Sunday, May 03, 2009

A Metric Experiment

When I listen to webinars or go to events regarding Social Media & Business, I hear A LOT about the need for solid metrics to see how successful efforts are (ROI). I've railed against this avalanche of opinion because I think statistics are flattening & superficial. It's hard to pick up nuances & the subtlety of complex human behavior by assigning a number to it.

I'm a qualitative researcher which means that I normally gather information through participant observation (being with people), interviews (talking with people) & content analysis (talking with texts).

Qualitative research is sometimes viewed by number-focused people as being more "fuzzy" and it's true that there is more room for interpretation. It's an art as much as a science. But qualititative analysis does have methods and is not merely subjective opinion but is based on theoretical constructs, methodological procedures and previous work in the field.

That said...Twitter is a hard subject for qualitative analysis. While I can refer to individuals, a Tweet or the tone of a conversation on the network, Tweets can be considered personal communication even though they are published on a public forum (unless the user has protected updates). There is not the expectation of privacy but people often behave as if they are writing to specific individuals alone (although they are not Direct Messages). An argument can be made that consent should be asked not only before analyzing someone's Tweets but before even ReTweeting or copying a message. Privacy & consent of internet communication has long been an area of debate and some issues still remain unresolved.

So, the past month I've been searching for alternative ways of using the numbers & statistics that are available for insight into Twitter culture. Lots of time, mistakes & false starts later, I'm ready to throw out one attempt later today. Consider it my practice run!