Even if a given online poll is never actually fixed in practice, they're still seriously bad quantitative measurement if they use a self-selected sample of peop

Even if a given online poll is never actually fixed in practice, they're still seriously bad quantitative measurement if they use a self-selected sample of people. You're not even able to say that n% of Engadget readers felt some way, because the population of people who happen to show up while the survey is open is still not randomly selected within that demographic: what day of the week did you post it? Did someone with an axe to grind promote the blog post for you on another site, on facebook, or some IRC channel...on Colbert? Respondents have to be randomly selected, and this is widely known. But since polls generate traffic and engage the readership everybody looks the other way. Even on esteemed sites like HN where everybody should know better - although I have faith that it's not done here for the purpose of generating traffic. I'm an old fart, and "journalism" is an accolade that I would never bestow upon a techno-lust blog. Still, this event is dubious. But it's not like the poll was actual quantitative measurement, so the world didn't lose anything of value beyond whatever dubious value it had to fanbois and rabid detractors alike.