People refer to “published” polls, so why are some unpublished?

Posted: August 5, 2013 in Politics, Uncategorized

These are the results from Roy Morgan

12-14, 2013
19-21, 2013
26-28, 2013
2-4, 2013
% Change
this week
% % % % %
ALP 42 41.5 38.5 38 -0.5
L-NP 41 (3) 41 (2.5) 41.5 (2.5) 43 (2) +1.5
Greens 7 9 10.5 9.5 -1
Ind. /Other 10 8.5 9.5 9.5
TOTAL 100 100 100 100  
ALP 52.5 52.5 52 50 -2
L-NP 47.5 47.5 48 50 +2
TOTAL 100 100 100 100  
Sample size 3,431 3,572 3,575 3,326   

This is from Galaxy

Two party preferred

Election Aug 2010 11-13 June 2013 27-28 June 2013 23-25 July 2013
Labor Party 50.1% 45% 49% 50%
Coalition 49.9% 55% 51% 50% 

As you can see, Roy Morgan has a  sample size of over 3000 which is much bigger than Newspoll. And, generally speaking, the bigger the sample the more likely it is to be accurate, unless there’s a lack of randomness in the sample.

Again, I make the point that opinion polls have 3% margin of error, so an individual poll is almost worthless in determining a trend – it may simply be the outlier- but the media seem to react to polls as though they’re the first figures in on election night. They often mention this, but then proceed to analyse the poll as though it’s 100% accurate. Given the closeness of the figures in all polls, it seems likely that a poll must at some point put Labor in front. Will we see this treated in the same way? Or will this be treated as an anomaly with people explaining it away by saying it’s just one poll and that Abbott has consistently led since 2010 so we can’t take any notice of ONE poll?

Well, there is one poll that we can take notice of. And then it won’t matter if Newspoll, Morgan or Galaxy had him leading by 55-45. And that poll is decided SEAT by SEAT. Abbott’s announcement that he won’t negotiate with minor parties could come back to haunt him if Katter’s party, for example, picks up a handful of seats in Queensland, or a couple of Independents are elected.

It’s a difficult task for Labor, but far from impossible.

  1. Stewart says:

    “it seems likely that a poll must at some point put Labor in front” What is your logic behind that?

  2. Stewart says:

    The most important thing at the moment is the western sydney numbers. That will make or break the labor party in this election.

  3. “it seems likely that a poll must at some point put Labor in front”.
    Just because an average of all the polls has it very close, and polls can swing by up to 3% without anything changing. If the “true” figure is 48-52, the Coalition’s way then the margin of error could have Labor in front in any particular poll. I suspect though, that the actual figure is tighter than that.

  4. The other fact which no-one talks about – because it makes speculation even more ludicrous – is that there are a significant number of uncommitted voters out there. (Galaxy put the figure at about 8%)

  5. Fed up says:

    How big are those undecided. or one could say fence sitters at this time. Has been up to 30% over the last year or so.

    While one has so many, no poll can tell us anything.

  6. […] People refer to “published” polls, so why are some unpublished?. […]

  7. Ben Russell says:

    If Abbott and the LNP get elected, there will be some international embarrassments for Australia in the next few years. We’ll most likely be behind the rest of the developed nations when it comes to social issues, and our public image will be represented by a man whose vocabulary is restricted to ‘um’ ‘ah’ and ‘stop the boats.’

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s