The Climate Institute of Australia's new glossy brochure is based on an online survey of 1,145 Australians. These 1,145 participants were selected by invitation by email and/or by selecting the survey from a list provided (and tailored) on the user's home page (example below).

The 1,145 are selected from 180,000 panel members of MyView.com.au, a site where you earn points for answering surveys.

I decided to join MyView.com.au to see what the selection process is. To my surprise the surveys weren't completely without description. For example one survey involving air travel was labelled as such. E.g.:

As soon as I answered that I will not travel by air within the next year, the survey was terminated. I suppose I won't be sent any air travel surveys in the near future.

Here's another poll in my selection list stating description "in the home":

The site also explained that my answers would be used for tailoring what surveys would be sent to me. For example I said I wouldn't buy a new car in the next three years, or that I bought one in the last three. Therefore I guess I won't be getting those surveys any time soon.

And every question answered is irreversible and recorded against your record. I got this message after answering a few questions from the "in your home" survey -- it seems I don't buy the kind of product they were surveying for:

Are these filtering methods are legitimate? Supposedly the site is certified to international standard ISO 20252. Whether that means every survey meets that standard...I don't know, but it would seem not.

If I ever made the statement that I was not interested in climate, or was a sceptic, it may be held against me, and I may not get sent the climate-related poll invites.

So what's the bet that an invite to do a climate-related survey has the label climate, and that therefore only people who are interested in climate -- or even joined myview.com.au just to do climate surveys -- are the main people to respond and do the survey?

The 2014 poll results were very alarmist-oriented in my opinion. E.g. this alarmist headline from the Guardian:

Australians want renewable energy target retained by big margin

Polling shows 72% of Australians want to keep or expand RET, as Abbott government considers abolishing it

Really 72%? That seems a lot. It would be interesting to see the exact question asked in that case.

So, could this poll be merely a survey of people who know in advance what the topic is, bringing automatic bias with it? I can only tell you in a year or so when the next Climate Institute poll comes up and I do, or don't, get an invite, and how such an invite is phrased at the outset.

I haven't been part of the site for long enough to get one of these invites, but so far that's the impression I get. If true, that would seriously throw the idea that it is a random survey into doubt.

Imagine this analogous scenario. You get a telephone call like this: "hello, would you like to complete a survey on the climate?" "..Click!" comes the response, as many who aren't interested in climate hang up the phone.

Or even worse: "hello I'm from the Climate Institute of Australia, I'm wondering if you'd like to answer..." "..Click!"

Many warmists and perhaps a few keen sceptics would stay on the phone for the survey, but many wouldn't. Plus, some people won't answer survey's on any topic; how does that skew the results of these polls? Does the very makeup of the myview panel skew results?

I guess to keep it truly random you'd have to express it as something like a "science survey", perhaps keep it obscure for as long as possible.

===================

Then there's this curious statement on page two of the brochure:

"..an online panel. The data collected was then weighted according to the most recent Census data for location, gender and age."

Weighted? Why? For example with age, let's say there are more answers from young people. So does that mean answers from old people should receive more weight? Or does it mean they should receive less?

And weighted for what? I don't see any reason for weighting by age, location or gender. It's all very odd.