Each city, Selfiecity posits, has its own selfie style. The faces in Moscow seem to trend older than those in Bangkok (relatively speaking – we're still just talking about 20-year-olds here). Sao Paolo seems to produce the cheekiest grins (and why wouldn't it? These people are in Sao Paolo!) The ladies of New York, meanwhile, appear to prefer the subtle head turn (just a hint of profile) over the dramatic ear-to-shoulder head tilt.
This synopsis comes from a delightful analysis of thousands of actual selfie photos from Moscow, Bangkok, Sao Paolo, New York and Moscow (where, yes, there is notably more selfie brooding). The project was created by CUNY's Lev Manovich, who also brought us an earlier playful study of Big Data in Instagram from various cities, alongside half a dozen other data scientists, curators and researchers.
The project relied on both Mechanical Turks to sort and assess the images, as well as automatic analysis of the degree of head tilt, nose position, and emotional expression on each face. The dataset originally included 20,000-30,000 Instagram images from each of the five cities. By the time they were sifted for verified selfies, the project yielded 640 images for each city. You can play with all of them yourself here, in the "selfiexploratory."
The project aims to investigate the medium with a "mix of theoretic, artistic and quantitative methods." In the latter group, here is the distribution of frowns-to-smiles in Sao Paolo at left, and Moscow at right:
The project views social media as a vehicle of voluntary interpersonal communication, thus becoming a study of human behavior that could as well be approached from [the] perspective of sociology or communication studies... And it all started so innocently: on January 16, 2011 Jennifer Lee from Oakland shared her selfie on Instagram, and on January 27 she was the first user to tag this selfie as #selfie.