In an article for Slate, Kevin Gold discusses the "leaky" nature of online privacy. Pattern recognition software has made it increasingly possible to determine a person's identity not by the data that they themselves have shared online, but by what their friends have shared. A researcher from Northeastern University found that only 20 percent of college students needed to participate in filling out profile information online "in order to deduce facts about the nonresponders who friended others," the report states. Using statistics about common characteristics, it's possible to make a "statistically motivated guess as to whether a person belongs to a particular community."     
Full Story