Wednesday, February 17, 2016

Fifth Bray – von Storch International Survey of Climate Scientists completed




In the past, beginning in 1995, Dennis Bray and Hans von Storch, conducted a series of interviews with an international sample of climate scientists concerning their opinions of climate change, climate models and the social and political dimension of climate change. These interviews provided the basis for a set of questions to be used as a survey instrument of the broader climate science community.

Surveys were conducted in 1996, 2001, 2007 and 2013. An overview is here. The full results and data sets of the surveys are also available on academia.edu.
A fifth survey was conducted from December 2015 to end of January 2016.  This survey has now closed and we are preparing a full descriptive report of the data. 

The new survey, which addressed climate scientists (for details, wait for the forthcoming detailed report), covers some of the earlier questions used, but has also some blocks which are new, namely
  • ·       Climate Service Centers
  • ·       Definition of Extreme Events
  • ·       Attribution of Extreme Events
  • ·       Climate and Society
  • ·       Science and Society
The Table below shows a comparison of the corresponding responses to one common question for 2013 and 2016

How convinced are you that most of recent or near future climate change is, or will be, the result of anthropogenic causes? (% rounded off to 1 decimal place)

Likert Scale

Survey #4 2013 (%)
Survey #5 2016 (%)
1
Not at all
2.5
1.3
2

5.0
2.2
3

3.2
2.2
4

7.9
6.7
5

9.4
13.7
6

28.5
25.
7
Very much
43.0
48.9
Agreeing to human influence
Sum of responses   5+6+7
80,9
87,8
Agreement (Pew)*
Sum of responses
5+6+7 + 4/2
84.9
91.3

* Earlier, the Pew Research Center (http://www.pewresearch.org/fact-tank/2015/01/29/5-key-findings-science/) published results of a survey of US scientists “connected with the American Association for the Advancement of Science (AAAS)”. One question was if one would agree (yes/no) to the assertion “Climate change is mostly due to human activity”, which is similar to our question. This assertion received 87% agreement; also well below highly publizied claim of 97%. For a better comparison with this yes/no question, we calculated from our data the sum of all positive assertions (5+6+7 plus half of the “neutrals” – see last row in table), and arrived at 91%. We would claim that our result, 91% is consistent with Pew’s 87%. Note however, that Pew seem to have asked not only climate scientists, while our survey was done only with climate scientists; also the 97% claim refers to this sub-group.

Thus, the overall agreement to the attribution of climate change to mostly human drivers, is very high, 88%, and has increased by about seven points since the last survey. Even if this number is very high, this is markedly lower than the often-quoted 97% agreement among climate scientists.

8 comments:

  1. The response rate is: 658 respondents out of 3882 invitations, 17%. This number is higher than what we had earlier (7% in 2013).


    ReplyDelete
  2. ich finde, Hans von Storch bewertet diese Studie doch arg über.

    Eine meiner Meinung nach korrekte Meinung über die Umfrage: http://julesandjames.blogspot.de/2015/12/trivial-pursuits.html

    Dazu kommt der Vergleich mit 97%. Die 97% gelten für Klimatologen, die aktiv an dem Thema Klimawandel arbeiten. Für Klimatologen im Allgemeinen: ca. 88% (verrückt, oder?). Damit passen diese 88% zu der Studie von von Storch und Bray bzw. Pew ganz gut, oder? Hm?!? War in Doran/Zimmermann 2009 oder so, oder? Darauf wurde doch angespielt oder? Krass.

    Man kann es auch anders formulieren: die Einteilung der Umfragegruppe ist doch sehr weich und bestimmt nicht einfach zu definieren. Man kann da schon viel diskutieren. Ich kenne Arbeiten, die sich allein mit der Feststellung und Einteilung der Teilnehmer einer Umfrage befassen.

    Aber es passt schon halbwegs zusammen. Die meisten Studien kommen da so auf ca. 90+-paar Prozent.

    Warum auch nicht? Auch in der post-normalen Zeit gibt es noch Physik, der man sich eben nicht so leicht entziehen kann.

    naja...

    Gruß, ghost.

    ReplyDelete
  3. Let me ask a question about how you expect respondents would answer on the 7 point scale IF:

    A) Respondent's opinion is that it is equally likely that most of recent or near future climate change is NOT, or will NOT be, the result of anthropogenic causes.

    B) Respondent's opinion is that it is more likely that most of recent or near future climate change is NOT, or will NOT be, the result of anthropogenic causes.

