Be carfeul what you tweet
Do we really want to be psychologically analyzed by our tweets?
Using Linguistic Inquiry and Word Count text analysis and the Regressive Imagery Dictionary Dan Zarrella designed TweetPsych to do just that. His blog post introducing the concept, and the beta site itself, explain that users who participate in Twitter conversationally rather than using it for content distribution will receive more insightful results. Despite my account being somewhat information oriented and in the 500-tweet range (1,000 or more tweets also return better results), I ran the posts through to see what I’d get.
The summary showed my highest score in Cognitive Content / Time followed by Primordial, Conceptual and Emotional Content / Temporal References and Abstract thought. My tweets also highlight the past tense and cognitive processes like learning, thinking, knowing, etc.
Furthering the applicability of TweetPsych, Zarrella’s post today introduces characteristic rankings showing a list of the top 20 users in categories including anxiety, oral fixation, and social processes.
Despite the disclaimer that it’s for entertainment purposes only, there is something about this concept that feels off to me. I can’t tell yet if it has more to do with sectioning off aspects of identity when using different social media tools (as in revealing a different component of yourself to Twitter than, say, to MySapce, and different still from Flickr) ... or if it’s because the analysis was approached from a social media marketing standpoint.