This is Item 7 of my 8 Critical Success Factors for Virtual Communities post.
Identity and privacy are key considerations for the hosts of virtual communities and for their participants. As more and more of our personal data appears online the user journey from storage of cookies through to what emails are sent are governed by norms and stringent laws including the UK’s data protection act.
Aside from ensuring that a website complies with the necessary regulation, the nuances of identity and privacy on-site are many. A virtual community has several options in terms of identity for its members and the route chosen will determine the appropriate level of identity verification needs be carried out by the software.
Here I consider the participant’s ‘name’ – the publicly visible label they have as they perform actions on-site. These are some alternatives a virtual community has. One or more may be chosen, depending on how each fits with the desired culture and existing norms of the community in question:
- Allow pseudonyms: A site discussing football results might contend with allowing a culture of participation using a pseudonym (e.g. footielover2008). On some occasions those using pseudonyms will interact alongside those using their real names, for example in political forums where participants have a varied level of sensitivity to the implications of what they reveal about themselves and their views online.
- Encourage the use of full names: The extent to which this can be enforced varies. It is relatively simple to set up a private email account using a fictional name and then sign up for a forum using those details. However univadis for example, the online community for medial professionals, requires a GMC number for members to join in the UK, which means they can be fairly certain a member is who they say they are. The LinkedIn community expects full names as it facilitates contacts between professionals (LinkedIn, 2009), and actually loses its potential career-enhancing value to members who are not honest about their identity.
- No login: Allow visitors, or give the option for visitors, to post without logging in (as with Slashdot’s ‘anonymous cowards’). To prevent an overwhelming volume of spam this should be enable with a captcha verification to ensure the user is a human, and teamed with robust moderation to ensure spam and inapropriate contributions are spotted quickly.
The above must be considered carefully as requiring verifiable identities and full names which appear on member profiles affects user behaviour. Anonymity can lead to disruptive behaviour, however on the flip side it can give users a feeling of freedom to speak and express themselves in ways they might not otherwise, if they feel their friendship circle or future employer could easily find and read their contributions.
Personal details can be further embellished by labels indicating user attributes such as gender, sexuality and ethnicity. However this again has implications for the way in which people behave and interrelate:
“Supporters of labeling suggest that this practice is beneficial, as it replicates the information available in physical interactions, allowing for appropriate behaviors and pronoun usage. Detractors claim that these coarsely drawn category labels are quite different from the more subtle cues present in real life and that they hinder the development of rich mediated communities by magnifying preexisting stereotypes, encouraging deception, and automatically sexualizing virtual spaces.”
Don’t ask me why danah always writes her name in lower-case. I’m just going along with it.
This potential impact should be borne in mind before facilitating the various nuances that might be included in visible online profiles. Depending on the type of community, there are also options to consider in terms of allowing / insisting on profile pictures, or the need for members to create avatars which display human or non-human features, such as visual avatars with gender specific characteristics.
Once how a user is indentified has been considered, who can see this information and how they can see it must be decided. The level of personal details to display must also be carefully determined: users should feel their privacy is respected, that they have some power over what is being shared, and that they can alter or remove data held about them. Options for the display of user profiles include making them:
- Visible to all: This will mean all of the profile content can be viewed by a site user who is not logged in and found via search engines
- Visible to logged in members: (The Groupsite platform, for example, currently allows communities to make their forums visible, but prompts the user to login if they want to view profile information)
- Visible to friends or contacts only: (LinkedIn use this level of privacy as standard but sell the ability to view further profile information as part of their premium packages)
- Member can choose: between two or more of the above options: Most social networks now allow the option to choose between a variation of the above, and Facebook now allow the grouping of friends and colleagues so that what is displayed to whom can be determined on an even more granular level
In addition to how individuals’ details should be treated online, their connections to one another should be considered. A website could allow users to identify whether their connection is an ‘online contact’ or a ‘work colleague’, as mentioned above with facebook.
This will become increasingly important as people manage more and more of their professional and personal networks online. Unless we end up with one massive social network /virtual community space online (perhaps beginning with G) which serves all of our professional, social and educational needs, identifying relationship type in a consistent, machine-readable format will be essential for any social website that wants to survive. Users will want the option to integrate their use of different social applications without re-explaining all of their associations, but on the understanding that there will not be damaging crossover between data intended for the eyes of one type of contact they have and not others. The new data portability standards include the ability to tag the nature of a connection (For example, rel-kin, rel-met, rel-neighbor – see more about XFN and microformats here) and should be considered when creating new social websites online.