Northern Ireland- a failure 99 years on?

jh79

FrancieBrady wrote: »

Which just happens to have a high concentration of Unionist opinion. Poll conducted for a programme heavily slanted in lamenting the breakup of the Union...do you need a diagram here?

Practically even split of Catholic /Protestant in the poll. How does geography skew that?

jh79

maccored wrote: »

and it just so happens a sizeable percentage of the unionist population lives in the east. very handy. :rolleyes:

Was the % of those in this poll living in the East higher than the % of the whole population that live in the East?

FrancieBrady

jh79 wrote: »

Practically even split of Catholic /Protestant in the poll. How does geography skew that?

Which Catholics would be more likely to vote for remaining in the Union, the wealthier more secure ones in the east or those in the west?

Not hard jh79, not hard to work it out.

blanch152

FrancieBrady wrote: »

Which Catholics would be more likely to vote for remaining in the Union, the wealthier more secure ones in the east or those in the west?

Not hard jh79, not hard to work it out.

This is unbelievable stuff.

From the information available, the poll appears to have been conducted perfectly normally and correctly, but you seem to think that because they didn't ask everyone in the Bogside their preference that it is somehow flawed.

FrancieBrady

In other news, wouldn't it be mad to see the old orders on both sides of the border having to coalesce/merge in a final effort to stave off the inevitable?

https://www.newsletter.co.uk/news/politics/dup-starts-2022-ni-assembly-election-campaign-early-calling-for-unionist-pact-to-defeat-northern-ireland-protocol-3158417

maccored

jh79 wrote: »

Was the % of those in this poll living in the East higher than the % of the whole population that live in the East?

in a population that has a high unionist percentage? Are you going to start making sense any time soon?

Does the east have 6 times the population of the west ?

maccored

blanch152 wrote: »

This is unbelievable stuff.

From the information available, the poll appears to have been conducted perfectly normally and correctly, but you seem to think that because they didn't ask everyone in the Bogside their preference that it is somehow flawed.

technically that would be kinda the equivalent to that poll

FrancieBrady

blanch152 wrote: »

This is unbelievable stuff.

From the information available, the poll appears to have been conducted perfectly normally and correctly, but you seem to think that because they didn't ask everyone in the Bogside their preference that it is somehow flawed.

It was meant to appear that way blanch.

What I am saying is, if you polled that demographic with that (tomorrow) question then I am not in the slightest bit surprised by the outcome.

jh79

FrancieBrady wrote: »

Which Catholics would be more likely to vote for remaining in the Union, the wealthier more secure ones in the east or those in the west?

Not hard jh79, not hard to work it out.

And you would exclude them from the real poll? A good poll is meant to be representative of the entire population. Where is the evidence that this poll isn't?

jh79

FrancieBrady wrote: »

It was meant to appear that way blanch.

What I am saying is, if you polled that demographic with that (tomorrow) question then I am not in the slightest bit surprised by the outcome.

We have a conspiracy now too! This is gas.

maccored

FrancieBrady wrote: »

In other news, wouldn't it be mad to see the old orders on both sides of the border having to coalesce/merge in a final effort to stave off the inevitable?

https://www.newsletter.co.uk/news/politics/dup-starts-2022-ni-assembly-election-campaign-early-calling-for-unionist-pact-to-defeat-northern-ireland-protocol-3158417

desperately clawing to bring back the good old days (which they thought they'd get with brexit)

maccored

jh79 wrote: »

And you would exclude them from the real poll? A good poll is meant to be representative of the entire population. Where is the evidence that this poll isn't?

granted its a great representation of the unionist population

Fionn1952

jh79 wrote: »

Purely from a technical point of view I'm curious what you think the flaws are? It took into account their religion and social grade, I'm not sure how geography would skew the data?

Because we are aware there is at least the potential for a geographical factor to have an impact with regards to a vote on Unification. That geographical factor may not be entirely corrected for by religion/social grade (a Catholic of the same social grade living in Belcoo, Co. Fermanagh may have a different propensity to vote a certain way compared with a Catholic of the same social grade living in Cushendall, Co. Antrim), actual proximity to the border being one obvious factor that would influence that. Living in/exposure to a more primarily Unionist community potentially another (being more likely in the East than the West)

So while the survey may correct for religion, gender and social grade, without a representative geographical sample (despite Blanch's protestations to the contrary, the sampling isn't heavily skewed East in a manner representative of population density) and with no mention of correcting for geographic factors (it doesn't seem to have been considered a factor to my reading), it certainly could influence the data.

Obviously quantifying the significance of that influence would require more in depth surveying to put some numbers on the level of geographical influence (indeed with less geographically biased sampling, it could've potentially have been investigated from this poll).

