[Exclusive poll] Maxime Bernier's tweet very divisive among Canadians

According to an exclusive poll conducted using Google Surveys (explanations below), the most infamous tweet from Maxime Bernier, former CPC member and now founder of the new People's Party, is highly divisive. While a short plurality actually agrees with him, you have almost as many people who don't or who simply aren't sure.

Here are the results:

Canadians seem very divided on this statement. Men were slightly more likely to agree with the statement that women, but the difference was small. Also, a very slight and positive correlation between the response time (in ms) and agreeing was there, but this is really weak. At the end however, a small plurality actually agrees with the tweet (and as discussed below, the methodology used here might underestimate this proportion). Once we account for the margins of error however, we can't be sure which option is preferred.

I personally think that showing such divided results is in itself interesting. Given some of the reactions in the medias, you'd have thought his tweet would be much less popular.

Let's talk methodology a little bit if you don't mind. I'm sure many of you will scream at the fact I'm using a survey done through Google Survey. But just to be clear, a poll conducted using Google survey is not the same as a non-scientific poll done on my website, Twitter or on the main page of a newspaper. Google Survey does sampling. In the 2012 US presidential election, Google Survey ranked 2nd. There is a dedicated page on electoral surveys. Look, I'm not trying to argue with you that a Google survey poll is better than one done by Mainstreet or Ekos. But this is a fairly cheap option allowing me to get data on questions not asked by regular pollsters. Using it for an election like I did for the Ontario one is nice (and results weren't too bad) but it doesn't provide me with anything (well, given the lack of polls in Quebec during this election, maybe it would be useful but that's another discussion). But using it for more "original" questions is interesting. At least I think so. If you can't get past the methodology, then so be it.

Another criticism some of you might have is the lack of context given to the tweet. You'd be right. I purposely didn't indicate it was from Maxime Bernier as to avoid having people judging the messenger instead of the message. Bernier isn't the most popular guy out there. The specific question I asked was "Do you agree or disagree with the following tweet from a Canadian Member of Parliament?" and then I showed the tweet (I only showed a picture of it with the name and number removed). Here is the tweet:

I only included tweet number 3 out of 6. Again this was on purpose, I only took the most controversial tweet. But this tweet, isolated from the first two in particular, should actually introduce a bias against supporting the statement. I mean, Bernier spent two tweets defending diversity, yet I ignored those tweets.

I collected 1001 responses from September 18th to 21st.

Do I have any agenda doing this? No. Again, I believe Google Surveys allow me to provide unique polls once in a while for cheap. I wanted to do it on a topic that would be interesting. I thought about the notwithstanding clause in Toronto but we already got "real" polls for that.

So, here you have it. I have the raw data and will share with anyone asking for it.