Poll: 51% “Concerned With Accuracy” Of Voluntary Census

Roadkill 73 Comments

“Poll based on the responses of 1,177 Canadians. Results are considered accurate within 2.99 percentage points, 19 times out of 20.”
Related…

73 Replies to “Poll: 51% “Concerned With Accuracy” Of Voluntary Census”

  1. Interesting; A polling firm which operates entirely on voluntary public data produces a poll stating Canadians don’t trust voluntary public data.
    Delicious irony.

  2. I’m surprised nobody’s mentioned the disconnect before. The same media organizations that commission weekly polls of 1,000 people or less to generate fully a third of their national political commentary … are suddenly twisted out of shape over the “unreliability” of voluntary data gathered on 1/3 of Canadian households?
    How does that work… exactly?

  3. Fellow Jedi Knights
    Does forcing people to do it make the data more reliable or less reliable?

  4. Would common sense not dictate that the data from a voluntary survey be more acurate from a forced one? If I get one of these “under penalty of law” census forms I am going to be pretty bitter about it and be creative with my answers. How about offer a $25 tax credit with the long form survey? I bet you a contract with a private polling company is more than that.

  8. I think we should all have a federal employee living in with us, reporting on us all the time. This is the only way for the government to be sure that they are getting all the right information. I already have this arrangement with a gentleman from Ottawa, and the flow of information is pretty much unbroken. From this, the government can control my movements and thoughts, and render me harmless to their all-important tasks. And I don’t need to worry about having any incorrect thoughts. This reduces my possible guilt before Judgement Day.

  9. Adding this link again in this thread. This article seems very pertinent to this issue.
    http://www.economist.com/node/16590962
    Denmark has been keeping track of its citizens without a traditional census for decades; Sweden, Norway, Finland and Slovenia, among others, have similar systems. Germany will adopt the approach for its next count, also due in 2011.
    There are two reasons for the change. The first is that computerisation allows statisticians to interrogate databases in a way that was not possible when information was stored on cards in filing cabinets. The second is that counting people the traditional way is getting harder, and less useful. Rising labour mobility and the accelerating pace of societal change mean that information goes stale more quickly than ever. Since its last census in 2001, for instance, Britain has seen hundreds of thousands of immigrants arrive from new eastern European members of the EU. Local governments complain that out-of-date information ignores these newcomers, leaving schools overcrowded, budgets stretched and houses scarce. At the same time, filling in the forms has become more onerous: what started as a short questionnaire about who lived where has turned into an inquisition about everything from toilet and car ownership to race and religion. As a result, compliance rates are falling. The decline of deference raises worries about reliability: last time, when asked about their religious affiliation, 0.7% of Britons replied that they were Jedi Knights.

  10. There seems to be a widespread (non-sequitur) meme where something being illegal makes it not happen. I remember reading about a distraught father insisting his son would be alive today if only texting and driving were illegal. Not true or even likely.
    Another thing is this idea that the median/statisticians seem to have that unwilling people are somehow compelled not only to submit information about themselves, but to submit accurate information about themselves to the government. Not willing to go to jail? Just make some stuff up. Heck, with some of the ridiculous questions on the census, just give them the most useless information possible. For instance, to the best of my knowledge, my most distant relatives were African. And so were everyone else’s.
    Anyhow, I’m rambling a bit here, but my point is that it disturbs me that statisticians who think introducing an incentive to a voluntary census will skew data are completely blind to the fact that “not going to jail” is, in fact, just an incentive and NOT, in fact a guarantee of accurate responses.

  11. A poll asking a single question is a little easier to get people to respond to than a 40 page questionaire.
    But hey… you guys hate this whole scientific method thing anyway, so I can see where you’re coming from.
    Data is the enemy of conservatism.

  12. That’s rich John. ‘scientific method’ with regards to polling.
    Yes the _math_ is scientific, but that’s about it.
    Social ‘sciences’ are not science.

  13. Does anyone recall the reaction of England to the “Doomsdaybook”?
    While this historical document is a unique, detailed, virtual snapshot of a Medieval country….it was not popular….hence it’s name.
    Making this voluntary is sensible/logical—-a “law” that is not enforced/unenforcable needs repealing.
    Note the desperate reaction to by the US Democrats/left to ACORN losing the contract for the US Census.
    The concern is not just losing a means to finance ACORN. The Census translates to political power….having control of the Census IS power.

