the Air Vent

Because the world needs another opinion

Design a poll

Posted by Jeff Id on September 28, 2010

This might work,  let’s design a reasonable survey of AGW opinions.  People don’t like the VonStorch blogs work very much but if the design was done better, it could better represent the various views.  Steve Mosher would probably be a perfect guy to do the whole thing but what I’m looking for is a survey along the Die Klimazweibel style representing all positions with reasonable questions and non-leading answers.

Do you believe increasing CO2 will:

a – increase global temperatures

b – decrease global temperatures

c – no change

d – nobody knows

Are IPCC temperature projections generally

a – accurate

b – inaccurate

c – nobody knows

That sort of thing.

This concept might flop completely but I bet with enough participation in the design, it could turn out to be a reasonable survey.  It should cover the readers positions on IPCC positions on solutions, projection and models with the simplest possible form.

23 Responses to “Design a poll”

  1. Paul said

    Do you believe Increasing Atmospheric CO2 levels will has:

    a – measurably increased global temperatures

    b – measurably decreased global temperatures

    c – no change

    d – nobody knows

  2. Jeff Id said


    How about one for how you got interested in AGW.

  3. Kenneth Fritsch said

    I continue to say that the survey needs to be essay in format and probably ask only 2 or 3 main questions about uncertainty attached to warming and its effects, where the survey participant stands on the level of certainty they require to initiate mitigation and finally where the participant stands on benefits and unintended consequences of mitigation.

    I would use a random sample of participants with perhaps one multiple choice question about how the participant might classify herself as belonging to the consensus or a skeptic or a few levels in between.

    Such a survey would require a good deal of thought and analyses in summarizing and is probably why survey takers shy away from these types of surveys.

  4. Jeff Id said

    An essay form is more accurate but too long. I doubt people would spend the time.

    If I started a thread on it, I wonder how many would have interest?

    As of late, I have less interest in writing anything about opinions on global warming. My own views are hardening from experience and they have already been expressed enough. This idea seemed like a way to express more of others views.

  5. D.T. said

    How about defining a skeptic first?

    A skeptic:
    a. disbelieves everything.
    b. doubts or denies a basic principle.
    c. suspends judgement until a basic principle is demonstrated objectively.
    d. seeks consenus to determine a basic principle.

    Or something like that.

  6. Bruce said

    Jeff, you have to go for more detail on the first question. Many people believe CO2 raises temperature by a small amount. Indeed, the empirical 2XCO2 values from Christy, Choi and Spencer appear to come in consistently around the 0.5-0.6 C/doubling mark.

    This seems the problem when most MSM polls are conducted, since when asked ‘do you believe CO2 causes global warming’ I have to answer yes. But they never ask ‘how much’ or ‘is it dangerous’ or ‘how long until we reach IPCC’s arbitrary 2 C limit’.

    Using Spencer’s data to answer the latter question we get the answer ‘centuries to millenia’, which means we’ve some time to develop better answers (even if we don’t tip over into pre-iceage cooling in that long time period).

  7. TinyCO2 said

    I agree with Bruce. It’s hard to reduce CO2 warming down to a yes, no answer. Perhaps something more like this-

    Assuming no major additional event (Krakatoa style eruption) Do you think CO2 will warm/cool the planet over the next 100 years:-

    a) -0.6ºC or below
    b) 0ºC
    c) 0.6C
    d) 1.2ºC
    e) 2.4ºC
    f) 4.8ºC
    g) 6ºC or higher
    h) don’t know

    Are you confident about this answer? Y/N

    Can you keep this topic as a sticky or something similar as it requires a bit of thought?

  8. TinyCO2 said

    One of the things that I’d like to see is what knowledge the person had in the specific question asked (eg if the question was about sea level rise, was the person experienced in that field or were they an expert on rainfall). Something like this but less wordy and cumbersome-

    Are you confident about this answer?

    a) No
    b) Yes but as an amateur
    c) Yes as an interested amateur (includes anyone who knows the subject but isn’t familiar with the detail)
    d) Yes as a citizen scientist (anyone, including climate scientists in unrelated subjects, who has spent some time examining the fine detail but is not involved full time)
    e) Yes as a junior climate scientist in this specific field.
    f) Yes as a senior climate scientist in this specific field.

  9. Brian H said

    “CO2 is a magic global thermostat.”

    A. Agree
    B. Disagree

  10. Brian H said


  11. Mark F said

    Part A: How much will the planet heat or cool….?
    Part B: What will be the portion of that due to CO2?
    Part C: Based on what?


  12. Amabo said

    A question I would like to see.

    How will any future change in climate occur?

    a. Globally and uniformly. (The globe will experience climate change of a single type/effect.)
    b. Regionally and uniformly. (Several regions of the globe will experience the same type/effect of change in climate, while others remain unaffected.)
    c. Globally, and non-uniformly. (Ie. Difference between northern and southern hemisphere, or all regions of the globe will experience different changes.)
    d. Regionally and non-uniformly. (Ie. Several regions will experience changes of different types depending on specific region, while other regions will remain unaffected.)
    e. All of the above, at the same time, in the same place with monkeys rampaging through the streets.
    f. Other.

    And maybe add a small space for specific notes, with maybe an added question of how the changes and effects could be predicted.

  13. Kenneth Fritsch said

    A broad sampling across the spectrum from consensus to skeptic would allow one to analyze how much of continuum existed while at the same time allow one to separate the wackos from the thinkers on both sides of the spectrum.

    And, of course, doing the survey right would require a lot of effort in the analyses and in the taking of it. The more simplistic and easy to handle surveys are usually taken to prove some political or other pre-conceived point.

