In response to the latest Roy Morgan poll – see National 54% in May Roy Morgan poll – lprent looks like he has rushed into an analysis a bit carelessly.
[lprent: Don’t get your panties in a twist. It isn’t that interesting, Just looks like the usual outliers that RM’s small sample size throws out. This poll is unlikely… It is too big a jump and the explanations for it are pure trash.
Caution is warranted for a single poll with a big shift in support.
But Roy Morgan state “both landline and mobile telephone, with a NZ wide cross-section of 866 electors from May 4-17, 2015″ – that’s not a notably small sample size.
That was a 8.5% shift to National pulling support from every other party. It doesn’t seem likely.
Allowing for a margin of error of about 3.5% this could be an unlikely high but National got 52% in January, then 49%, 46.5%, 45.5% and now 54% so fluctuations over a wide range are normal.
The usual shit analysis at Roy Morgan as well. Talking about the budget last week when their polling period was May 4-17, 2015 and the budget wasn’t released until the 21st. Even the damn spin wouldn’t have been there for most of the polling period.
The headline states “Positive news in lead up to New Zealand Budget helps National”, and in his analysis Gary Morgan, Executive Chairman, Roy Morgan Research, says: “National has enjoyed a strong increase in support in the lead up to last week’s New Zealand Budge”.
They are clearly referring to the lead up to the budget, which happened through the polling period.So the shit analysis is from lprent.
If it carries on through the next couple of polls I will get interested. But I’d expect a big adjustment down in the next poll.
Why expect a big adjustment down? Anything could happen in June – including a negative response to Labour’s poor budget response.
It is a pity that Roy Morgan are only releasing polls every month rather than every two weeks now. It means that their small sample size has 4 weeks until it corrects.
They have been polling monthly since the election last September, far more often than anyone else. So they provide the most useful data for seeing trends and fluctuations.
Again “their small sample size” is inaccurate.
Still good for a spinner to get excited about. I’d expect to hear you sprout your usual shallow analysis over the next weeks eh?]