  14. To repeat what I put in the last thread, Jon. If the statisticians were interested in the scientific method they would be curious to compare the results between a voluntary and mandatory long form census. If they can prove that Canadians need to be forced to provide personal info to the government then they better argue that they have a scientific case to reinstate the mandatory form.

  15. John said: “But hey… you guys hate this whole scientific method thing anyway, so I can see where you’re coming from.”
    So John, you’re obviously a big brained science guy. Tell me, which is harder to control for: self selection bias or “Jedi Knight” style false answer bias?

  16. Voluntary or mandated – the FACT is that surveys filled in by a respondent on their own are not necessarily accurate.
    Rather than deal with the accurate data, respondents will write in any number.
    Surveys require multiple sources to deal with these basic ‘fault-lines’.
    What is puzzling is why the Liberals/NDP are so adamant about the ‘mandatory’ survey. How does making something mandatory guarantee any accuracy or completeness of data?

  17. correction- …then they can better argue that they have a scientific case to reinstate the mandatory form.
    I am sure that the world as we know it won’t end if the worst happens and Canadian en masse refuse to let the bureaucrats, activists and private business snoop into citizens private lives.
    The outrage by the MSM over such a trivial issue is so typical of a “profession” that really needs to get out of their bubble (as Akin hints at in his article).

  18. Like I said at Ezra’s blog, Harper really is an evil genius.
    He picks something that 99.99% of Candians could not care less about and changes it in a way that he knows will drive the chattering classes absolutely bonkers. So they spend their inches of ink and precious airtime blubering along about something that means nothing to Joe and Jane Canuck.
    But it is made clear to Joe and Jane that if this monstrosity winds up in the mailbox, they can happily ignore everything beyond the basic “How many people live there” questions, which are, of course, the only questions that the govt needs to know about.
    Now if only the man were a conservative!

  19. Kate’s “The same media organizations that commission weekly polls of 1,000 people or less to generate fully a third of their national political commentary..” is a huge, huge issue.
    The media uses it’s poll “results” to drive it’s own agenda and to direct public opinion.
    I have had many email exchanges with the Yaffes, Heberts, Simpsons and McQuaigs of Canada and often received pertinent responses.
    However, whenever I suggested media polls were intentionally skewed, drawn out of thin air, sample biased … silence. In fact they all stopped responding about the same time – probably filtered my addy. I guess they are not interested in hearing from their customers.

  20. ET asks, “What is puzzling is why the Liberals/NDP are so adamant about the ‘mandatory’ survey.”
    Because they are all guys like John, ET. They think they are better and smarter than all us idiots who “…hate this whole scientific method thing anyway”.
    I can’t match wits with the great John-like minds of our elite political masters, I only have a frickin’ master of science degree. You too, ET. You’re only a lowly PhD of Smartology, you don’t have a chance.

  21. John – that’s the whole point of the mandated survey – short or long. It’s not scientific.
    How on earth can you say that a self-reported questionnaire, mandated or optional, has any reliable scientific accuracy?
    And what’s your data base to prove that conservatives reject data? Hmmm?

  22. Progressive rulers need the data to properly engineer our society, otherwise they won’t know who to kill off first.
    ‘Murder by Government’ got a little out of hand with Stalin and Mao only because they didn’t have proper cencus information.

  23. ET, I think you probably know why the Liberals/NDP are in such a state over the elimination of the long form being mandatory by force of law. It is because it is a decision made by the present government. The Liberals will stop at nothing in order to regain their “divine right” to power.
    John, you are perfectly free to provide the government with all your personal information you like from what you eat, your sexual habits to your toilet habits. Using the law to force people to answer questions does not guarantee accurate information, as totalitarian regimes have discovered. Indeed, people quickly learn to provide the answer that is required, but that does not make it accurate.
    Oh well, other countries have real problems while we, according to the chattering class, have the elimination of the mandatory long form census. To listen to them Canada is now doomed. Yet I have raised the subject with everyone I meet here in the Fraser Valley of BC and not a single one agrees that the government’s decision is bad. By the way that included university professors, teachers, health care workers, farmers and a whole lot of other workers. I suggest that this is not unique to people in the Fraser Valley.

  24. I don’t think this question requires a degree in Smartology. It’s girl-simple to understand that there is no case beyond “what if” as the basis for the outrage. The MSM just assumes that voluntary = inaccurate (except when it’s one of their polls).

