A discussion of social media research design would be a bit shallow if devoid of the role technology plays in altering any one person’s true reality. Computer-mediated communication, online impression management, and self-presentation tactics are just a few of the concepts often discussed in conjunction with how someone communicates (voluntarily or otherwise) via the electronic medium. Computer-mediated communication is not new but an idea that quickly sprouted when virtual reality began to receive lots of attention in the 1990’s. In 1996 I wrote an article for the American Marketing Association –“Virtual research exists, but how real is it?” – touching on this very issue.
Back in 1996 I stated that online research “provides the researcher with a solution that is sensitive to both budget and time constraints,” a key justification for online research designs today. Because our understanding of how people think and communicate in the online world was cloudy at best, I go on in this article to offer “fast, economical” alternatives to online designs –
- Developing an annual corporate research program (while minimizing costly ad-hoc research)
- Reducing sample size in survey as well as qualitative research (e.g., greater use of mini groups)
- Cutting out research services that are underutilized, e.g., written transcripts or full reports that are rarely read
- Asking for “volume-discount pricing” from research providers
- Moving the research function up the corporate ladder to create efficiencies and focus on less-costly design solutions
While these alternative approaches are as appropriate today as they were 15 years ago, the appreciable advancement of online technology has greatly increased the viability of online research designs. And, although the near silence in the marketing research community concerning computer-mediated communication is a bit deafening, it is encouraging to see MarketTools TrueSample and other initiatives designed to address online respondent fraud.
But what about social media research where validation is difficult? Moving forward, it would be useful for social media researchers (corporate side and consultants) to entertain the ideas espoused by those in communication studies, psychology, computer science, and other disciplines that examine online behavior and attitude formation. For example, Jenny Rosenberg and Nichole Egbert discuss in the Journal of Computer-Mediated Communication their study of the “self-presentation tactics” Facebook users employ to maintain a particular impression on their intended audience. And Stephanie Rosenbloom in her New York Times article, “Putting Your Best Cyberface Forward,” references a variety of sources including Mark Leary, a psychologist at Duke, who studies impression management and explores the images people choose to create of themselves in the online sphere.
In the relatively controlled environment of online survey and community-style research designs, we may be learning to identify whether there is a dog at the other end of the computer or mobile phone screen; but social media researchers are strapped with the more daunting task of understanding how people think and who they choose to become in the virtually social context. This – and its ramifications for research design – are worthy of more dialog.
“understanding how people think and who they choose to become in the virtually social context”
Brilliant!! While I believe we are already in a “privacy by demand” world, the level of anonymity of the source of conversation will also affect the evolution of your “cyberface” – of course this will vary by the objectives for using this media.
Thank you for your comment. Yes, the level of anonymity certainly plays a role, but I wonder if we take on a particular persona in the virtual world..one that we adopt regardless of the venue and sense of privacy. A type of thinking — about people, our relationship with other people, attitudes toward issues — that just becomes who we are online.
Could not agree more. Do we then want to analyze interactions on the social web by persona types? Does that also mean the choice of platform one uses (mainly) defines the level to which they unmask themselves?