May 3, 2007

Lit Review: Information Revelation and Privacy in Online Social Networks (The Facebook Case) (Gross & Acquisti)

The researchers conducted an analysis of over 4,000 college students using Facebook at Carnegie Mellon, utilizing the lens of information revelation and related privacy implications.
-Sought to examine the openness of individuals in revealing information (such as contact information, political and sexual orientation, and intimate details of one's personal life) freely posted in the public realm of the Internet.
-Collected actual field data, rather than surveys or experiments.

Information Revelation and Online Social Networking
1. Identifiability
-Varies according to the nature of the site, though most encourage identifiable photos.
2. Types of information elicited
-range from the semi-public to the private to entirely open-ended (diary communities).
3. Visibility of information
-Can be viewed by all members or limited to one's personal network.

-Anecdotal evidence reveals an utter willingness of members to reveal private information.
-Social Network Theory and Privacy: Discussions center on the complex nature of one's propensity to disclose personal information, the importance of weak ties in the formation of social capital, and expectations of privacy.
-In the offline world, relationships are dynamic and can exist at multiple levels of intimacy. Online, relationships are reduced to simply "friend or not".
-Though not necessarily supportive of strong ties, the Internet facilitates the formation of a large and dispersed network of weak ties.
-Situating the Internet as a vast network of rather weak ties, it has been described by some as an imagined community (Anderson), and thus the meaning of trust must be renegotiated, as well as the meaning of intimacy.
-The Internet slightly facilitates meaningful interaction while greatly enhancing the ability of others to access your information.
-Privacy Implications: Photos, demographic data, unique tastes may lead to a re-identification of an individual belonging to more than one SNC. This occurs either through recognition of a pseudonymous user by searching for this information, or knowledge of unknown characteristics of an identified subject on another site.
-Members are often not fully aware of a hosting site's privacy policies concerning information disclosure, or the magnitude of the site's user population and/or data archival.
-Risks include identity theft, stalking, embarrassment and blackmailing.
-Factors in information revelation include peer pressure, perceived benefits outweighing potential harm, casual attitudes regarding privacy, lack of awareness of threat, trust in the service and its members, or the SNC interface itself.
-College-oriented SNCs are often based on a shared real space that is extended to a bounded virtual domain.
-Increased sense of trust and intimacy, however outsider access and rapid network expansion quickly challenge the "realness" of the community and expectations of privacy.
-Photo: 91%; Birthday: 87.8%; Phone: 39.9%; Residence: 50.8%; Dating Preferences, Relationship Status, Religious and Political Views
-Facebook encourages validity of information and a valid e-mail address.
-->89% real names, 8% fake names, 3% first name only
-->91% provide images: 61% directly identifiable, 80% useful for identification, 12% unrelated - in comparison to Friendster: 23% joke images, 55% directly identifiable
-->CMU students average 78.2 friends at CMU and 54.9 at other schools
Data Visibility and Privacy Preferences: Default settings allow everyone at same institution to view full profile, and full name/institution/status/photo show up in any general search. However, visibility and searchability are able to be defined by the individual user. Less than 3% of users alter their privacy settings.

Privacy Implications
-Facebook users appear generally unconcerned about information disclosure and potential ramifications.
1. Stalking: Physical presence can be determined based on location and class schedule; AIM (listed by 77.7% of users).
2. Re-Identification: the linkage of non-explicit information (name, address) with explicit information (common attributes). This can be based on demographics (all one needs is zip code, gender, and birthdate- provided by 44.3% of users), face (provided by 55.4%), social security number and identity theft (birthdate, residence, phone number)
3. Building a Digital Dossier: Sensitive data revealed in college, such as sexual orientation and political reviews, is archived and can potentially be mined in the future.
4. Fragile Privacy Protection: Social networks can be hacked! E-mail addresses can be hacked, manipulation of users (when 250,000 users were sent a friend request, 30% were willing to make all of their information available by accepting), advanced search features are available to anyone in the network looking to search for someone at any college

This article is slightly dated (2005), and concerns over privacy on the Internet have since grown exponentially due to media dramatization and new features implemented by Facebook (namely, the News Feed, which encouraged many to finally implement some of the privacy options made available to users). A simple survey tapping into perceived privacy, protective behaviors, and perceived audience would be easy to implement- Facebook does make recruiting participants a lot easier! Also, Facebook has since updated their privacy policy- a little highlighted review:

We may use information about you that we collect from other sources, including but not limited to newspapers and Internet sources such as blogs, instant messaging services and other users of Facebook, to supplement your profile. Where such information is used, we generally allow you to specify in your privacy settings that you do not want this to be done or to take other actions that limit the connection of this information to your profile (e.g., removing photo tag links).

