The Polls Thread

[Statsman]

Just a quiet corner for the poll anoraks among us to discuss the finer points of statistical analysis, margins of error, secular trends and temporary blips, and other matters of great moment.

And we had a poll in the Sindo which showed, once again, the relative stability of party standings since the GE, with the big three of SFFG sharing 60% of the decideds

Screenshot 2025-10-07 at 14.22.41.png (55.96 KiB) Viewed 1101 times

https://bsky.app/profile/gavreilly.com/ ... eyjbtlvk2z

Given their relatively low slice of support, Aontú's + 2 and the Indos and others -2 are the only movements that fall outside of the MoE here, but are they a blip or a trend? We'l have to wait and see.

[midlander12]

Just a quiet corner for the poll anoraks among us to discuss the finer points of statistical analysis, margins of error, secular trends and temporary blips, and other matters of great moment.

And we had a poll in the Sindo which showed, once again, the relative stability of party standings since the GE, with the big three of SFFG sharing 60% of the decideds

https://bsky.app/profile/gavreilly.com/ ... eyjbtlvk2z

Given their relatively low slice of support, Aontú's + 2 and the Indos and others -2 are the only movements that fall outside of the MoE here, but are they a blip or a trend? We'l have to wait and see.

Don't tell me the Big Four aren't a thing now!

[Statsman]

Don't tell me the Big Four aren't a thing now!

Don't mention the war.

[Irish History]

Just a quiet corner for the poll anoraks among us to discuss the finer points of statistical analysis, margins of error, secular trends and temporary blips, and other matters of great moment.

And we had a poll in the Sindo which showed, once again, the relative stability of party standings since the GE, with the big three of SFFG sharing 60% of the decideds

https://bsky.app/profile/gavreilly.com/ ... eyjbtlvk2z

Given their relatively low slice of support, Aontú's + 2 and the Indos and others -2 are the only movements that fall outside of the MoE here, but are they a blip or a trend? We'l have to wait and see.

What about the drop in support for Fianna Fail?

Aontú's + 2 and the Indos and others -2 are the only movements that fall outside of the MoE here,

That poll states that Fianna Fáil at 20 (-2) also fall outside your same margin of error.
.

[Statsman]

What about the drop in support for Fianna Fail?



That poll states that Fianna Fáil at 20 (-2) also fall outside your same margin of error.
.

Marginally, yes.

[Irish History]

Marginally, yes.

I take then that you are a supporter of Fianna Fail and did not want to highlight the FF drop in support within the same MOE..

[Statsman]

I take then that you are a supporter of Fianna Fail and did not want to highlight the FF drop in support within the same MOE..

You can take whatever you like, but you'd be wrong.

On the whole, I'm a Soc Dems supporter.

The margin of error is not a fixed number; the higher your share of the vote, the nearer you are to the declared MoE, and vice versa. Aontú's MoE is much smaller than the baseline.

[Irish History]

You can take whatever you like, but you'd be wrong.

On the whole, I'm a Soc Dems supporter.

I don't want you to feel I'm having a go at you personally because I'm not.

My ire was with the Fianna Fail drop in support 20 (-2) not being highlighted when it is within the same MOE you mentioned for Aontú's + 2 and the Indos and others -2.

[Statsman]

I don't want you to feel I'm having a go at you personally because I'm not.

My ire was with the Fianna Fail drop in support 20 (-2) not being highlighted when it is within the same MOE you mentioned for Aontú's + 2 and the Indos and others -2.

They are not the same margin of error.

[Irish History]

They are not the same margin of error.

The poll clearly states they are.

Fine Gael 21
Sinn Féin 21 (+1 in four weeks)
Fianna Fáil 20 (-2)
Social Democrats 8 (+1)
Aontú 6 (+2)
IndIreland 4
Labour 4
PBP-Solidarity 3
Greens 1 (-1)
Inds/others 11 (-2)

POLL: Sunday Independent/Ireland Thinks
(Oct 2-3, MoE 2.3%)
.

[Statsman]

The poll clearly states they are.

Fine Gael 21
Sinn Féin 21 (+1 in four weeks)
Fianna Fáil 20 (-2)
Social Democrats 8 (+1)
Aontú 6 (+2)
IndIreland 4
Labour 4
PBP-Solidarity 3
Greens 1 (-1)
Inds/others 11 (-2)

POLL: Sunday Independent/Ireland Thinks
(Oct 2-3, MoE 2.3%)
.

The MoE stated in any poll assumes one party on 50%; their MoE would be, in this case, 2.3%.

For parties on less than 50%, there's a sliding scale, and the further you are from 50%, the more your MoE diverges from the baseline.

[Irish History]

The MoE stated in any poll assumes one party on 50%; their MoE would be, in this case, 2.3%.

For parties on less than 50%, there's a sliding scale, and the further you are from 50%, the more your MoE diverges from the baseline.

Would you stop - you are now like a cat trying to cover up its sh1t on a concrete floor - not going to happen.

