>The more interesting part comes when we use the "how often do you use/post" questions to ask not what the distribution of users looks like, but what the distri

>The more interesting part comes when we use the "how often do you use/post" questions to ask not what the distribution of users looks like, but what the distribution of tweets looks like: >Though only around 30% of Twitter users identify as "Liberal" or "Extremely liberal", those users are evidently responsible for around 60% of tweets. Where did this 60% come from? Did the author really translate survey responses along the lines of a user "sometimes" posting political content and uses Twitter "a few times each week" to a direct percentage of all tweets? This isn't even getting into that "using" Twitter doesn't necessarily mean posting tweets.