  26. Honestly I think the whole strategy is to get the long form census shortened, like the American version.

  27. Lo these many years ago, the authors of the great little tome How To Lie With Statistics (look it up on Amazon, I’d link but I don’t want to trip the spam filter) pointed out that when surveyers asked people what magazines they read regularly, they got answers that skewed towards the highbrow and urbane – Atlantic Monthly, the New Yorker, the Economist. When they sent people around incognito to ask the same households if they had any old magazines they’d like to donate to charity, they got armfuls of trashy celebrity gossip rags. In short, what people said they were reading didn’t match what they were actually reading.[1]
    This is Statistics 101. Anyone with any training in the field knows to be sure that you’re collecting the information you actually want, rather than what information the subject wants you to have. This is why StatsCan’s head honcho resignation holds no water with me – he has to know that the census in its current form is largely unverifiable anyway.
    [1] for bonus Smartology points, spot the flaw in the previous statement.

  28. [1]That Atlantic Monthly etc are any better than trashy celebrity gossip rags.
    I haven’t ticked off anyone yet today, Phantom, so I had to say it. The bruising will have to start soon because I’m off for my usual afternoon activity of lounging around the pool.

  29. All the govt is doing is decriminalizing the completion of the long census form, should you happen to get one. Nobody has ever been charged with failure to complete the census long form. This is nothing but a bit of housekeeping to decriminalize another lieberal paper crime.
    In the past those who thought the questions were too intrusive either left them blank or filled in bogus info. The StatsCan folks might get fewer completed forms, but the volunteered data will likely be more accurate than the bogus data filled in as protest.
    So I won’t have to tell them I live in a house with 1 bedroom and 99 bathrooms making $100 a year total income as a Jedi-knight shaman.
    The opposition is grasping at straws for just another lame excuse for Harper bashing. Such children.

  30. I never thought about this until you said it, ET. NDP and Librals want the mandatory long census form.
    Probably another of my technicolor dreams, but, maybe, just maybe, some people are going to look at this and think:
    NDP? Mandatory
    Liberal? Mandatory
    CPC? Voluntary
    Hmmmmmmmm.
    A few more CPC votes at election time.

  32. Can I call MSM retards, honestly urnalists’ infer Harper is a fascist dictator for giving Canadians a choice on which form to fill out and that freedom is deemed facism but imprisoning me for not filling out the form is democracy. Retards, stupid utterly contemptable retards. No actually if you are born retarded it’s no fault of your own, the left choose to be farking morons.

  33. 1. Perhaps I keep (and catalog) all my old copies of the Economist and the New Yorker but I give away my copies of trashy gossip mags.

  35. I’m more worried about the accuracy of the mandatory one, after all we do have 20,000 Jedi in our country.

  36. OK, let’s start with a mandaTORY long form census for every member of parliament.
    And make that PUBLIC (and anonymous) so we can see the emporers’ have no clothes.

  37. The more I think of it the more I think that being counted in a census and voting should be a mandatory duty. Kind of a small price to pay for living in a dominion. Of course the census form wouldn’t be as confusing as your income tax form either but I guess that’s another rant.
    btw, I spent a wonderful evening in a mountain lodge by a stoked fireplace reading a 1910 or so Atlas of Canada. A lot of the statistical information was from the census which was interesting in a trivia sort of way but then again I don’t think they cared about daycare spaces for children of left-handed, transgendered, enviro-vegans who wanted to grow chickens in your neighbour’s yard.

  38. Rob R. wins the Smartology points.
    Both “what are you willing to tell me you read” and “what are you willing to give to charity” don’t address the real question, which is what do you actually read. They’re at best weak proxies for the real question, and figuring out how close the the answer you’ve got is to the answer you want is the cornerstone of good survey design.
    I don’t doubt that StatsCan knows this and cross-references the census data to compensate for it. And it is harder to correct for self-selection bias than incorrect info, expecially if the incorrect info is prima facie ludicrous.
    But all that obscures the real issue – should the government be putting people in jail for not coughing up their private information? That’s the issue that gets lost in all this more accurate-less accurate jawboning.
    As an aside, I’ve been needling several of my Toronto urbanite firends lately with “So, a cop putting you in jail for being too close to a fence is out of bounds, but a cop putting you in jail for not telling the government about your sexual preferences, that’s just fine? That’s your thesis?”

  39. Let the 77% (CBC poll) who want the long census form to be mandatory volunteer to do it. For me, I’ll take the short form.

  40. >>after all we do have 20,000 Jedi in our country.
    Don’t forget to extrapolate. It’s actually 100,000 since only 20% of the population is polled.
    And what about the ‘closet Jedi’? It’s actually probably closer to 1 million when that is considered.