You already conceded in post #5 that the poll states Fianna Fáil at 20 (-2) ALSO fall outside your same margin of error.
.

[Statsman]

Would you stop - you are now like a cat trying to cover up its sh1t on a concrete floor.

You already conceded in post #5 that the poll states Fianna Fáil at 20 (-2) ALSO fall outside your same margin of error.
.

Marginally, whereas Aontú and II, because of the way things work, fall significantly outside the MoE.

If you can't deal with your own lack of understanding in any way other than personal abuse, on my foes list you'll go.

[Irish History]

Marginally, whereas Aontú and II, because of the way things work, fall significantly outside the MoE.

If you can't deal with your own lack of understanding in any way other than personal abuse, on my foes list you'll go.

You are trying to split hairs now - the poll clearly states the following because that is how it wants people to understand the following.

Fine Gael 21
Sinn Féin 21 (+1 in four weeks)
Fianna Fáil 20 (-2)
Social Democrats 8 (+1)
Aontú 6 (+2)
IndIreland 4
Labour 4
PBP-Solidarity 3
Greens 1 (-1)
Inds/others 11 (-2)

POLL: Sunday Independent/Ireland Thinks
(Oct 2-3, MoE 2.3%)

[Brabantje]

Would you stop - you are now like a cat trying to cover up its sh1t on a concrete floor - not going to happen.

You already conceded in post #5 that the poll states Fianna Fáil at 20 (-2) ALSO fall outside your same margin of error.
.

He's 100% correct. The further away you are from 50% the less of a margin of error.

This is why a party on 1% such as the Greens can't have an MoE of 2.3%.

Because they can't be on minus 1.3%.

Think about it that way.

There's also confidence levels - typically 95%. This means that if you carried out the same poll 20 times you'd be within the confidence interval 19 times.

[Brabantje]

A good explainer on margin of error here.

https://yougov.co.uk/politics/articles/ ... rgin-error

However, it is important to note that a margin of error applies to the whole sample. All pollsters who are members of the British Polling Council, like YouGov, will publish computer tables showing the detailed results of the poll, which will include crossbreaks breaking down respondents by age, gender, social class, region and other demographics. While these offer great insight into patterns of public opinion, they do, naturally, have smaller sample sizes. For example, a poll of 1000 people will normally have around 500 men and 500 women, and the margins of error on those figures will be around +/- 4%

For smaller demographic groups, sample sizes are even smaller and these bring with them much larger margins of error. For example, a poll of 1000 people would have a margin of error of +/- 3%, but if there were only 100 Scottish respondents within that poll the Scottish figures would have a margin of error of +/- 10%. This means unless the difference between what Scottish respondents said was different to what the rest of the sample said by more than 10 percentage points, it would not be statistically significant. It could just be random error.

[Brabantje]

To put this another way, an MoE of 2.3 related to a sample size of around 1900.
Each 1% equates to 19 people.

The MoE on a sample of 19 would be ±23% or roughly 4 people.

For the example of Aontu they went to 6% or 114 people. The MOE on a sample of of 114 is 9.3%, or roughly 11 people. 11/1900 is 0.56% of the whole sample. Which is why a 2% move is outside the MOE.

Contrasted with FF, the MOE on 20% or 380 people is 5.1% or 19 people, which is 1% of the sample. Which, while outside the MOE, is much less so than Aontu's relative movement.

[Statsman]

To put this another way, an MoE of 2.3 related to a sample size of around 1900.
Each 1% equates to 19 people.

The MoE on a sample of 19 would be ±23% or roughly 4 people.

For the example of Aontu they went to 6% or 114 people. The MOE on a sample of of 114 is 9.3%, or roughly 11 people. 11/1900 is 0.56% of the whole sample. Which is why a 2% move is outside the MOE.

Contrasted with FF, the MOE on 20% or 380 people is 5.1% or 19 people, which is 1% of the sample. Which, while outside the MOE, is much less so than Aontu's relative movement.

This!

Bookmarked for future reference.

[Irish History]

He's 100% correct. The further away you are from 50% the less of a margin of error.

This is why a party on 1% such as the Greens can't have an MoE of 2.3%.

Because they can't be on minus 1.3%.

Think about it that way.

There's also confidence levels - typically 95%. This means that if you carried out the same poll 20 times you'd be within the confidence interval 19 times.

Hang on a second - is the margin of error referring to the individual percentage support for parties or to the amount of people, be it the 100 - 1000 - 10000 sample who responded, meaning the poll itself?

[Brabantje]

Hang on a second - is the margin of error referring to the individual percentage support for parties or to the amount of people, be it the 100 - 1000 - 10000 sample who responded, meaning the poll itself?

The margin of error as stated at ±2.3% applies to the whole poll (statistically the population) and equates to roughly 1900 people.

The moe on individual parties is smaller - often much smaller the further they are from the 50% mark.

E = 1/(SQRT sample).

[Irish History]

The margin of error as stated at ±2.3% applies to the whole poll (statistically the population) and equates to roughly 1900 people.

