < .001); photo and updateinformation (R2 = .05, F(1, 289) = 16.46, p < .001, b = �.23,p < .001); education information (R2 = .04, F(1, 289) = 10.82,p = .001, b = �.19, p = .001); photo album and profile picture infor-mation (R2 = .09, F(1, 289) = 28.24, p < .001, b = �.30, p < .001);and relationship information (R2 = .07, F(1, 289) = 20.78, p < .001,b = �.26, p < .001). Overall, as age increased, disclosure of thesetopics decreased.

4.2. Discussion

For the most part, the outcomes of Study 3 support the out-comes reported in Study 2. Specifically, gender (male versus fe-male) and network membership (community versus university)were not important variables for distinguishing who would orwould not be likely to disclose information. Unlike traditional

Fig. 2. Presence of personal information, photo and update, photo album andprofile picture information, age information, work information, view information,and relationship information as a function of relationship status (single, in arelationship and status not indicated/missing).


face-to-face interactions (Dindia & Allen, 1992), online social net-working profiles appear to provide a means of communicating thatfacilitates disclosure among males to the same level as evidencedin females. Also, whether an individual belongs to a university-based or community-based network does not impact on theamount of disclosure of information. Apparently, both universityand community samples share similar knowledge or attitudes to-ward revealing information online.

Like Study 2, results from Study 3 revealed that voluntarilyproviding personal information related to gender, and relation-ship status was related to greater disclosure of certain topics(i.e., for 5 of the 11 scales). One notable difference between Study2 and Study 3 concerns revealing personal information related toage. In Study 2, revealing any information about age (i.e., whetherthey indicated their age or not) had an impact on the likelihoodto disclose information. In contrast, for Study 3, whether an indi-vidual supplied information regarding their age or not did not im-pact on information disclosed. Out of the possible 10 topiccategories, revealing age was only significant for one (i.e., educa-tion information).

On the other hand, in both Studies 2 and 3, when individualsprovided their age, it proved to be an important factor in distin-guishing levels of disclosure online. As age increased, self-disclo-sure decreased on 5 out of 10 scales in Study 3 (i.e., photo albuminformation, relationship information, personal information, photoand update information, and education information) and for allcategories examined in Study 2. Items in these particular 5 areastended to overlap with the items in the three grouping categoriesused in Study 2 relative to the 5 areas that were not significantin this analysis.

Consistent with earlier discussions, older adults may be lesslikely to reveal information due to less familiarity and trust withtechnology or due to more experience, wariness or social prohibi-tions regarding disclosure of private information (Bucur et al.,1999; Mann, 2007).

Relationship status was very important in distinguishing lev-els of disclosure for the different topic categories. Overall, thoseseeking a relationship were far more inclined to disclose on avariety of topics including views, relationship information, and

personal information than were their counterparts who were ina relationship or who did not disclose a status. The current out-comes highlight the importance of relationship status for identi-fying those users who are more likely to reveal highly personalinformation. Those seeking a relationship may be using FACE-BOOK™ as a less overt dating site, and, thus, may be differen-tially motivated to disclose highly personal information across avariety of topics regardless of the dangers or threats associatedwith disclosing this information. Motivation therefore may proveto be an important factor, and merits further investigation in fu-ture studies.

Interestingly, contact, other wall information (i.e., Fun Wall,Advanced Wall and Super Wall), and acceptance of messagesand pokes yielded non-significant results. In the case of accep-tance of pokes and messages, a very high percentage (99.3%and 99% for pokes and messages, respectively) of users had thesefeatures present in their profiles. It may be that because thesefeatures are a prompted or automatic feature present in a FACE-BOOK™ profile, less variability may result for these particularitems. Alternatively, these may also be perceived as more ‘‘play-ful” and social opportunities to interact and may not carry anyperceived threat in having them as part of the profile. For otherwall information, those that occurred in less than 10% of the pro-files, there might be less use as a function of the amount of effortand time required to generate the materials required to developthese features.

Interestingly, items involving contact information did not re-sult in any significant outcomes as a function of age, gender, net-work or relationship status. Perhaps this is a function of theacknowledgement of the high sensitivity of this information withthis information being reported very infrequently. Given themany other alternatives for communicating with someone inthe social network, including sending a wall post, these otheralternatives may be accepted more readily than highly sensitivepersonal information because these former alternatives leavethe door open for connection without exposing oneself toinformation theft or invasion. Thus, it appears that regardless ofgroup membership, people did not voluntarily disclose thisinformation.

