Slashdot Mirror
Survey Finds Canadians Support Net Neutrality Law
An anonymous reader writes "A new public opinion survey conducted in Canada finds overwhelming public support in that country for net neutrality legislation. Three-quarters of Canadians believe the government should pass a law to confirm the right of Internet consumers to access publicly available Internet applications and content of their choice — even though most of those surveyed did not know the term 'net neutrality.' The survey was commissioned by eBay." Of course the devil is in the wording. Given the survey's sponsorship, it's unlikely that respondents were presented with examples of the value that ISPs say packet shaping can bring, or asked to weigh such against net neutrality.
Does it mean that bandwidth providers can charge more for high demand customers? Probably fair enough. Does it mean that they can charge end users more for extra speed. No complaints. What is not acceptable is that the owners of the backbone can make deals with "partners" and give them a special rate and stiff other customers. Or they can charge their customers more for bytes from one source than another. The concept of a "common carrier" has served will in the the fields of communication and transportation. Regulation is necessary. I don't want a top down controlled Internet where I am merely a content consumer.
The funny thing is that there are well known effects that skew the effects of polls, among which:
1. People are nice social beings. They tell you what they think you'd like to hear. It's a reflex and enculturation effect that, well, I suppose helps us live with each other. If you know someone, say, likes pink, the nice social reflex is to say "yes, it's a nice colour."
Why does that matter? Most people, even on a perfectly anonymous poll, tend to answer what they think would please the poller. If they're polled by eBay, of course they'll say what they think eBay would like to hear.
2. (Or 1.b.) The wording is very important. If you present a skewed view where option 1 is pure good and option 2 is pure evil, you've already told them what you think on that matter. So they'll subconsciously try to be nice and agree with what you told them you like, regardless of what they actually think on the matter, and regardless of whether they even give a damn at all.
3. All things being equal, there's a bias towards answering more "yes" and less "no". I guess we've all been educated that it's not nice to disagree all the time. So well design polls actually randomize the questionnaires so 50% will ask the question one way, and 50% ask the negative version.
E.g., if half the questionnaires ask "should we stay in Iraq?", the other half must ask "should we pull out of Iraq?", because otherwise you get it skewed towards "yes". If you only ask "should we stay in Iraq?" you'll get your results skewed as some people will vote "yes" just because it's, you know, a "yes."
4. Biased sample fallacies. Was that sample representative, or was it, say, only the people who visit site X? E.g., if you were to make a poll about computers or OSes on Slashdot, I hope you can see how the results wouldn't really reflect what the whole population thinks.
Etc.
Now I don't know how the poll in TFA was done, so I'm not commenting on that. But basically if you want to know what people _think_, then you _don't_ do a poll along the lines of "do you think we should stop ISP extortion?" If you do that, you'll just get a false result that's good for self-shoulder-patting, but won't reflect what they actually vote for in the next elections.
Just saying...
A polar bear is a cartesian bear after a coordinate transform.
In this case, I don't question my neighbor's use of his pipe but the company's selling policy. Appearantly they sell more bandwidth than they can sell. They should not sell him a pipe fat enough to interfere with the TV broadcast.
God beware my neighbor actually uses what he pays for!
We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.