The moe on individual parties is smaller - often much smaller the further they are from the 50% mark.

E = 1/(SQRT sample).

Exactly - which is why I asked.

The results of the poll as presented to people does not split hairs as far as the moe on individual parties being smaller- it just shows margin of error 2.3% for the whole poll - up 1 or 2 or down 1 or 2 for the individual parties.

So my point stands in post #4.

Fine Gael 21
Sinn Féin 21 (+1 in four weeks)
Fianna Fáil 20 (-2)
Social Democrats 8 (+1)
Aontú 6 (+2)
IndIreland 4
Labour 4
PBP-Solidarity 3
Greens 1 (-1)
Inds/others 11 (-2)

POLL: Sunday Independent/Ireland Thinks
(Oct 2-3, MoE 2.3%)
.

[Brabantje]

Exactly - which is why I asked.

The results of the poll as presented to people does not split hairs as far as the moe on individual parties being smaller- it just shows margin of error 2.3% for the whole poll - up 1 or 2 or down 1 or 2 for the individual parties.

So my point stands in post #4.

Fine Gael 21
Sinn Féin 21 (+1 in four weeks)
Fianna Fáil 20 (-2)
Social Democrats 8 (+1)
Aontú 6 (+2)
IndIreland 4
Labour 4
PBP-Solidarity 3
Greens 1 (-1)
Inds/others 11 (-2)

POLL: Sunday Independent/Ireland Thinks
(Oct 2-3, MoE 2.3%)
.

No, your point doesn't stand. At all. Because it fundamentally misunderstands the concept.

MoE is a mathematical concept that is ENTIRELY related to sample size. The sample size for parties is of its very nature smaller than that of the poll. Unless you live in China.

The 2.3% relates to the whole poll and would be spread across the parties in proportion with their individual sample sizes.

So talking about a party moving a few points up or down in relation to poll MoE is generally statistically insignificant. This isn't helped by general (deliberate or otherwise) ignorance in the media of how polls work.

Again, margin of error is distributed across the parties.

[Irish History]

No, your point doesn't stand. At all. Because it fundamentally misunderstands the concept.

MoE is a mathematical concept that is ENTIRELY related to sample size. The sample size for parties is of its very nature smaller than that of the poll. Unless you live in China.

The 2.3% relates to the whole poll and would be spread across the parties in proportion with their individual sample sizes.

So talking about a party moving a few points up or down in relation to poll MoE is generally statistically insignificant. This isn't helped by general (deliberate or otherwise) ignorance in the media of how polls work.

Again, margin of error is distributed across the parties.

Wrong - my point in post 4 does stand and was conceded in post 5 by the person I made the point to.

I know 2.3 refers to the whole poll, which is why I asked you to clarify what you meant - because you seem to think that I think the 2.3 moe is subtracted from the individual percentage support for parties. That's what you wrote in post 19- not me.

This is why a party on 1% such as the Greens can't have an MoE of 2.3%. Because they can't be on minus 1.3%

It is you who does not understand and jumped in out of context to post 4.

The poll clearly states the following because that is how it wants people to understand the following.

Fine Gael 21
Sinn Féin 21 (+1 in four weeks)
Fianna Fáil 20 (-2)
Social Democrats 8 (+1)
Aontú 6 (+2)
IndIreland 4
Labour 4
PBP-Solidarity 3
Greens 1 (-1)
Inds/others 11 (-2)

POLL: Sunday Independent/Ireland Thinks
(Oct 2-3, MoE 2.3%)
.

[Statsman]

Wrong - my point in post 4 does stand and was conceded in post 5 by the person I made the point to.

I know 2.3 refers to the whole poll, which if why I asked you to clarify what you meant - because you seem to think that I think the 2.3 moe is subtracted from the individual percentage support for parties. That's what you wrote in post 19- not me.

It is you who does not understand and jumped in out of context to post 4.
.

My 'concession' was that FF's drop was marginally outside *their* MoE whereas Aontú's jump was significantly outside theirs.

As for the rest, all you've done is shown yourself to be more or less innumerate. No shame in that, really.

[Irish History]

My 'concession' was that FF's drop was marginally outside *their* MoE whereas Aontú's jump was significantly outside theirs.

As for the rest, all you've done is shown yourself to be more or less innumerate. No shame in that, really.

Don't lie - you did not differentiate between significantly and marginally.

You do now, after I questioned you why you did not highlight Fianna Fail who also dropped 2 percent.

You conceded my point in post 4 in your post 5, even if you added marginally.

You then tried to split hairs, which as far as the way the poll is presented to the people, does not.

The poll clearly states the following because that is how it wants people to understand the following.

Fine Gael 21
Sinn Féin 21 (+1 in four weeks)
Fianna Fáil 20 (-2)
Social Democrats 8 (+1)
Aontú 6 (+2)
IndIreland 4
Labour 4
PBP-Solidarity 3
Greens 1 (-1)
Inds/others 11 (-2)

POLL: Sunday Independent/Ireland Thinks
(Oct 2-3, MoE 2.3%)