longweekend58 wrote on Mar 2^nd, 2013 at 2:41pm:


No, longy, resorting to ad hominem, erodes your credibility. Even Newspoll concedes that changes to telephone usage has brought into question the reliability of landline polls. However, you have made a good point, low margins of error would suggest this behavioural change may not be having the influence we might suspect. Or it might be that several biases are cancelling each other out.

It's also worth checking out if the methodology used at election time is the exact same methodology used throughout the year. Polling companies have a vested interst in getting these ones as accurate as possible, it wouldn't at all surprise me if a more resource intensive methodology was used just before the election.

Outside the possibility of bias in the methodology, there is also the issue of what questions are asked, what areas are sampled, the timing of the polls and the results that are published.

Just one example is the Rudd v Gillard poll. Why just conduct it in Queensland and why not publish the 2PP result.

MOTR wrote on Mar 2^nd, 2013 at 3:20pm:


once again MOTR. the questioning of the methodology is one thing but if they continue to be stunningly accurate does it matter if they stir fry the entrails of a chicken and examine them to get their answers?  and the answer to your last question is simple: they were commissioned (and paid) to do just that survey. You do realise they dont do all this for free do you?

I just find it curious and a bit humorous that you question newspoll's methodology. It is well known, quite public and industry standard and yet you question it, despite its proven accuracy. But on the other hand you accept without question the methodologies of climate scientists whose methodologies have been crucified by other scientists and whose results and predictions have a near 0% success rate.

Just something to think about.

What I'd also like you to consider, longy, are the circumstances under which a bias comes into play. If one group of people are over-represented in a poll, when does this show up in the poll results. Well, if the group of people over represented in the poll have views similar to the population as a whole the bias will have no impact on the polls. However, if the over represented group are affected by a government decision, this is likely to impact on the poll results, creating a larger margin of error.

I'll give you an example, longy. What happens if the government decides to remove some single mothers from the pension. What happens if these mothers are more likely to be interviewed than the general public because they are more likely to be at home. Under these circumstances, the voting intentions of the over represented group is likely to move further away from intentions of the whole population, creating a bias that favours the opposition.

longweekend58 wrote on Mar 2^nd, 2013 at 4:01pm:


I wasn't posting here during that period, so you can hardly comment on my opinions about polling biases at that point in time. Bob's opinion at this time, would be interesting, but is nothing more than a red herring.

longweekend58 wrote on Mar 2^nd, 2013 at 4:07pm:


Like most of us, Bob is not immune from making stupid and wrong comments. You of all people, longy, should empathise with that. However, he has also been guilty of sharing some quite insightful analysis. He is very much a mixed bag, which is why his ideas should be weighed carefully but not dismissed without thought. Much the same can be said about Alan Jones. I vehemently disagree with much of what he says which is often wrong, flawed in logic or shaped by a very different set of values. However, some of his arguments are impossible to dismiss. I don't assume Jones is wrong because his logic was faulty in the past.

Hi, melielongtime, 'smee again.  Tell us again how Abbott beat Rudd.  Your next attempt might be better....And while you are at it:

You said:

Quote:
....  make sense.  Okay, where in what follows does he not make sense?



Quote:

This uncorrected omission of numbers comes under criminal fraud I would think, and O’Shannessy should be arrested; in my view.

I ask the Senate to investigate Newspoll and ask it to justify its methodology.

A million is a tidy few. On top of which are, oh, fifteen million mobile phones not being rung.

And those mobiles aren’t rung, not one of them. Which means three million voters, probably, aren’t contacted.

the whole of the sample that is, belong to and are used by preponderantly older people; who mostly, not always, vote Liberal.

t is tempting for these reasons to call this poll entirely worthless, or the equivalent of mass forgery,

It has not only ‘been weighted to reflect the population distribution’ (translation: we made up the figures)

So, if you un-tweak the figures, throw in a margin of error, and some Katter votes trickling down to Labor in preferences, and some ten million mobile phones unrung, Labor wins this poll or scores 49.5, and Gillard leads Abbott 39-37.

It’s a pack of lies, probably.

Wikipedia http://en.wikipedia.org/wiki/Opinion_poll#Coverage_bias

Another source of error is the use of samples that are not representative of the population as a consequence of the methodology used, as was the experience of the Literary Digest in 1936. For example, telephone sampling has a built-in error because in many times and places, those with telephones have generally been richer than those without.

In some places many people have only mobile telephones. Because pollsters cannot call mobile phones (it is unlawful in the United States to make unsolicited calls to phones where the phone's owner may be charged simply for taking a call), these individuals are typically excluded from polling samples. There is concern that, if the subset of the population without cell phones differs markedly from the rest of the population, these differences can skew the results of the poll. Polling organizations have developed many weighting techniques to help overcome these deficiencies, with varying degrees of success. Studies of mobile phone users by the Pew Research Center in the US, in 2007, concluded that "cell-only respondents are different from landline respondents in important ways, (but) they were neither numerous enough nor different enough on the questions we examined to produce a significant change in overall general population survey estimates when included with the landline samples and weighted according to US Census parameters on basic demographic characteristics."

This issue was first identified in 2004, but came to prominence only during the 2008 US presidential election. In previous elections, the proportion of the general population using cell phones was small, but as this proportion has increased, there is concern that polling only landlines is no longer representative of the general population. In 2003, only 2.9% of households were wireless (cellphones only), compared to 12.8% in 2006.[15] This results in "coverage error". Many polling organisations select their sample by dialling random telephone numbers; however, in 2008, there was a clear tendency for polls which included mobile phones in their samples to show a much larger lead for Obama, than polls that did not.

The potential sources of bias are
Some households use cellphones only and have no landline. This tends to include minorities and younger voters; and occurs more frequently in metropolitan areas. Men are more likely to be cellphone-only compared to women.

Some people may not be contactable by landline from Monday to Friday and may be contactable only by cellphone.

Some people use their landlines only to access the Internet, and answer calls only to their cellphones.

Some polling companies have attempted to get around that problem by including a "cellphone supplement". There are a number of problems with including cellphones in a telephone poll:
It is difficult to get co-operation from cellphone users, because in many parts of the US, users are charged for both outgoing and incoming calls. That means that pollsters have had to offer financial compensation to gain co-operation.

US federal law prohibits the use of automated dialling devices to call cellphones (Telephone Consumer Protection Act of 1991). Numbers therefore have to be dialled by hand, which is more time-consuming and expensive for pollsters.

An oft-quoted example of opinion polls succumbing to errors occurred during the UK General Election of 1992. Despite the polling organizations using different methodologies, virtually all the polls taken before the vote, and to a lesser extent, exit polls taken on voting day, showed a lead for the opposition Labour party, but the actual vote gave a clear victory to the ruling Conservative party.