|
| |
March 9 2006
Further to yesterday's post (South Park Nonsense
II) re Nielsen 2004 polling data, Tim
Colebatch (The Age) replies to an email.
"Peter ... Your request for precise
figures on the numbers polled by age group forced me to hunt out my files of the
original detailed tables. And I discovered that the precise numbering was more
complicated than I had thought.
In the final three weeks, AC Nielsen polled 6251 people, of whom 5733 expressed
a two-party preference. The breakdown was
18-24, 702 polled, 639 expressed preference
25-39, 1778, 1639
40-54, 1893, 1729
55 +, 1878, 1726
So my 702 compares with their 127.
The Coalition 2pp was of course 52.7 per cent, which was precisely the average
of AC Nielsen's polls over that period (51, 54, 52, 54).
But that said, it is interesting to look at the demographic breakdown for the
first three of those four polls (Newspoll has never published the demographic
breakdown of its final poll). It totally contradicts The Australian's thesis
about the youth vote, and is broadly in line with what our poll found. On a
simple arithmetic average, the Newspolls showed the two party vote as
Coalition Labor
18-34 43 57
35-49 47 53
50 + 55 45
Which again, is what anyone with their head right way up would expect. And for
all Newspoll's limitations, it is backed by a much more solid sample size than
the AES, whose value is rather in what it tells us about the questions the
professional polls don't ask.
Tim"
A few points
 |
Tim also said the 2001-04 swing is from Nielsen's 2001 polling data,
rather than the result. That data overstated the Coalition lead by about one
percent, so swing-wise perhaps we could add about a percent to all age
groups below (so 18-24 didn't really move, while all the rest swung even
more to the government.)
|
But none of this alters the fact that Nielsen's is good quality data
that refutes (as Newspoll's no doubt does overall) the 'South Park Howard'
stuff, which is 90% sophistry. The problem is that this is how rumours
start, and before you know it otherwise sensible people are
quoting the Overington 'analysis'.
|
|
|