  41. The way to figure out what people actually read is to data-mine their Chapters loyalty card. Compare that to what’s in the garbage and compare both to what they self report. That will give you a good idea, but it is still inference.
    The only way to know -for sure- what they read is to observe them 24/7 and take notes. Hence the Liberal/Left love of CCD cameras.
    Shoving a form up somebody’s nose and threatening them with jail gets you 70,000 Jedi Knights in Canada. I’m thinking of being Jedi for the next mandatory census, if there is one.
    And yeah, chucking you in jail for not filling out a form -is- their policy. Two words, gun registry. The gun registry form contains questions about sexual preferences/partners. Ten years in jail for lying, baby.

  42. Today’s edition of the Grope and Flail is comical:
    Apparently we are in the midst of a disaster of epic proportions because Stats Canada will be unable to ‘rely on rich data’ anymore. It is so bad that Liberal shill, Munir Sheikh, has quit his job! What is sure to be felt across Canada in the coming weeks and years (decades?), Stats Canada employees are now ‘demoralized and somewhat adrift’. A shocking revelation by a bunch of overpaid civil servants whose livelihood is entirely dependant on the Census (specifically the long form) is that they don’t like the scrapping idea.
    Another bombshell: The evil Harper government has been steering this venerable institution away from their vital work focusing on social issues, nuanced analysis, and storytelling and has instead told them to… brace yourselves here… ‘stick to the facts.’ Needless to say the gigantic vacuum created by the loss of federally regulated ‘nuanced analysis’ will cause Canadians no end of suffering in the near future. What is happening here is a snowball really… when we lose our source of nuanced analysis AND storytelling all at the same time, what’s next? It would be like Canada Post not having a monopoly on delivering letters. We will become a hollow shell overnight.
    Amazingly, Mr. Clement has not taken this ‘extraordinary rebuke’ to heart and is ‘plowing on’. But of course once he realises that no statistics can happen in Canada with out Mr. Sheikh and his frikken huge moustache, he will see this for what it is a ‘fatal blow’ to his proposed changes. Or if he does follow through with it the government must surely fall and we will have to go to an election to decide the fate of statistics collection in Canada.
    Oh and no Canadian crisis would be complete without input from Ralph Goodale: After inhaling – and effectively halting and then reversing prevailing wind patterns throughout Western Canada, disrupting 8 of the 10 wind farms in operation in Saskatchewan – he said: his party plans to “call Mr. Clement to the carpet and to invite other affected groups to speak out.” NO doubt Canadians across Canada will leap at this opportunity.

  43. You either support the state putting people in jail for not filling out a form, or you don’t. It is really that simple.
    I for one think its ludicrous that anyone would support the state arresting people for simply not filling out a survey.
    Why not send every Canadian the long form on a voluntary basis? Or have it so people can request it if they didn’t receive it as part of the 1 in 3?

  44. So here’s another field of science that pseudo-conservatives choose to ‘disbelieve’ in. Here’s a hint, nobody ‘believes’ in science, you either understand it or you don’t.
    And as for the brilliant argument that it’s easier to correct for self-selection bias (How? Please be specific) then for false answers, this is begging the question. You assume that voluntary answers are more honest on the basis of ‘common sense’, or ‘thinking from the gut’, but this claim lacks evidence. I could just as easily assert the opposite, that the form being mandatory sends a message that the results are serious business, and therefore people will be less likely to fib. Of course, I would never argue like this, because I like to take the effort to not be stupid.
    And one final word, if you’re name-dropping your degrees on the internet, you’re doing it wrong. ‘Nullius in verba’, or ‘on no one’s word’, is a common saying in science. An argument is measured on its own merits, not on the letters which come after the author’s name. You can’t post hogwash and then defend it by saying you have degree XYZ. Well, you can, but you’ll be laughed at.
    One last thing, about how liberals and statisticians are being elitist and insulting when they say your arguments are dumb: tough, your arguments are dumb. This isn’t intellectual pissing contest, this is about using your brain and thinking for yourself. I sense more laziness, complacency, and orthodoxy here than I do stupidity. Most of the people here can do better, but choose not to.

