Saturday, November 23, 2013

4th "CLISCI" survey by Dennis Bray - response frequencies published

Fourth "CLISCI" survey among climate scientists by Dennis Bray (with minor help by Hans von Storch) made public at

Bray, D., and H. von  Storch, 2013: A survey of the perceptions of climate scientists 2013

References to results in earlier results:

3rd survey:

Bray, D., and H. von Storch, 2010: CliSci2008: A Survey of the Perspectives of Climate Scientists Concerning Climate Science and Climate Change, GKSS Report 2010/9, 121 pp.

2nd survey : 
Bray, D. and H. von Storch, 2007: Climate Scientists’ Perceptions of Climate Change Science. GKSS-Report 11/2007

1st survey:
Bray, D. and H. von Storch, 1999: Climate Science. An empirical example of postnormal science. Bull. Amer. Met. Soc. 80: 439-456


MikeR said...

Interesting survey, though I'm sorry that the response rate was so low. Of course, it's a very long survey.

I am astonished at the very positive opinions by most of respondents about the ability of the climate models to predict the future. They are quite confident that the models can hindcast the last fifty years; I would have agreed except for the last ten years or so (which was actually forecasting), where the models really did very badly indeed. What in the world would make so many climate scientists still confident about the next ten years?

MikeR said...

Ah - forgot to mention that I was also astonished at the respect respondents have for regional climate models, which as far as I know currently have no skill for the past _or_ the future. Is my information wrong, or do these climate scientists just not know what they're talking about?

Karl Kuhn said...

I guess that a lot of strategic answering is to be expected in such surveys. You will think twice before bringing the reputation of your discipline into trouble

To the author: How is the problem of responding strategically dealt with in the survey?

Dennis Bray said...

@Karl Kuhn

There is no means to control 'strategic answering' as far as I am aware. There are however some tactics that are sometimes employed - i.e. asking the same question twice with slightly different wording, reversing the order or the value labels (i.e. one question good - bad, subsequent question bad - good. This survey was very long as it is, without the addition of further questions.

However, I do not think it is necessary to find a conspiracy plot behind every corner. If a person is so inclined, I suppose one way of satisfying the suspicion might be to adjust response values by 10% or whatever is a satisfying rate. The same would apply to face to face interviews. Short of using a waterboard, I think it something we have to live with.