Look, if you're going to do a survey about Twitter habits ( via ), you really should get a sample larger than 135 people. I admit, that in most cases, 135 people is enough to make a study statistically relevant, but we're talking about Twitter here. Twitter has about, what, 100 million registered users? OK, let's be conservative and say 50 million registered. More conservative still? Let's say 10 million active users. 135 out of 10 million?