Again, I'm not suggesting that the poll is wrong in its final conclusion (I don't think a majority would vote for Unification tomorrow, in fact I would be a No vote in that particular poll; I wouldn't vote for Unification tomorrow because we don't have a feckin clue what it would look like), but I wouldn't be getting too bogged down in specific numbers from a State of the Union poll (potentially the very topic title introduces a degree of sampling bias) carried out online, or holding said poll up as Gold Standard (to the point of ignoring other recent polls which suggest a much closer margin). As with any reasonable discussion on the topic, we should probably focus on a broad range of information available rather than trying to over egg the pudding with one specific poll because it paints a picture we like the look of to the exclusion of others which paint a somewhat less desirable picture.

jh79 wrote: »

Originally it was the sample size that was the problem until two regular Republican posters who happen to be data analysts debunked that claim.

Still, was expecting a better comeback than the question! Even worse excuse than the "issue" with the poll in the Republic.

Despite your snide comments, I have more than passing experience with statistical analyses though that would've been at a much earlier point in my career. Certainly I don't doubt that I'm a bit rusty, but perhaps address the points I've raised rather than trying to attack my credentials. My points can stand or fall on their own merit; I certainly have enough of a background in the field to not try and argue in defensive of a statement that a sample size of ten would have a margin of error of 10% and a sample size of 50 would have a margin of error of 2%.

FrancieBrady

jh79 wrote: »

And you would exclude them from the real poll? A good poll is meant to be representative of the entire population. Where is the evidence that this poll isn't?

It is representative of the population of those areas.

No conspiracy. Exactly the answer I would have expected.

blanch152

FrancieBrady wrote: »

It was meant to appear that way blanch.

What I am saying is, if you polled that demographic with that (tomorrow) question then I am not in the slightest bit surprised by the outcome.

They polled a representative sample of the population, distributed fairly by age, gender, geographical location and religion. There is nothing to suggest otherwise.

That means the poll accurately represents the views of the population within a certain confidence interval. There is nothing inaccurate about it.

57-43 is a huge gap, it means it is impossible to justify a border poll.

jh79

Fionn1952 wrote: »

Despite your snide comments, I have more than passing experience with statistical analyses though that would've been at a much earlier point in my career. Certainly I don't doubt that I'm a bit rusty, but perhaps address the points I've raised rather than trying to attack my credentials. My points can stand or fall on their own merit; I certainly have enough of a background in the field to not try and argue in defensive of a statement that a sample size of ten would have a margin of error of 10% and a sample size of 50 would have a margin of error of 2%.

I think you've read my post wrong. Francie was trying to claim the numbers were wrong. To both yours and Bonnie Situations credit , ye both debunked that nonsense.

blanch152

Fionn1952 wrote: »

Because we are aware there is at least the potential for a geographical factor to have an impact with regards to a vote on Unification. That geographical factor may not be entirely corrected for by religion/social grade (a Catholic of the same social grade living in Belcoo, Co. Fermanagh may have a different propensity to vote a certain way compared with a Catholic of the same social grade living in Cushendall, Co. Antrim), actual proximity to the border being one obvious factor that would influence that. Living in/exposure to a more primarily Unionist community potentially another (being more likely in the East than the West)

So while the survey may correct for religion, gender and social grade, without a representative geographical sample (despite Blanch's protestations to the contrary, the sampling isn't heavily skewed East in a manner representative of population density) and with no mention of correcting for geographic factors (it doesn't seem to have been considered a factor to my reading), it certainly could influence the data.

Obviously quantifying the significance of that influence would require more in depth surveying to put some numbers on the level of geographical influence (indeed with less geographically biased sampling, it could've potentially have been investigated from this poll).

Again, I'm not suggesting that the poll is wrong in its final conclusion (I don't think a majority would vote for Unification tomorrow, in fact I would be a No vote in that particular poll; I wouldn't vote for Unification tomorrow because we don't have a feckin clue what it would look like), but I wouldn't be getting too bogged down in specific numbers from a State of the Union poll (potentially the very topic title introduces a degree of sampling bias) carried out online, or holding said poll up as Gold Standard (to the point of ignoring other recent polls which suggest a much closer margin). As with any reasonable discussion on the topic, we should probably focus on a broad range of information available rather than trying to over egg the pudding with one specific poll because it paints a picture we like the look of to the exclusion of others which paint a somewhat less desirable picture.




Despite your snide comments, I have more than passing experience with statistical analyses though that would've been at a much earlier point in my career. Certainly I don't doubt that I'm a bit rusty, but perhaps address the points I've raised rather than trying to attack my credentials. My points can stand or fall on their own merit; I certainly have enough of a background in the field to not try and argue in defensive of a statement that a sample size of ten would have a margin of error of 10% and a sample size of 50 would have a margin of error of 2%.

Nonsense.

https://www.assemblyresearchmatters.org/wp-content/uploads/2019/02/population-density-for-PC-2017.jpg

Any population density map of Northern Ireland shows a clear difference between West and East.

jh79

FrancieBrady wrote: »

It is representative of the population of those areas.

No conspiracy. Exactly the answer I would have expected.

So which areas are not represented? Have you actually looked at the data?

FrancieBrady

blanch152 wrote: »

They polled a representative sample of the population, distributed fairly by age, gender, geographical location and religion. There is nothing to suggest otherwise.