5. General discussion

The current study provided an initial, systematic exploration ofFACEBOOK™ profiles. The study identified what kinds of informa-tion can be disclosed through these profiles, as well as when infor-mation is likely to be disclosed, and by whom.

Specifically, the first contribution of this research was the con-struction of a comprehensive checklist to identify the types ofinformation available on the FACEBOOK™ profiles. This first stepwas important for establishing a general checklist that can beused in future FACEBOOK™ research. No known checklist, thataccounted for all of the possible dichotomously items containedwithin online FACEBOOK™ profiles, existed prior to this research.This checklist indicated that, for some online social networkingusers, a great deal of information is shared through this mediumand that some information is more likely to be revealed thanothers.

5.1. Understanding what is disclosed

In order to make meaningful conclusions about the data foundin the profiles, grouping procedures were used to aggregate var-ious items within descriptive or associated categories (i.e., de-fault/standard, sensitive personal and potentially stigmatizing,as well as according to specific topics). The grouping strategies


permitted a novel, exploratory examination of the impact thatindividual difference factors have on disclosure. For the mostpart, age and relationship status were salient in describing whowould or would not be likely to disclose information, while tra-ditional variables such as gender were not significant. In fact,across the grouping strategies there was considerable overlap inoutcomes suggesting that some types of information were partic-ularly salient for some groups.

Overall, this study was important because it provided evidencethat highly personal, sensitive, and potentially stigmatizing infor-mation is being disclosed on social networking sites such as FACE-BOOK™. However, the study also depicts users who are expressingdiscretion regarding at least some personally revealing informa-tion. In light of the prevalence of online identity theft, and socialthreat issues, the results of this study can be used to support theneed for developing programs and interventions that further cau-tion users of online social networks against placing themselves atrisk.

Acknowledgements

The authors would like to thank Wilfrid Laurier University forfunding this project through Operating funds and through a SocialSciences and Humanities Institutional grant. We would also like tothank the following individuals for their comments and support increating this manuscript, Vanessa Buote, Elizabeth Flynn-Dastoor,Amy Grant, Linzi Williamson, and Alissa Anderson.

Appendix A

Variable checklist used in coding each FACEBOOK™ profile.

Variable name
Variable description
1
FACEBOOK™username link
Web address of the FACEBOOK™profile accessed

2
Network searched What network (community oruniversity) you belong to (e.g.,Toronto, ON)
3
Size of overall network Number of members in thenetwork
4
Status yes/no Update information that notifiesother users of your whereaboutsand actions (e.g., what you arethinking, feeling, doing etc. . .)
5
Gender yes/no Sex (e.g., male, female) 6 Male 7 Female 8 Interested in (sexual
orientation) yes/no
Whether you are interested in menor women
9
Men 10 Women 11 Relationship status
yes/no
Whether you are: single, in arelationship, engaged, married, it’scomplicated or in an openrelationship
12
Single 13 In a relationship 14 Engaged 15 Married 16 It’s complicated 17 In an open relationship 18 Former name yes/no If you are married and changed
your name, what your maidenname was

Appendix A (continued)

Variable name
19
Looking for(relationshippreference) yes/no
Whether you are looking forfriendship, dating, a relationship,or social networking

20
Random play 21 Whatever I can get 22 Friendship 23 Dating 24 Relationship 25 Networking 26 Birth day yes/no The day or month you were born 27 Birth date The day you were born 28 Birth year The year you were born 29 Hometown yes/no The town you grew up in 30 Political views yes/no Your political stance (e.g., liberal,
conservative etc. . .)
31 Religious views yes/
no
Your religious stance (e.g., Catholic,Christian, Muslim etc. . .)
32
Mini-feed yes/no Updates/list of your FACEBOOK™activity (e.g., events you areattending, friends you have added,pictures that have been posted ofyou etc. . .)
33
Email address yes/no Your email 34 Mobile phone yes/no Your cell phone number 35 Land phone yes/no Your land line phone number 36 School mailbox yes/no Your school mailbox 37 Residence yes/no The residence you live in 38 Room yes/no The residence room you live in 39 Address yes/no Your home address 40 Current city/town yes/
no
The city or town you live in
41
Zip yes/no Your zip or postal code 42 Website yes/no Personal or interest website
address (e.g., justudents website)
43 Activities yes/no Things you like to do (e.g., sports,
hobbies, leisure activities etc. . .)
44 Interests yes/no Your personal interests (e.g.,
painting, photography)
45 Favorite music yes/no Bands/songs or genres of music
that you like
46 Favorite TV shows
yes/no
TV shows/genres of shows that youlike
47
Favorite movies yes/no
Movies that you like to watch

