This again is the case of a couple of links on the same topic piling up in my reader (this tends to happen if you take a pause in blogging :-)):
- Gmail's Spam Filter No Longer Effective?
- 60% of Facebook users consider quitting over privacy
- Poll: 93% say Facebook should make you 'opt-in' to sharing rather than 'opt-out'
- Marketing study: avid gamers average seven hours per day
The commonality between all these articles is that they make statements based on faulty questions (PHD Comics says it best). A website poll is not the same as a scientific study (to name just one the problems - it has a selection bias towards the reader of the particular site - which wouldn't be a problem if the results wouldn't be presented as applicable for the general population). And even if they were scientific studies, the purpose of a scientific study isn't to find the absolute truth! It is to present a hypothesis which doesn't contradict any of the current observation. But it doesn't exclude the possibility that in the future there will be an observation which contradicts the hypothesis, and as such, it must be changed.