Of course, anyone can file a lawsuit. My hope is the judge dismisses their case and hands the Ottawa citizens a big winfall. Mind you, most of these convoy twats probably don't have a pot to piss in so not sure if they will ever see any money from a judgement.
Court has ruled Emergencies Act during Trucker convoy was unconstitutional
- Thread starter Phil C. McNasty
- Start date
You're the one who said it should have a lower margin of error.
I'm not arguing with the math of how you get that lower margin of error.
I am asking why you think it matters that the margin of error isn't 1.
Again - if this is just you thinking that all polls need to have 10,000 respondents and a 1% margin of error then fine, I will keep that in mind whenever we talk about polls in the future.
It appeared at the time that you were not saying that, you were saying that this poll should be discounted because the margin of error was too big. If your complaint is only about this poll, it is very hard to discount the idea that you are just bringing up the MOE to try and find an excuse to ignore a poll you don't like.
There are polls whereby the MOE is fine if it's higher.
But considering this politically-charged poll, in the context of a specific place, it would warrant a lower MOE in my opinion...and by virtue, a larger sample size.
OK.
So.
Given it didn't have that, but then showed a result where your preferred MOE would make no difference at all to the result, you agree in the end it doesn't matter and you're just nitpicking, right?
How can you predict that a survey of 1600 people would have the same results as one given to 10000 people and then generalize your prediction without even knowing who the respondents were and where they were located?
But if only looking at the numbers, a larger sample size is better in most cases.
Because it doesn't matter in the slightest?
If you decide to do it with 10,000 you get one MOE. If you do it with1600 you get another.
Unless there is an actual flaw in the methodology, they are both perfectly sound, with slightly different margins of error.
Since this is a poll, and not engineering, no one should care about the difference between 1% and 2.5%.
Anything even close to the MoE shouldn't be taken as gospel in a poll.
And one poll is never good enough on its own as I mentioned before.
So acting like they didn't plan it in advance for 1% therefore it is useless or not helpful makes no sense.
They obviously planned it because of the sample size they used. They didn't pick the number of respondents out of the air.
Phil,
I believed from your early reactions that you thought the sample size was too small for 40 Million population size, and wanted to reassure you that it was not. Whether the population is 100,000 or up into 10, 20, 30, or 40 Million, the sample size is roughly the same given the margin of error and confidence level. It doesn't change that much. (100000 requires 1514 people, for example, and 40 million requires 1537.)
But, let's get back to what the survey was saying...
Now, regarding the margin of error, because the ratio they are trying to estimate is so large (75%), they don't need to be very precise. It doesn't make sense to spend the money on a more precise poll.
They are saying that if they did 100 similar polls, they would get an answer between 72.5% and 77.5% in 95 out of 100 of those polls. It's a strong indication that there isn't some vast pool of people out there who feel differently.
You seem troubled by this and want to see that it lies between 74% and 76% in those 100 polls before you believe that well over 70% of Canadians wanted the convoy protestors to "go home now."
That's great for you and for arguing for arguments sake, but not really going to add anything substantial to the knowledge on how the general public felt. They wanted it over.
Have a great day.
Who would you say the "general public" is? I assume you mean those polled. Who were they? Where were they from? Demographics? Political preferences, etc
I simply thought a larger sample size with a lower margin of error would be better and more accurate. That's my opinion. That's what I still maintain.
When I saw the post say 75% of Canadians, my first thought was what was the sample size and what was the MoE used. I obviously know that they wouldn't need to poll millions of people.
If that's what I would have actually thought I would have stated as such. I didn't. You are reading too far into my comments.
Yes.
They planned to interview people enough to do a MOE of 2.5%, which was a perfectly good choice for what they were trying to find out.
They have lots of information on how they do their surveys.
How we poll -
The Angus Reid Institute conducts its polling entirely through online research— surveys or studies conducted on the Internet via a personal computer, tablet or smartphone. We have made the decision to conduct our research solely through online means because traditional telephone based approaches...
angusreid.org
Generally, a survey has a target population that they are trying to draw insights on. That target population will be defined in the survey parameters - could be all adults, just teens, etc, etc.