48
Favorite books yes/no Favorite books you have read 49 Favorite quotes yes/
no
Quotations that you enjoy
50
About me yes/no Personal details about yourself(e.g., you love hot chocolate andare the eldest of 3 children)
51
College/universityyes/no
The university/college youattended or are currently attending

52
Concentration Yes/No Your discipline or area of interest(e.g., biology)
53
Degree yes/no What degree you are working on(e.g., bachelors, masters etc. . .)
54
Diploma 55 Bachelors 56 Masters 57 Doctorate 58 Post-doctorate 59 Alumni 60 Graduation year yes/
no
When you graduated or areplanning to graduate
61
High school yes/no The high school you are attendingor attended

Appendix A (continued)Appendix A (continued)


Variable name
62
Courses yes/no Any courses you took 63 Awards yes/no Any awards you won 64 Job listed (have listed
at least one job) yes/no

Who you work for (or have workedfor in the past)

65
Employer yes/no What your job position is (or wasin the past)
66
Position yes/no What your job duties were 67 Description yes/no The city in which you work(ed) 68 Job city/town yes/no The time period in which you
worked
69 Time period yes/no An upgraded bulletin board where
users post messages and largegraphics for each other to see, userscan add applications

70
Advanced wall yes/no An upgraded bulletin board whereusers create drawings and postgraffiti for each other to see
71
Fun wall yes/no An upgraded bulletin board whereusers post messages and largegraphics for each other to see
72
Super wall yes/no 73 Wall yes/no A bulletin board where users post
messages for each other to see
74 Profile picture yes/no The main photo seen at the top of
your profile (e.g., picture ofyourself, friends, animals etc. . .)

75
Self 76 Activity 77 Friends 78 Relationship partner 79 Family 80 Work 81 School 82 Animal(s) 83 Random picture 84 Photos of. . . yes/no Users can upload photos of
themselves and others
85 Tagged photos yes/no Photos that have been uploaded by
another user, and where the profileuser has been identified or labeledin a photo

86
Self-selected photosyes/no
Photos that have been uploaded bythe actual profile user themselves

87
Tagged videos yes/no Videos that the user has beenidentified in and tagged by anotheruser
88
Friends viewable yes/no
All the friends on your friend list

89
Can you send them agift? yes/no
Whether you can send a gift to theuser without prior permission

90
Can you messagethem? yes/no
Whether you can send a messageto the user without priorpermission

91
Can you poke them?yes/no
Whether you can send a poke (likea virtual nudge to let them knowyou are there) to the user withoutprior permission

92
Applications yes/No Any additional features that are notdefault to FACEBOOK™ (e.g.,scrabulous, must be added by theuser by signing up for theapplication)
93
Photo album(s) yes/no Users can create online photoalbums where they can pick andchoose pictures and upload themto their profile
Variable name
94
Groups (joined at leastone group) yes/no
Any groups that users belong to(e.g., groups based around sharedinterests, activities, or anythingyou like)

95
Events (as indicated intheir mini-feed) yes/no
Any parties, get-togethers or socialevents you are attending

96
Gifts yes/no Gifts that other users have sent you 97 Posted items yes/no Posted websites, blogs, videos,
anything users choose to post forothers to view

98
Market place listingsyes/no
A classified ad section online whereusers can buy and sell advertisedgoods

99
Notes yes/no Written entries detailing personalthoughts, opinions or generalcomments
100
Personal pages yes/no Personal web-pages or links thatare connected to the users profile
References

Acoca, B. (2008). Scoping paper on online identity theft (Ministerial BackgroundReport DSTI/CP (2007) 3/FINAL. Retrieved May 27, 2009. Available from http://www.oecd.org/dataoecd/35/24/40644196.pdf.

Bucur, A., Renold, C., III, & Henke, M. (1999). How do older netcitizens compare withtheir younger counterparts? CyberPsychology & Behaviour, 2, 505–513.

Christofides, E., Muise, A., & Desmarais, S. (2009). Information control anddisclosure on Facebook: Are they two sides of the same coin or two differentprocesses? CyberPsychology and Behavior, 12, 1–5.

Dindia, K., & Allen, M. (1992). Sex differences in self-disclosure: A meta-analysis.Psychological Bulletin, 112, 106–124.

Ellison, N. B., Steinfield, C., & Lampe, C. (2007). The benefits of Facebook ‘‘friends:”Social capital and college students’ use of online social network sites. Journal ofComputer-Mediated Communication, 12, 1143–1168.