That means the poll accurately represents the views of the population within a certain confidence interval. There is nothing inaccurate about it.

57-43 is a huge gap, it means it is impossible to justify a border poll.

Wow, that matches your opinion. Extraordinary poll result for you, you must be delighted.

blanch152

jh79 wrote: »

So which areas are not represented? Have you actually looked at the data?

The Bogside, any border poll should be confined to those living in the Bogside, and the result applied to the rest of Northern Ireland. It is the only fair way.

Fionn1952

blanch152 wrote: »

Nonsense.

https://www.assemblyresearchmatters.org/wp-content/uploads/2019/02/population-density-for-PC-2017.jpg

Any population density map of Northern Ireland shows a clear difference between West and East.

Yes, there is a difference. The poll is not representative of that difference.

Antrim/Down/Armagh have a population of ~1.3 million, Fermanagh/Derry/Tyrone around 600,000. A 2:1 East to West split is not anywhere near what is reflected in that poll.

FrancieBrady

And the lads are 'surprised' the poll hasn't 'featured'.

Much the same reasons an issue of An Pobllacht with the headlines 'A UI Tamarra' wouldn't 'feature' I reckon.

jh79

FrancieBrady wrote: »

And the lads are 'surprised' the poll hasn't 'featured'.

Much the same reasons an issue of An Pobllacht with the headlines 'A UI Tamarra' wouldn't 'feature' I reckon.

Jaysus, ye could hardly claim victory again on another losing poll.

Annd9

blanch152 wrote: »

If you go through the details of the poll, the Brexit ****show doesn't feature much as influencing voting patterns.

One thing that may affect it, because it is everyday life is the vaccination rate. The UK success there may make people think twice.

Brexit doesn't feature much because it's only begining for the north .Whatever way it plays out I guarantee you it will have an effect on the middle ground voters , whether they become pro Union or pro UI depends on the next few years being smooth or continually chaotic.

Do you honestly think people will look back in 5-10-20 years and think about the vaccine rollout when deciding on a UI ? Unless we absolutely **** it up and create a pack of zombies I'm not so sure that will be an issue ....
I'm open to correction but shouldn't we be close enough to the UK on vaccine completion when both injections are finally given ?( I'm not complimenting our system in any way btw)

Fionn1952

FrancieBrady wrote: »

And the lads are 'surprised' the poll hasn't 'featured'.

Much the same reasons an issue of An Pobllacht with the headlines 'A UI Tamarra' wouldn't 'feature' I reckon.

I just find it highly amusing that I'M being accused of bias for suggesting that perhaps the poll isn't perfect and should be considered in combination with a suite of other polling data by the lads half way through running a victory lap celebrating NI staying part of the UK for all time because of one poll....the same lads who were telling everyone that one poll shouldn't be taken in isolation when earlier polling was showing a potential Brexit dividend splitting in favour of Unification.

The same lads getting ultra defensive and refusing to accept any potential for flaws within this poll (despite flaws being highlighted, unsurprisingly as every poll is flawed) being the first to start picking over any minor issue with a poll they don't like.....

jh79

Fionn1952 wrote: »

Yes, there is a difference. The poll is not representative of that difference.

Antrim/Down/Armagh have a population of ~1.3 million, Fermanagh/Derry/Tyrone around 600,000. A 2:1 East to West split is not anywhere near what is reflected in that poll.

Wouldn't Antrim / Derry be North, Down would be East, Tyrone be West and the rest South?

FrancieBrady

jh79 wrote: »

Jaysus, ye could hardly claim victory again on another losing poll.

You claim 'victory' on the basis of a poll?

Wow, pretty revealing stuff there jh79.

jh79

Fionn1952 wrote: »

I just find it highly amusing that I'M being accused of bias for suggesting that perhaps the poll isn't perfect and should be considered in combination with a suite of other polling data by the lads half way through running a victory lap celebrating NI staying part of the UK for all time because of one poll....the same lads who were telling everyone that one poll shouldn't be taken in isolation when earlier polling was showing a potential Brexit dividend splitting in favour of Unification.

The same lads getting ultra defensive and refusing to accept any potential for flaws within this poll (despite flaws being highlighted, unsurprisingly as every poll is flawed) being the first to start picking over any minor issue with a poll they don't like.....

I think you are giving constructive criticism and are realistic about the impact any bias would have.

Francie is just anomaly hunting because of the poor result for unification.

jh79

FrancieBrady wrote: »

You claim 'victory' on the basis of a poll?

Wow, pretty revealing stuff there jh79.

I was referring to the ridiculous excitement over the previous poll showing the majority wanting to remain in the UK. For some bizarre reason you were claiming it as some sort of victory.

FrancieBrady

jh79 wrote: »

I think you are giving constructive criticism and are realistic about the impact any bias would have.

Francie is just anomaly hunting because of the poor result for unification.

I'm saying the exact same thing as Fionn, albeit less eloquently.

Relax jh79...I'm happy with that poll funnily enough, it is exactly as I would expect it to be at this stage and with the parameters.