    ReplyDelete
  4. Sorrry, #3, I do not understand your question? Is it a matter of the wording of the
    question?

    ReplyDelete
  5. @model2

    It is a little difficult to determine what you are asking. Looking at your question A, you seem to be asking how would a person respond if that person was convinced that recent change that has occurred was not related to anthropogenic causes and change that will occur in the near future will also not be related to anthropogenic causes. If that person believed that recent change that has occurred is not related to anthropogenic cause and change that will occur in the near future is not related to anthropogenic causes then I would assume he or she would respond that he or she is not at all convinced that either recent change or near future change are or will be the result of anthropogenic causes. The respondent would answer 1. not at all convinced that most of the recent or near future is, or will be, the result of anthropogenic causes.

    Question B seems to be a rephrasing of question A. Again, the respondent would answer 1. not at all convinced that most of the recent or near future is, or will be, the result of anthropogenic causes.

    ReplyDelete
  6. Not surprisingly Bray' poll shows an overwhelming believe in AGW-basics (or shall I write "consensus"?). The results are in very good agreement to Verheggen et al. Bart Verheggen gives in his blog post a reasonable explanation about the difference to the famous 97%, see https://ourchangingclimate.wordpress.com/2014/08/11/survey-confirms-scientific-consensus-on-human-caused-global-warming/

    Andreas

    ReplyDelete
  7. Hans von Storch, Dennis Bray,

    A) Respondent's opinion is that it is equally likely that most of recent or near future climate change is NOT, or will NOT be, the result of anthropogenic causes.

    The question above can be more re-stated more clearly.

    I am thinking of the opinion that it is equally likely (50/50 chance) that most (more than 51% ?) of recent warming is due to anthropogenic causes. That is, it is equally likely that the recent warming was due to natural causes as it was due to anthropogenic causes.

    How you you expect such a person holding the opinion above would answer the question posed on your survey? My first guesss was that such a person would answer with a 4 -- but the answer is not completely clear to me.

    ReplyDelete
  8. The question on survey the reads “How convinced are you that most of the recent or near future climate change is, or will be, the result of anthropogenic causes?”.
    Model2 above poses 2 questions:
    Question 1. ‘Respondent's opinion is that it is equally likely that most of recent or near future climate change is NOT, or will NOT be, the result of anthropogenic causes.’ – 50% convinced that climate is the results of anthropogenic casues and 50% not convinced that climate change is the result of anthropogenic causes. This asks the level of conviction a respondent has regarding such a claim. On a scale of 1 (not at all) to 7 (very much) 50/50 would fall at the response of 4.
    To ask about ”Respondent's opinion is that it is equally likely that most of recent or near future climate change is NOT, or will NOT be, the result of anthropogenic causes.” is simply the inverse of ”Respondent's opinion is that it is equally likely that most of recent or near future climate change is, or will be, the result of anthropogenic causes.”
    The 7 point scale used in the question of the survey is devised to capture both the positive and the negative.
    Question 2 asks “I am thinking of the opinion that it is equally likely (50/50 chance) that most (more than 51% ?) of recent warming is due to anthropogenic causes. That is, it is equally likely that the recent warming was due to natural causes as it was due to anthropogenic causes.”
    This asks ‘What is the level of contribution to climate change made by anthropogenic causes?” and is not the intention of the question. The question posed in the survey is concerned with how much a person is convinced that anthropogenic causes are responsible for climate change, not how much climate change is the result of anthropogenic causes.
    In the survey a response of 1 indicates that the respondent is not convinced at all, a response of 2, the respondent is somewhat convinced but still has a lot of uncertainty, and so on until a response of 7 indicates that the respondent is (almost) 100% convinced that most of the recent or near future climate change is, or will be, the result of anthropogenic causes.
    The question could have been posed as “Are you convinced that most of the recent or near future climate change is, or will be, the result of anthropogenic causes?” with response categories of 1= yes, 2=no, however this allows for no variations in strength of conviction. To deal with a question that asks ‘if most (more than 51% ?) of recent warming is due to anthropogenic causes’ the question could also have been posed as “How much of recent or near future climate change is, or will be, the result of anthropogenic causes?” with response categories being, for example 0%, 10%, 20% ….100%. At which point, someone might ask, what if the respondent thinks it is 15%? The question could also be posed as an open ended question in which the respondent is free to insert, for example, 27% but then it would be unclear as to the meaningful difference between 24% and 27%. What we attempt to do in these questions is capture general perceptions of the climate science community.

    ReplyDelete