We do not provide contact information to third party marketers without your permission. We share your information with third parties only in limited circumstances where we believe such sharing is 1) reasonably necessary to offer the service, 2) legally required or, 3) permitted by you.

We may be required to disclose user information pursuant to lawful requests, such as subpoenas or court orders, or in compliance with applicable laws. We do not reveal information until we have a good faith belief that an information request by law enforcement or private litigants meets applicable legal standards. Additionally, we may share account or other information when we believe it is necessary to comply with law, to protect our interests or property, to prevent fraud or other illegal activity perpetrated through the Facebook service or using the Facebook name, or to prevent imminent bodily harm. This may include sharing information with other companies, lawyers, agents or government agencies.

If the ownership of all or substantially all of the Facebook business, or individual business units owned by Facebook, Inc., were to change, your user information may be transferred to the new owner so the service can continue operations.

Individuals who wish to deactivate their Facebook account may do so on the My Account page. Removed information may persist in backup copies for a reasonable period of time but will not be generally available to members of Facebook.

May 2, 2007

Lit Review: Unraveling the Taste Fabrics of Social Networks (Liu, Maes & Davenport)

-Sought to uncover a semantic fabric of taste derived from the language used in 100,000 social networking profiles, dubbed the social Semantic Web.
-First mapped users onto taste-spaces, then compared the taste-similarities of participants.
-Moving away from formal semantics toward implicit and emergent semantics that are organized from the bottom up- folksonomies that include taste neighborhoods, identity hubs, and taste cliques.

Theoretical Background
1. Authentic Identity and Aesthetic Closure
-Contemporary culture is marked by consumeption preferences of diverse demographic categories- a culture of plenitude, in which identities are described using the vocabulary of preferences (McCracken).
-Simmel: the individual is born as an unidentified contents that evolves into identified forms, a truly authentic identity.
-Lacan: the self is a mediated construction in the Other (supported by the Sapir-Whorf hypothesis and feminism)
-Csikszentmihalyi and Rochberg-Halton: consolidates the above two theories- an individual's "symbolic environment" both echoes and reinforces her identity. This is the framework from within which this study works.
-Aesthetic Closure: when an individual's interest can be regarded as unified, interconnected, sharing a common aethetic.
-Diderot Unity: the compulsion of consumers for consistency, to like that which we consume in a consistent and unified manner- provides support for aesthetic closure.

2. Upper Bounds on Theoretical Ideal
The above theory is problematized by a number of factors:
-Goffman's theory that performance is inherent in socialization- we all wear different masks depending on the social context. Identities are viewed as multiplicitous- online profiles provide only a single flat view.
-boyd's theory that, because profiles may be viewed by myriad social circles, the individual is forced to take such publicity in account, resulting in self-censorship.
-boyd also points out profiles are often abandoned over time, resulting in static representations/artifacts of past identity performance.

3. Identity Keywords vs. Interest Keywords
-Examined both broad interests as well as special interests (such as cultural identities).
-Special Interests are usually placed at the top of profiles, while more specific interests are listed later on. The former is used to place individuals into categories, while the latter serve as more detailed descriptors.

