Actually, the problem is that the questions are very poorly designed.  Some of them break the most basic rules of questionnaire-making, and put this survey into the realm of push-polling. I've been fuming about this questionnaire since completing it last night.  I actually got NDP as my result, but the Green Party was almost even, and a bunch of my NDP-supporting Facebook friends were pegged as "Green Party".

They have compound questions in the survey - that is, questions with more than one question in them, but that you have to answer with only one answer.

Example: "The government should fund daycare instead of giving money directly to parents."  That's actually two statements, not one, and should be separate questions: "The government should fund daycare" and "The government should not give money directly to parents."  Because if you believe the government should fund daycare AND give money directly to parents, there is no way to answer this question.  And if you believe the government should not fund daycare AND they should not give money directly to parents, there is also no way to answer it.

And that makes it a trick question.  Because anyone who votes NDP will answer this question "Strongly agree" or "Somewhat agree" since we all agree that the Conservatives giving a hundred bucks a month to parents was no fit replacement for universal day care.  But they have placed the NDP into the "neither agree nor disagree" choice because the NDP supports both universal daycare AND giving money directly to parents.  But if forced to make a choice, you can bet that the NDP would support universal day care over $100 per month payments to parents, which they roundly criticized at the time.  This question makes it seem like the NDP has no position on daycare at all.

It's a scam question.  Only the Green Party position (fund daycare INSTEAD of payments to parents) or the Conservative Party position (Give payments directly to parents instead of funding daycare) can be reflected in the answer to the question.  The likert scale they use doesn't adequately represent the position of the NDP or Bloc on this question.

Another question that's like this is this one: "Environmental regulation should be stricter, even if it leads to consumers having to pay higher prices."  This is another compound question because it assumes that the effect of stronger environmental regulations is that consumers will have to pay higher prices.  This is mirroring the Green Party's position of using market solutions to environmental problems.

So, again, the only answers that you can agree or disagree with are the Green Party's (increase environmental regulation and increase prices), or the Conservative Party's (do not increase environmental regulation so that prices will not increase).

Whereas NDP supporters are likely to support strict environmental regulation, but do not necessarily think that increased prices always have to follow.  I, for instance, support strict environmental regulation, and also support cost controls on basic levels of consumption of necessities, and increased funding for social programs and subsidies for people who cannot afford the necessities.

So anyone who votes NDP either has to put "somewhat agree" or "strongly agree" with this statement to reflect their support for increased environmental regulations, but then they get counted as a Green Party supporter because this survey doesn't give the NDP credit for their support for increased regulation.  This is due to the question being a compound question that also assumes that all people should have to pay higher prices for everything as a result of increased environmental regulations, something the NDP does not support.

Not to mention that they soft-pedaled the positions of the Conservative Party on many of the questions too, as outlined by Vansterdam Kid above.  I think it's a travesty that this survey is being touted by the CBC as a guide for people to help them figure out who to vote for.  It is an extremely poorly-designed survey that skews the results for progressives in favour of the Green Party.  And call me a conspiracy theorist, but I don't think that's an accident.  I doubt that highly-trained research academics somehow don't understand the most basic rules of creating unbiased survey questions!