BREAKING:
Just minutes after the Slovenian parliament confirmed the new gov. of @JJansaSDS, workers removed the Palestinian flag from the government palace facade & replaced it with Ukraine’s flag
Slovenia now honors a country defending itself instead of massacring civilians
@DanielBent67126@harveymorgan47@GBPolitcs@Survation Once again…
It. Is. Not. About. The. Total. Population.
A sample size’s validity is the same regardless of whether the population is 100,000 or 10,000,000. A sample size is generally seen to be fairly valid at 1000-2000.
It has the same margin of error in each scenario.
@DanielBent67126@harveymorgan47@GBPolitcs@Survation Correct. And if the polls had operated with smaller sample sizes, it is statistically likely would have been even further off.
So what is your point here?
@DanielBent67126@harveymorgan47@GBPolitcs@Survation I’m not sure what you think I’m talking about, and I’m not sure you know what you think I’m talking about either.
All I’ve said is a poll of 500 is less accurate (in other words - has a higher margin of error) than a poll of 2000.
All of this is demonstrably factually correct.
@DanielBent67126@harveymorgan47@GBPolitcs@Survation Sooo… it is less accurate then?…
It has a higher margin of error. We are not talking about accuracy in proportion to sample size, because we are not judging the quality of the poll’s sample given its strained circumstances, we are simply judging the poll, and it’s less accurate
@DanielBent67126@harveymorgan47@GBPolitcs@Survation I’m not denying that this is a good indication of who will win! I said so in my second reply to you, and noted later how I think Burnham will easily sweep it! I am not contesting that, I am contesting your earlier claim that a poll with a 500 sample size is better than a 2000!
@DanielBent67126@harveymorgan47@GBPolitcs@Survation I didn’t say they do usually have polls with 1000+ people. I made no claim to the contrary of any of what you are saying here. I am simply stating the fact that these small polls usually used in by-elections must always be taken with a grain of salt due to their small sample size
@DanielBent67126@harveymorgan47@GBPolitcs@Survation I think Burnham will easily sweep it personally, but that isn’t the point of this. I would push back on the idea that people who are voting Reform know they’re voting Reform, because, nation-wide I agree, but in this specific context it is harder to say and people are more split
@DanielBent67126@harveymorgan47@GBPolitcs@Survation You are the one not getting it. Look up how sample size works. A sample size’s reliability relies on ability to create a representative sample of the population, which analysts and experts generally agree 1000+ is best for. It is NOT to do with proportion to the target population
@DanielBent67126@harveymorgan47@GBPolitcs@Survation But it should still, of course, be called a poll, despite the larger margin of error, bc it’s still a decent indicator, even though it may not be correct
@DanielBent67126@harveymorgan47@GBPolitcs@Survation Eh not really. The validity of a sample size is more about the absolute number that can create a representative sample, which applies to pretty much all elections (generally agreed 1000+ is accurate), but the sample size’s proportion with the population doesn’t matter so much.
My ranking of who performed best tonight on #bbcqt#Makerfield
1.🔴Andy Burnham (Labour)
2.🟢Sarah Wakefield (Green)
3.🔵Michael Winstanley (Tory)
4.🧡Jake Austin (Lib Dem)
5.🩵Robert Kenyon (Reform)
🟢Sarah Wakefield had absolutely no response to the housing/immigration question 😬 her worst, and ➡️Robert Kenyon’s best, moment tonight. However, I still think she was the second best tonight (after Burnham), and he was the worst.