Weaving the Taste Fabric
-A single crawl of two SNCs mined information from 100,000 profiles.
-To preserve anonymity, only the text of descriptors was used.
-Because language fragments are often used in specific categories, 90% of them were successfully segmented.
-In the case of general interests, about 75% were successfully segmented, as they often contained more idiosyncratic speech.
-Descriptors were then coded in order to create a common language of categories using sources such as wikipedia's article on subcultures, IMDB, AllMusic, AllRecipes, etc;
-21,000 interest descriptors and 1,000 identity descriptors coded.
-Correlation analysis was then conducted via numeric strength of semantic relatedness.
-Identity Hubs: One's location in the fabric is described in terms of proximity to the various identity hubs, which serve as an index of identities.
-Taste Cliques: cliques of interest based on taste- for instance, "Soccer," "Manu Chao", "tapas" and "Samba Music" would be an example of a Latin taste clique.
-Taste Neighborhoods: larger, more permanent taste cohesions

What is a Taste Fabric Good For?
-InterestMap: taste-based recommendation system- an interactive map where users can input descriptors and receive recommendations based on a navigatable map of descriptors
-Ambient Semantics: facilitates interaction between two strangers who share taste.
-IdentityMirror: makes identity self-management possible
-A dynamic model of taste would take context into account- current events, location


One of the implications I derived from this article was the potential for the creation of true interest-based communities that are capable of a radical clarification and reconfiguration of the networked individual. How addictive it would be, I imagine, to have a system before you powerful enough to know what you like before you're even aware that you did. Oh yeah, check this out:

Pentagon to Merge Next-Gen Binoculars With Soldiers' Brains

'Sup, Big Bro?

Lit Review: An Evaluation of Identity-Sharing Behavior in Social Network Communities (Stutzman)

-Though academic institutions have been working to protect student identities, their work is increasingly being undermined by social networking communities (SNCs).
-The goals of this study were twofold: obtaining quantitative data about SNC participation on college campuses, and analyzing member attitudes pertaining to SNC participation and online identity sharing. This data was gathered from the perspective of an outsider to these communities.
-A random survey of 200 students (38 of whom responded) inquired about the specifics of their involvement in SNCs as well as their feelings regarding online identity sharing. The researcher then created a disclosure matrix for each participant by examining the data made available in their SNC profiles.
-Limitations: Small sample size, internet survey may be biased toward the tech-savvy, outsider status, lexical differences in coding identity elements of the SNCs (favorite movies, sexual orientation, academic status, etc;).

-71% involvement in SNCs: 90% of undergrads, 44% of grads.
-Most popular was Facebook (90% of undergrads), followed by Friendster and MySpace.
-Though participants expressed doubt that their identity information was protected online (2.66 on a 5-point Likert scale), they were nevertheless okay with friends accessing this information (4.55), but markedly less so with strangers (3.15).
-Information of particular interest: location, sexual orientation, political status
-Urges discussion of new identity disclosure threats posed by SNCs.


The very small sample size of this study makes it almost entirely worthless to review, but it is worth noting that academic institutions are working to protect the identities of their students. In another vein, the enormous discrepency between SNC participation by undergrads and that of graduate students suggest that the undergaduate community may possess certain qualities or needs that SNCs fulfill, such as maintaining high school ties.

Seeing as identity disclosure would seem to be a pertinent issue, it would be interesting to reassess users' feelings on the matter now that SNCs have become both mainstream and problematized by media discourse. How is "stalking" defined (it is a commonly used term in Facebook discourse)? What sort of activities and degree of involvement are deemed acceptable by today's norms?

Lit Review: Rhythms of Social Interaction: Messaging Within a Massive Online Network

-Extensive empirical analysis of 362 million message sent by 4.2 million Facebook users over a 26 month period.
-Results found a temporal rhythm that extended across campuses and seasons.
-Nearly all communication occurred between a small proportion of "friends".
-Social Network Research: how people make friends, number of friends, and forms of social support.
-Their understanding of the Poke: Users can ascribe whatever meaning in the context of their relationship to the poker or pokee; described as a "virtual intimate object", an active meaningful social gesture that necessitates reciprocity. Such a situation is a marker of a strong social bond.
-The privacy inherent in messaging/poking frees the act from the pressure of self-presentation.

-Average of about 180 friends per user.
-About half of messages sent to friends at same school, 41% to friends at another school.
-Strangely, over 98% of pokes were sent between people from the same school.
-Reciprocity of messages occurs 59% of the time if senders are at the same school, but only 41% of the time if the sender is from a different school.
-Messaging/poking highest at the beginning of the week, declining drastically Friday and Saturday.
-The rhythm of activity differs from that of a corporate network, where most activity takes place during working hours.
-Trend of messages sent to those at different schools during the daytime, to nonfriends in the same school during the late-night hours.
-No change in rhythm, even during the summer, with the exception of a dramatic increase in messages sent to school friends during school break times.
-Different universities consistantly show either a disproportionately large or disproportionally small number of Facebookers who are active during the weekend.

