Demographic bias exists in Social Networks

There was some controversy last June when uber Ph.D. student Danah Boyd wrote an essay for her blog Apophenia reporting on four years of ethnographic research on social networks. That essay claimed there was a demographic segregation by class between Facebook and MySpace. The process and results were both challenged, prompting a lot of discussion among bloggers and traditional press (Mashable looked at the most popular artists, for instance, and Danah offered a response to criticism a month later.). The online discussion was especially interesting given the utopian glow around Facebook application platform at the time.

Boyd and Nicole Ellison were guest editors of the Journal of Computer Mediated Communication for the past issue, contributing a piece on history of social networks. Among the works published in the October issue of JCMC is a formal study by Eszter Hargittai (Northwestern) entitled, “Whose Space? Differences Among Users and Non-Users of Social Network Sites.”

Hargittai noted that when looking at social networks in general, as an aggregation of all site data, there is only a relationship between gender and SNS. Context of use and experience are also factors. However, when the stats of specific sites are considered, there are pronounced demographic differences in who uses which site:

In particular, Hispanic students are significantly more likely to use MySpace than are Whites in the sample, while Asian and Asian American students are significantly less likely to use MySpace. Additionally, the latter group is much more likely to use Xanga and Friendster than are Whites, a practice that may be due to these services’ popularity in the Philippines, Singapore, Malaysia, and Indonesia (boyd & Ellison, this issue), where—given the immigrant nature of the sample—many students may have extended family and friends from earlier parts of their lives.

Regarding parental education, students whose parents have lower levels of schooling are more likely to be MySpace users, whereas students whose parents have higher levels of education are more likely to be Facebook users. These associations are not evident when aggregating all social network site usage, probably because the various relationships cancel each other out.

It is significant bit of justification to Boyd’s informal observations and evidence that differences exist. Like the early explorations of internet relationships that inherently dubbed online connections as “weak,” however, the volatility of Internet communities should temper the results. The distinctions in the short histories of the major social networking sites forces us to stop short at viewing these demographic differences as either inherent in the services or indicative of a dangerous divide. They are merely the current state of evolution in SNS, a snapshot of what is happening now.

The next questions become:

1. If ignored, will this study ultimately show a phase in natural development toward demographic equality, or are all SNS destined to have demographic identity?
2. Can or should SNS organizations react to this news by encouraging representative diversity?