[Slick50]

Originally Posted by dvpower

Take Cork for example. They had around 500 protesters. It looks like this campaign is being fought by hard core activists only and they haven't got the public on board.

The numbers we saw yesterday aren't going to lead to the overturning of the tax.

You're probably right on both counts above, not including "the public on board" bit, but the proof of the pudding will be how many register and pay.

Originally Posted by dvpower

Quote:

You're badly double counting here.

Are you actually a politician?, because you have an uncanny knack for miss-interpreting or miss-representing my posts

But to give you the benefit of the doubt, let me clarify. I was merely using dxhound's calculations on the NO figures to demonstrate how his sum was wrong. Didn't you spot that for yourself?

Originally Posted by dvpower

Speculation? It was survey findings.

It was a survey of 912 people. This poll has a larger sample between the yes's and no's alone, and completely contradicts their's.

[dvpower]

Originally Posted by Slick50

But to give you the benefit of the doubt, let me clarify. I was merely using dxhound's calculations on the NO figures to demonstrate how his sum was wrong. Didn't you spot that for yourself?

No. I've had another look and I can't see what's so wrong with what dxhound says.
He assumes half of the 'Don't Knows' will pay and he has the numbers who have already paid as 10% - not unreasonable numbers - that does bring it to around 60% paying by the deadline.
You may be missing the base of the survey i.e. "All who have not yet paid".

Originally Posted by Slick50

It was a survey of 912 people. This poll has a larger sample between the yes's and no's alone, and completely contradicts their's.

c. 1000 is a pretty standard sample size for a professionally carried out survey. I'm sure you don't need me to tell you the various ways a professional survey is far superior to a boards poll.

[mikom]

Originally Posted by Slick50

It was a survey of 912 people.

Originally Posted by dvpower

c. 1000 is a pretty standard sample size for a professionally carried out survey. I'm sure you don't need me to tell you the various ways a professional survey is far superior to a boards poll.

I'm sure you don't need me to tell YOU how an absence of 78 people from a poll of 1000 can greatly skew its results.

[dvpower]

Originally Posted by mikom

I'm sure you don't need me to tell YOU how an absence of 78 people from a poll of 1000 can greatly skew its results.

Not quite sure what you're getting at - the sample size was 914, not 1000. There is no 'absence of 78 people'.

The technical appendix of the poll has the accuracy of this poll being +/- 3.3% at a 95% confidence level.

[mikom]

Originally Posted by dvpower

Not quite sure what you're getting at - the sample size was 914, not 1000.

So not a pretty standard sample size as you alluded to.

[Slick50]

Originally Posted by dvpower

No. I've had another look and I can't see what's so wrong with what dxhound says.

Enouh said.

Originally Posted by dvpower

You're badly double counting here.

Where did you see the error?

Originally Posted by dvpower

He assumes half of the 'Don't Knows' will pay and he has the numbers who have already paid as 10% - not unreasonable numbers - that does bring it to around 60% paying by the deadline.

I used exactly the same method to calculate the "won't pay's" as dxhound used for the "will pay's". You spotted it quick enough then. None more blind than those who won't see.

Originally Posted by dvpower

You may be missing the base of the survey i.e. "All who have not yet paid".

Has no bearing on the error of dx's calculations

Originally Posted by dvpower

c. 1000 is a pretty standard sample size for a professionally carried out survey. I'm sure you don't need me to tell you the various ways a professional survey is far superior to a boards poll.

Yes, just like all their GE polls.

[Ghandee]

Originally Posted by dvpower

Speculation? It was survey findings.

Speculation/assumption. (dxhound presuming don't knows will pay etc)

More or less the same thing.

[dvpower]

Originally Posted by mikom

So not a pretty standard sample size as you alluded to.

I said circa 1000 was a standard size - this was 9% off 1000, so it is in the region.
They've published their margin of error figures, for a poll of the size they conducted, in the poll documentation. I'm not sure what point, if any, you're attempting to make.

[dvpower]

Originally Posted by Slick50

Enouh said.
Where did you see the error?

36% won't pay, add the 90% that haven't paid

You're adding the 36% of people who 'haven't paid and won't pay' to the 90% who haven't paid - you're counting the 'haven't paid' twice (and then attributing them to the 'won't pay' figure).

[dvpower]

Originally Posted by Ghandee

Speculation/assumption. (dxhound presuming don't knows will pay etc)

More or less the same thing.

dxhound presumed that 50% of the don't knows will pay (and the other 50% won't pay). It is an assumption for sure, but a generous one for the anti-charge side, since most of the people who had made up their minds are going to pay.

Originally Posted by Ghandee

A lot of speculation and hoping is all this is.

I take it that you accept that the opinion poll results aren't just 'a lot of speculation and hoping'?

[Ghandee]

Originally Posted by dvpower

dxhound presumed that 50% of the don't knows will pay (and the other 50% won't pay). It is an assumption for sure, but a generous one for the anti-charge side, since most of the people who had made up their minds are going to pay.



I take it that you accept that the opinion poll results aren't just 'a lot of speculation and hoping'?

I accept it, yeah.

But at the end of the day, its an opinion poll, that's all tbh.
The real figures that matter are the ones who've registered/payed to date, and the reality of that figure is a paltry 9%.

I'm not arguing by the way, just separating opinions from facts.

[mikom]

Originally Posted by dvpower

I said circa 1000 was a standard size - this was 9% off 1000, so it is in the region.
They've published their margin of error figures, for a poll of the size they conducted, in the poll documentation. I'm not sure what point, if any, you're attempting to make.

I'll be back to you about your "margin for error" when the concrete results come in at the end of next month.

[Am Chile]

Originally Posted by Ghandee

I accept it, yeah.

But at the end of the day, its an opinion poll, that's all tbh.
The real figures that matter are the ones who've registered/payed to date, and the reality of that figure is a paltry 9%.

I'm not arguing by the way, just separating opinions from facts.

Currently 9% near the end of february, so just to calculate it, The rate of registration/payment for January was 5%, its currently at 9%, meaning for february an extra 4% registered/payed for the month of february.

[mikom]

Originally Posted by Am Chile

Currently 9% near the end of february, so just to calculate it, The rate of registration/payment for January was 5%, its currently at 9%, meaning for february an extra 4% registered/payed for the month of february.

A drop of 1% in momentum.
Warms the cockles of my heart.

[ebixa82]

Why would you not pay it?