-Concludes that internet sociality is an activity that frequently occurs alongside work-related tasks rather than as a leisure activity in and of itself.
-Though messages are sent primarily to friends, most friends do not receive messages. What does this say about the strength of Facebook "friend" ties?
-Seasonal variation in same/different school messaging demonstrates the importance of Facebook in supporting geographically distant relationships.


Not sure about the importance of this bit of research (though it could serve useful as a statistical supplement), however it was interesting to discover that such an intensive statistical analysis has been conducted on Facebook. With all the information available on the site, a vast array of studies concerning Internet activity could be conducted, as well as looking at the relationships between group memberships, interests, demographics (political and sexual orientation, gender, "looking for", religion) etc; The study would probably have been much more interesting had they focused on wall posts, the most active form of Facebook communication by far.

Lit Review: Digital Relationships in the ‘MySpace’ Generation: Results From a Qualitative Study (Dwyer)

A qualitative study of online social networking sites and instant messaging.

-CMC reduces the exchange of social context cues, affecting perceptions of truthfulness, interpretation and response to messages, and the formation of impressions.
--Social Information Processing Model (Walther): CMC relies on paralinguistics, slowing the rate at which social cues are received.
-Impression Management (Goffman): The subtle process of controlling another's perception of something by managing the information exchanged in a social interaction. When that something is one's own identity, it is referred to as self-presentation. We interpret others through inference of their roles, derived from the information they or others present to us.
-Social-Technical Gap: The space between what technology can support and what actually happens in the social world.

The Study
-Examined the use of technology to manage relationships, and the ways in which these technologies mediate behaviors pertaining to the management of these relationships.
-The semi-structured interview designed inquired about self-presentation/impression management, pros and cons of these systems, usage and dependency for social communication. It also probed participants for information on how they used the tools provided by this systems in developing new relationships, restricting access, and responding to negative events. Expectations of privacy were also investigated, pertaining to what individuals felt comfortable with sharing and why.
-Interviews were conducted by 6 undergraduates, who interviewed a total of 19 college-aged participants. The transcripts were content-analyzed and coded.
-Participants reported heavy use of communication technologies, heralding their low cost, entertainment value, and convenience.
-Profiles provide the opportunity for impression management. Authenticity plays a large role here- profiles that appear (or are known to be) false or contrived trigger a very negative impression. However, they also discussed the need to create a "cool" persona and intense awareness of how others would perceive their self-presentations. Nevertheless, the act of constructing one's profile was generally considered a fun, entertaining activity.
-As one participant put it, "The defining characteristic of social networking sites is extreme impersonality. The people that one talks to on these sites are not treated as other human beings. They appear more like characters in a story."
-Though privacy concerns have been well-documented, the participants expressed general apathy, countering that they as members are responsible for the content and management of their virtual personas.
-Acknowledged that relationships formed online are superficial in nature.
-General enjoyment of these systems' ability to maintain bonds with those one doesn't see every day, as well as reunite one with old friends.
-Instant Messenger Away Messages: A user is able to monitor others while behind the "barrier" of the away message.
-Comfort level increased as the degree of their own anonymity rose, decreased with the anonymity of others.

Communication technology features (profile, visibility, and identity management) enable interpersonal relationship management (forming new relationships, maintaining existing relationships), which is in turn influenced by individual attitudes (impression management, concern for information privacy).

Questions Raised
How is impression management carried out within CMC?
How to explain the apparent contradiction between privacy concerns and the overwhelming popularity of social networking sites?


Individuals are fluid, not static, and in the act of creating a profile of the self one undergoes a strangely simplified process of impression management. I would like to examine the paralanguage of Internet communities, the ways in which social cues are subtly communicated, as well as the complex ways in which impression management is enacted.