It looks like Angus Reid aims to target all adult Canadians in their surveys.
Random surveys work by choosing a random subset of the population, and using that small sample to make assessments of the views of the whole target population. By choosing the sample randomly, researchers can get very accurate predictions from a relatively small sample. They will also take care to make sure the sample is balanced across known characteristics like income, etc, so the random sample has a better chance of being representative.
All of this to say that over 70% of canadians wanted the protests over with.
It would be.
But since you can get 2.5% at just over 1,000 people, and it is more than accurate enough for what you are asking, why would they waste their time trying to get 10,000 people for next to no benefit?
You saw a poll you disagreed with and decided to discount it by saying the sample size was too small.
Even now that you know the sample size was fine you refuse to admit you were wrong or that the poll is reasonable.
I have stated that the nature of this poll would be special which is why a larger sample size would be warranted in my opinion.They have lots of information on how they do their surveys.
How we poll -
The Angus Reid Institute conducts its polling entirely through online research— surveys or studies conducted on the Internet via a personal computer, tablet or smartphone. We have made the decision to conduct our research solely through online means because traditional telephone based approaches...
Generally, a survey has a target population that they are trying to draw insights on. That target population will be defined in the survey parameters - could be all adults, just teens, etc, etc.
It looks like Angus Reid aims to target all adult Canadians in their surveys.
Random surveys work by choosing a random subset of the population, and using that small sample to make assessments of the views of the whole target population. By choosing the sample randomly, researchers can get very accurate predictions from a relatively small sample. They will also take care to make sure the sample is balanced across known characteristics like income, etc, so the random sample has a better chance of being representative.
All of this to say that over 70% of canadians wanted the protests over with.
For example, what if they polled people who were directly affected by the protest as part of that smaller sample size?
You can see various polls of different types and the results vary.
This was the article I had responded to: https://www.ctvnews.ca/canada/nearly-three-quarters-of-canadians-want-convoy-protesters-to-go-home-now-survey-1.5781022
Here are others with different results:
https://www.baytoday.ca/local-news/poll-shows-sharp-divisions-over-emergencies-act-8194173
https://angusreid.org/emergency-act-review-freedom-convoy/
https://www.ctvnews.ca/politics/wha...cies-act-according-to-nanos-polling-1.6758343
In all cases I believe the sample size should be larger.
I've also seen similar polls that I agree with but I still think the sample size and/or MoE should be different.
Tell me how that opinion is wrong and why your opinion is right?
I've explained my position here several times and you just rely on others to comment for you.
They use other variables to try to make sure the sample is representative. People think about this stuff.
Survey results also depend on the question asked and the timing - opinion can change.
Those later surveys ask a different question. It seems clear from the first angus reid survey that Canadians typically wanted the protestors to go home (over 70%), but were divided about using the emergencies act (closer to 50% or 60% approval of the use of the act to some degree in the later surveys.)
Don't try to get hired as a pollster, you'll not compete due to the high cost of your proposals.
I'm not really trying to change your mind, just pointing out the flaws in your understanding of how surveys work. Others may benefit from this too.
So is cost of conducting the poll more important than a more accurate outcome? Your main argument here is how much more expensive it would be?
I didn't factor that into my opinion or I would have said that they should have increased the sample size and incurred a higher cost to do so.
They held a reunion today in Ottawa.
Its such a popular movement.
Its such a popular movement.
fyi That photo is misleading.
There were hundreds of people there and they marched peacefully through the city and just didn't hang out in front of the parliament buildings the whole day.
https://ottawacitizen.com/news/loca...ll-for-two-year-anniversary-of-emergencys-act
Here a video showing the march...
https://www.ctvnews.ca/video/c28692...ry?playlistId=1.6278633&__vfz=medium=sharebar
