Gotcha.

And I agree.

He's really backed himself into a corner on this issue. Many many Canadians care about this issue a great deal. A tiny fraction of a minorty of Canadians – all far right conservative/libertarians care about it strongly and want it gone. A much larger minority of Canadians – including people from all parties and ideologies – see the value and the importance in having sound and reliable data and want it mandatory.

Will it change anyone's vote? Not likely. But you can just add it to the pile of decisions that turn people just a little bit more off who might have supported you in the past or the future. It is inevitable with any government that this happens over time. It is surprising when they voluntarily and deliberately try to accelerate that disengagement and disconnect.

" as evidenced by the fact that the only people who are about this story are trolls"

Your evidence for this?

You seriously believe that the only people concerned about this issue are on Maclean's blog comments board?

Or is it that, to your way of thinking, anyone who disagrees with you is by definition a troll?

I love it when people go to their computer, seek out the Macleans site, go to Wheery's thread and then spend the next hour or so, telling the lengthy list of people on the thread that nobody is interested. (makes one wonder what they do with the rest of their day)

I get it… you must be like Aaron's mom, just helping him out by providing some controversial spice to an otherwise dry discourse.

"Your tirade is filled with accusations about people misunderstanding you, despite quoting you directly"

I believe that there is one direct quotation of anything of anything that I've written in this thread. On the other hand, you have posts such as brooster's which are sheer fabrications.

"calling people who disagree with you stupid"

People who think invent and attribute positions to me wholesale or engage in inane name-calling do indeed seem stupid.

"and accusing the of making sh*t up to (alternately) misrepresent you or make themselves feel good."

This is a fact. Well, the motivation is an imputation. But look at brooster's post below, for instance. Purely imaginary.

"overwhelmed by the internal contradictions"

What internal contradictions?

"misplaced hostility"
There's not hostility. My original post was meant ironically, but Liberal partisans on this site are too dogmatic to be able to take light criticism.

"vitriol fueling your responses to others"
Eh. I don't have vitriol so much as ironic contempt.

"question the value of public debate on this topic"
I've never questioned it. I've just suggested that it's not going to have much political impact.

No, your mistake was to have a "dialog" exclusively with positions which you invented. Far from attempting to "dialog with a sentient being", you were attempting to construct an imaginary dialogue with a strawman who exists nowhere outside of your own head. Having a dialogue with a sentient being would involve responding to their actual statements, as opposed to inventing a slew of claims for an imaginary opponent which don't correspond to anything any being which exists outside of your fevered imagination had claimed.

Actually, I post under one name, and, unlike you, I'm not paid by the taxpayer to write.

I take it back. You're not retarded because, frankly, it would be an insult and injustice to the mentally ill to call you one.

You are actually choosing to be who you are i.e. a petulant imbecile. Jarrid and wilson, our other Conbot friends, at least have a sense of humour.

You, on the other hand, seem … well, rather disturbed; alarmingly so, in fact. Smile. Call yourself a jackass once in awhile, it'll loosen you up :-)

I'm a little late to the party, but here's my quick answer:

There are fundamentally two kinds of sampling techniques: truly random ones, and technically non-random ones. Truly random sampling is difficult, expensive, and never really carried out in practice. Non-random sampling techniques, including in one sense those using clustering or stratification, and in another sense those use voluntary mailouts and internet surveys (e.g. Angus Reid) are in general more cost effective, can be controlled, and can reach target populations more easily if need be. All inferential statistics, from basic descriptors (mean with confidence interval for example) to any sort of modeling relies fundamentally on an assumption of random sampling. However, as I mentioned above, non-random techniques are much more common and open up types of research that are difficult randomly, and also lower cost barriers to research a large amount. As such, the assumption of random sampling is met (with increased standard errors) by weighting the non-random sample on relevant demographic factors (e.g. gender, age, income, education). Really the only such data which is generally trusted, has a large enough sample, is sufficiently random, etc. is the mandatory census and without it the viability and usefulness of non-random sampling techniques is greatly hindered, if not rendered nil. Of course, the other problem with the change is that it renders all comparisons with historical data moot.

Uh huh. Looks to me as though you've moved on to the second round of "where are the goalposts?"

Why don't you try it the other way: ask yourself how public policy can be conducted without a census.

Further, I trust that because the figures will be so inaccurate under the new regime, so thoroughly tainted, so statistically impious, that you'll disavow using any of the census numbers collected by a voluntary form, and laugh heartily at the numbers derived from voluntary surveys (where any selection bias would have otherwise been cured by a mandatory long form), in your analyses.