Lit Review: Personal Relationships: On and Off the Internet (Boase & Wellman)

From Computer-Mediated Small Groups to the Internet
-The authors surmise that we are experiencing an era of "networked individualism", shifting from tight-knit, geographically local communities to dispersed, sparsely-knit personal networks.
-Past research on computer-mediated communication (CMC) failed to situate CMC within broader social contexts.
-Past research demonstrates that, thus far, the Internet has had no salient destructive or radical effect on society.

The Social Affordances of the Internet

-supports a greater number of geographically dispersed interactions.
-asynchronous: these interactions can occur between those with very different temporal rhythms.
-rapidity: increased velocity of interpersonal exchange.
-reduced social presence may lessen commitment, complexity, and/or strength of virtual bonds.
-textuality: reduces image-based hierarchies, such as race and class.
-absence of direct feedback --> increased likelihood for flaming.
-societal connectivity: transitive, indirect contact facilitated in a manner that often crosses group boundaries.
-mass messaging: allows contact with multiple social circles.
-HOWEVER, often social networks do not interact, and information diffuses quickly.

Utopianism and Dystopianism
-Utopianism represented by the transformative ideals of Barlow, McLuhan, and others.
-Dystopians warned of the isolating and alienating potential of the Internet, warning ominously of societal decline (Kroker & Weinstein, Stoll) and fractured identities (Turkle).
-Both views are overly simplistic, lacking ethnographic and empirical data.
-Technological Determinism: Attributing causal effects to the technology rather than the way in which people choose to utilize it. For instance, the Internet was the effect rather than the cause of the desire for distant communication.

Contact With Friends and Family- Online and Off
-Though early findings demonstrated negative effects of Internet use, later findings uncovered that Internet use is inextricably tied to one's preexisting personality characteristics (Kraut).
-Other studies used time diaries to examine the everyday practices and effects of Internet use in comparison to offline interaction and found little to no causal relationship (various studies by Katz).
-Evidence exists that demonstrates the correlation between using the Internet to meet new people and decreased TV watching (Kraut, Kiesler, Boneva, & Shklovski).
-The needs of the individual must be examined as causally affecting the everyday usages of the Internet.

Forming Relationships Online
-Research has shown that only a very small percentage of Internet users have formed new relationships online (Katz).
-The Internet serves as an important tool for forming relationships for those who are physically or psychologically disadvantaged, including those with very low self-image (McKenna).
-Often, most users who form relationships online eventually express the desire to meet offline.

Neighboring and the Internet
-A study by Hampton and Wellman on a networked Toronto suburb, "Netville" demonstrated an increase in neighborhood contact, offline visits, and neighborhood ties. Furthermore, the networked residents showed an increase in contact with geographically distant social bonds. The combination of these two effects is defined by the researcher as "glocalization".

Towards a Theory of Networked Individualism
-Since the Industrial Revolution and the rise of mass transit and telecommunications, there has been a shift in social relations that Wellman coins "networked individualism".
-Networked Individualism:
1. Both local and long distance relationships:
Most likely, individuals virtually interact with those they are close to, but just far away from that it is inconvenient to visit.
2. Sparsely-knit personal networks that include densely-knit groups:
Ease in coordinating group events through mass messaging; direct and autonomous nature of the Internet helps one to maintain a large network of contacts with relatively little work.
3. Relationships more easily formed and abandoned:
In this era of mobility and frequent change in environment, CMC helps people stay in touch with those they've left behind. Additionally, the Internet may aid the sociality of those who have trouble forming relationships offline. It also aids in neighborhood connectivity, as demonstrated by the Netville study.
4. Many relationships with people from different social backgrounds:
Devoid of many of the social cues implicit in offline interactions, the Internet facilitates the formation of relationships between individuals of differing backgrounds.
5. Few strong ties and many weak ones:
The asynchronous nature of the Internet means that interactions need not take place in the same space at the same time. This can be an asset in everyday activities (such as shopping) that involve coordination with strong ties, for communicating in a way that lacks intrusion or disruption. Weak ties can be more easily maintained as well, allowing for affirmation of a connection to even the most geographically distant acquaintance.

In conclusion, the authors point to a need to examine the psychological effects of networked individualism, such as information overload and dissatisfaction or, more positively, cognitive flexibility, social tolerance, and increased knowledge.