Fogel, J., & Nehmad, E. (2009). Internet social network communities: Risktaking, trust, and privacy concerns. Computers in Human Behavior, 25,153–160.

Goodstein, A. (2007). Totally wired: What teens and tweens are really doing online.New York, NY: Macmillan, St. Martin’s Griffin.

Henderson, S., & Gilding, M. (2004). ‘I’ve never clicked this much with anyone in mylife’: Trust and hyperpersonal communication in online friendships. New Media& Society, 6, 487–506.

Joinson, A. N. (2001). Self-disclosure in computer-mediated communication: Therole of self-awareness and visual anonymity. European Journal of SocialPsychology, 31, 177–192.

LaRose, R., & Rifon, N. (2006). Your privacy is assured—of being invaded:Websites with and without privacy seals. New Media and Society, 8,1009–1029.

Lee, D. H., Im, S., & Taylor, C. R. (2008). Voluntary self-disclosure of information onthe internet: A multi-method study of the motivations and consequences ofdisclosing information on blogs. Psychology & Marketing. Special Issue: NewDevelopments in E-commerce, 25, 692–710.

Madden, M., & Lenhart, A. (2006). Americans who are seeking romance use the internetto help them in their search, but there is still widespread public concern about thesafety of online dating. Pew Internet and American Life Project. Available fromhttp://www.pewinternet.org/pdfs/PIP_Online_Dating.pdf.

Mann, B. (2007). The polite way for a man to ask someone’s age. RetrievedJune 1, 2009. Available from www.associatedcontant.com/article/299592/the_polite_way_for_a_man_to_ask_someons.html.

McCandlish, S. (2002). EFF’S Top 12 ways to protect your online privacy. Retrieved onJune 1, 2009. Available from http://www.eff.org/wp/effs-top-12-ways-protect-your-online-privacy.

Messmer, E. (2007). Study: Facebook users easy targets for identity theft. Retrieved onJune 1, 2009. Available from http://www.macworld.com/article/59488/2007/08/facebook.html.

Rosen, L. D., Cheever, N. A., Cummings, C., & Felt, J. (2008). The impact ofemotionality and self-disclosure on online dating versus traditional dating.Computers in Human Behavior, 24, 2124–2157.

Statistics, FACEBOOK™. (2008). Retrieved January 20, 2008. Available from http://www.facebook.com/press/info.php?statistics.

Sullivan, D. (2008, January 14). Web 2.0 identity theft. Message posted to http://www.realtimewebsecurity.com/type/mt-tb.cgi/593.

Tidwell, L. C., & Walther, J. B. (2002). Computer-mediated communicationeffects on disclosure, impressions, and interpersonal evaluations: Getting to

http://www.oecd.org/dataoecd/35/24/40644196.pdf

http://www.oecd.org/dataoecd/35/24/40644196.pdf

http://www.pewinternet.org/pdfs/PIP_Online_Dating.pdf

http://www.associatedcontant.com/article/299592/the_polite_way_for_a_man_to_ask_someons.html

http://www.associatedcontant.com/article/299592/the_polite_way_for_a_man_to_ask_someons.html

http://www.eff.org/wp/effs-top-12-ways-protect-your-online-privacy

http://www.eff.org/wp/effs-top-12-ways-protect-your-online-privacy

http://www.macworld.com/article/59488/2007/08/facebook.html

http://www.macworld.com/article/59488/2007/08/facebook.html

http://www.facebook.com/press/info.php?statistics

http://www.facebook.com/press/info.php?statistics

http://www.realtimewebsecurity.com/type/mt-tb.cgi/593

http://www.realtimewebsecurity.com/type/mt-tb.cgi/593


know one another a bit at a time. Human Communication Research, 28,317–348.

Walther, J. B., Van Der Heide, B., Kim, S.-Y., Westerman, D., & Tong, S. T. (2008). Therole of friends’ appearance and behavior on evaluations of individuals onFacebook: Are we known by the company we keep? Human CommunicationResearch, 34, 28–49.

Willard, N. E. (2006). A briefing for educators: Online social networking communitiesand youth risk. Center for Sage and Responsible Internet Use.

Yum, K. (2007). Facebook says ‘Thanks, Canada’. National Post; May 18.Zhao, S., Grasmuck, S., & Martin, J. (2008). Identity construction on Facebook: Digital

empowerment in anchored relationships. Computers in Human Behavior, 24,1816–1836.

Documents

All about me: Disclosure in online social networking profiles: The case of FACEBOOK