  45. Apparently the Liberal govt, which mandated the mandatory long census, didn’t dare actually charge or jail anyone who refused to fill it in.
    The fuss now, isn’t really about the reliability of statistics but the usual Bash Harper and Promote the Liberal Way.
    As I said, a self-administered questionnaire is not inherently valid. The respondent can fill in anything he wants, even with the most bizarre answers. Oh – and it’s considered unethical to force people to respond to a survey.
    So, this survey has zilch to do with data and statistics and everything to do with, yet another political Bash Harper. It’s endless.
    Statistics can’t come just from self-administered questionnaires and it’s irrelevant whether they are voluntary or mandatory.
    Statistics can lie; we know that. How about the oft-repeated statement that the gun-registry is vital because its data base is tapped into at least ‘5,000 times a day’. When I heard that, I was stunned. Did this mean that in Canada there were 5,000 gun-related crimes a day?! Astonishing. But this statistic is cited to support that registry. What’s the truth?
    The registry is simply connected to the entire data base and if the police even check an address, a car licence, a street address – that hits the gun registry. The gun registry data base is not itself of any interest.
    And, given the fact that in Toronto, gun-related crimes and deaths are now accepted as ‘normal’ – and no-one says a word about ‘how the gun registry was supposed to stop this’…well, we can see that the gun-registry is, in reality, a Liberal Make-Work project. More civil service jobs. More Liberal voters.

  46. The G&M article bob Crooks refers to also says:
    “Statscan employees say privately the agency has already undergone a significant shift in emphasis under the Harper government – away from social issues and towards more economic subjects. It’s also scaled back nuanced analysis – something that made it unique in the world, sources there say. The tone of reports is tilting away from detailed storytelling about Canadian life as workers are ordered to stick to reporting the facts.”
    The above quote is buried half way through an article about the census. I think it reveals the real battle between Harper and Statscan.
    http://www.theglobeandmail.com/news/politics/statistics-canada-chief-falls-on-sword-over-census/article1647348/

  47. Here’s why I suspect that Stats Can is so freaked out over a voluntary census: like so many statist bureaucracies, Stats Can measures success by the sheer number of responses it can obtain, not by analyzing the quality or reliability of those responses. It’s kind of like industries in the Soviet Union that measured success according to sheer tonnage of output. The output was often unusable or of terrible quality, but it served the purpose of keeping bean counters in the Kremlin happy.
    The central planners are upset that they might miss some data that they need for central planning, while not bothering to consider that the whole enterprise of central planning is and always will be a colossal failure.
    The worst aspect of this issue is that even some nominally conservative media types have decided to throw a whining fit over Harper for making this move.
    As an alternative, why not pay people for filling out the census? Try using a much smaller sample, but give each respondent a gift certificate or cash. You’d be amazed at how many people would be only too happy to volunteer their information then. For Pete’s sake, Stats Can charges money for a lot of their information right now; why can’t they pay for it too?
    For that matter, I’d be content if they’d just shut down the whole organization and fire the lot of them. I find it disturbing that this issue even warrants a front page news headline.

  48. Filling in “Canadian” won’t suffice either they want your heritage, English, French, Aboriginal, etc. They’re heavily counting on the hyphenated side of things. Why? Are we not all equal under Canadian law or do they want to play favs, making some chosen ones more equal than others.

  49. People are focussed on the NDP and Liberals……but where is the Bloc? The last poll I saw said 62% of Quebecers supported making the long form voluntary.
    Goodness, what other issue would even exist that the Cons would be the sole owner of in agreement with 60+% of the Quebec voting public?
    We will see what the Bloc says. If the Bloc supports the government in this one then forget those committee inquiries, or if they do allow them then there will more fireworks than one expects.
    I suspect the Bloc would kill it, since they wouldnt want more publicity for the Cons on this issue.
    I dont care either way if the long form is voluntary or mandatory, but I dont belive we are headed to data armageddon if it isnt…..I believe they have ways of mitgating the impact. Although the rise in refusniks and the lack of enforcement makes the survey look like its already voluntary.

  50. And as for the brilliant argument that it’s easier to correct for self-selection bias (How? Please be specific) then for false answers, this is begging the question.
    And this is a straw man, Dunbar. No one here is arguing that. In fact, I said the exact opposite.
    What everyone has been pointing out is that mandatory or voluntary, survey answers can and will be faked. Accuracy don’t enter into it.
    And I notice you’ve cunningly dodged the simple question of whether it’s moral to throw people in jail for failing to provide intimate and largely irrelevant information to the government.
    Also, statistics is not a science. It’s a branch of mathematics, which isn’t a science either.

Proudly powered by WordPress & Skyrocket Themes

MENU

Navigation