This paper was published in 2004, around the time online social networking services were becoming mainstream in popularity, and before the advent of social media services such as YouTube. It focuses primarily on e-mail as a form of CMC. Examining society's relationship to the Internet in general, the authors are a bit too empirical and generalistic for my taste.

In developing surveys regarding interaction in online social networks, it would be interesting to examine a subpopulation of those who have recently transitioned into a new environment (such as college alumni) and their use of these tools in maintaining important social ties. How useful and satisfying is this medium? Furthermore, does the Internet actually facilitate group formation, or are these group memberships illusory in nature (serving the purpose of identity development and performance)?

May 1, 2007

Lit Review: Trust in Electronic Environments (Chopra & Wallace)

The article examines the widely dispersed literature on trust with regard to information technologies. Trust is considered a crucial element with regard to social capital, and exists on four levels: the individual (psychological), the interpersonal (one to another), the relational (social glue), and the societal (functioning). The authors put forth a unified definition of trust as "the willingness to rely on a specific other, based on confidence that one's trust will lead to positive outcomes (3)."

When an individual possesses uncertainty or vulnerability in a particular domain, she may seek a dependent relationship with that which one is confident can fulfill that particular need, and this often entails risk on the part of the vulnerable individual. Systems trust (and here we bring the Internet into play) necessitates the establishment of social norms shared by other trusting individuals.

Trustworthiness is defined along four dimensions: competence and credibility, positive intentions, ethics, and predictability/consistency.

There are multiple processes invoked in the development of trust:
-prediction based on past behavior.
-intentionality of the trustee.
-emotional bonding and reciprocity of trust.
-reputation or institutional trust established through the trust of others.
-identification with the trustee (shared goals/values)

Four domains of trust are important in virtual environments:
1. Trustworthiness of information on the Internet.
-Users in need of information place themselves at risk for potential inaccuracy of information viewed. Factors at work include accuracy, bias, methodology, stability (of website address, currency), and validity.
2. Trustworthiness of the information/computing systems themselves.
-This includes interpersonal trust in one's own system as well as societal trust in network structures. One's emotional attitude, or technological bias, toward technology (technophilia/technophobia) is a crucial factor.
3. Trustworthiness of the economic stability of e-commerce.
-Both the buyer and the seller take risks. Reputation hinders on offline reputation and/or online ratings.
4. Trustworthiness of the individuals with whom one interacts in virtual environments.
-Motivations include information, friendship, or simply entertainment, and the truster has the authority to sever the relationship in case of fraudulent identity, abuse, or harassment. One's propensity to trust in this situation is highly dependent on one's technological bias, disposition, referrals, and the context of the relationship.

Computer technology challenges traditional notions of personhood, as machines increasingly take on roles and duties once assigned to people. Thus, it is apparent that trust plays a role in the relationship between human and computer in much the same way as in interpersonal relationships.

Social capital will be a central issue of discussion in my ensuing ethnography, and trust (as well as reciprocity and shared values and norms) is a crucial element.

MySpace: Little trust amongst generations outside the MySpace generation (today's teenagers), however a great deal of trust within that generation due to established cultural norms of participation in the community. Trust on MySpace is diminished by the proliferation of fraudulant identities and predatorial behavior, the finicky nature of the site, censorship, and the popular reputation I have just summated.

Facebook: High degree of trust in other users in one's network, who are usually preexisting relationships. However, growing wariness in the corporation itself due to the enormous amount of data that is collected and recorded and legal reprecussions to naive assumptions of privacy. Its quickly evolving nature, as well as its mass popularization (thus losing its niche identity) have led many to become suspicious of the intentions of the site.

Tribe: Very high degree of trust due to Tribe's subcultural history (as evidenced by its enormous popularity in the San Francisco Bay area). A haven for free speech, Tribe is also locally-based. Furthermore, many of Tribe's users possess shared cultural and social values (such as neotribalism). Users form tribes that often extend to the offline world, and vice versa.

Very high degree of trust, though few strong ties. What ties do exist are based around common interests (such as music genre or production tools) as well as reputation (established through rankings, popularity, degree of involvement on the site). SoundClick's niche audience authenticates the site, gives it legitimacy, and its users are able to share information with disparate others interested in the same thing (shared values and goals... globalizing potential).