  14. TG O'Donnell said

    Von Storch’s attempt to explore the mindset and motivation of AGW sceptics is certainly interesting. Unfortunately many of his questions are difficult to understand,ambiguous or obscure.
    I believe that the primary reason for this (apart from obvious translation issues) is the lack of clearly defined objectives for the study. In market research terms, it is critical to understand the goals of a study and indeed how the results will be used before one designs a research methodology. Therefore, if you/Air Vent wants to delve into the attitudes and motivations of visitors, it would be best to define clearly what you want to understand about them and why.
    For instance, you say that you want a survey ‘representing all positions’, in which case use a combination of closed ended and scalar questions allowing respondents to position themselves in terms of the fundamentals.
    For instance; The first requirement might be to identify where respondents are on the basics e.g.
    Q1. Do you accept that global temperatures have increased since 1975? (Yes; No; To some extent)
    Q2. Do you believe that atmospheric CO2 is a green-house gas? (Yes; No; To some extent)

    Scalar questions can then explore respondent attitudes and degrees of warmism or scepticism on key topics e.g.
    Q3. To what extent do you believe that the global temperature rise observed over the last 30 years is due to manmade CO2 emissions? (Where 1 = ‘No relationship’ and 10 = ‘Human activity is primarily/exclusively responsible’). (Respondents who do not believe that temperature has increased at all should be routed through)
    Similarly, other issues such as the impact of increasing levels of CO2, the credibility of the IPCC, confidence in ‘climatology’ as a science, the need for massive intervention etc and any other matters of opinion and belief can all be explored using attitudinal scales.
    Where issues cannot easily be expressed as scales or asked as closed-ended questions, then open ended ones e.g. “Why did you first become sceptical about the AGW consensus?” provide much of the key information that would be included in Kenneth Fritch’s ‘essay’ format.
    Bearing in mind that blog visitors are almost all, by definition, interested in the subject, I would suggest inviting all readers to participate and would anticipate a high response level …. probably in excess of 50%.

  15. Kenneth Fritsch said

    Do you believe that atmospheric CO2 is a green-house gas? (Yes; No; To some extent)

    I would not think that there is a “to some extent” answer here. Either the physics apply or they do not. Maybe the question could be used to determine how the responder thinks about these issues. I suspect that question would need to define better and more specifically what green-house gas means in the survey question.

  16. Jeff Id said

    I think tonight, I’ll start incorporating the suggestions here into the head post. That way if there is enough interest a reasonable pole can be built.

  17. Gary said


    A format suggestion:

    If you are looking to access opinion, a Likert scale (1 to N, going from highly negative to highly positive) is more appropriate than multiple choice because subjective feelings (informed by fact, perhaps, but feelings nonetheless) tend to be continuous. Use multiple choice items when the responses are clearly discrete.

  18. Dr T G Watkins said

    Nice idea, but I agree with Bruce and others that a more nuanced approach is needed. I suspect that the people who follow your blog would find it difficult to answer yes/no black/ white questions.
    Only retired old codgers like me would have the time to write essays, although that could be quite amusing ( like an exam ) to see the misconceptions revealed.
    I mean my ignorance although I like to think I’m up to speed after several years following the stories my underlying maths and physics isn’t strong enough.

  19. Brian H said

    There is also the possibility, many including me suspect a probability, that GHG increases (particularly CO2) set off a negative feedback response that results in net cooling.

  20. Paul said

    What is the ideal temperature of the earth?

    a) Early Holocene
    b) MWP
    c) Preindustrial(little ice age)
    d) Current

  21. Paul said

    Which mitigation proposals would you support(select all that apply):

    - Large scale deployment of Nuclear Power
    - Large scale deployment of Hydro-electric
    - Large scale deployment of Clean Coal
    - Painting asphalt under weather stations white
    - Relocating the UN from NYC to zero-emission buildings in Gaza
    - Removing Universities from the electric grid and requiring Green Energy only

  22. Geoff Sherrington said

    TG O’Donnell notes correctly that there are pitfalls in design. It would not be so good to do a poll that attracted the criticism that Dr Wegman gave (not enough employment of specialist statisticians in climate science, in his case).
    Although nfar from exact, polling has learned from many past attempts and there is some formalism about which a new pollster should take advice to avoid the criticism “Not enough professional pollsters in tAV effort”.

    For example, a first weeding out question might be “Are you or close associates involved in carbon trading to the extent that you would profit from a general acceptance of global warming?” This might help curb a bulk mail-in by organised interest groups. But then, people are not compelled to be truthful.

    There is quite an art in design of questions so that the interpretation can go past the first level of response, to multivariate analysis and even detection of contradiction/faking. It is easier to get good results when you poll random people. Polls involving a known demographic with general properties (like readers of tAV)are quite hard to pose and the results are often merely confirmatory of prior impressions. Even when there are surprises, what do you do with them?

    How do I know? My boss once told me to set up a polling system to be worked by head office staff after hours one day each month or so. The analytical software was good, the system worked with volunteers and bar code wands, the questions were written by professionals. Typically over 1,000 responses per question. I would use it when visiting pollies, who would pull out old charts saying “Here’s what our polls show”. It was a winning move to pull out a bigger poll from the night before.

    IMHO, it’s much more valuable to work at tactics like those that caused the Royal Society to have a rethink today. Maybe this just means keeping up pressure on important questions and concepts that still lack credible answers, or working out and presenting more credible answers – which I guess is part of the scientific method.

  23. Brian H said

    Not much of a rethink. The boilerplate stating there was sufficient evidence (unspecified) of anthro-warming discredits the entire effort. The RS covered itself in ingloriousness once again. Newton must be barfing in his grave.

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


Get every new post delivered to your Inbox.

Join 133 other followers

%d bloggers like this: