Tag Archives: Social science

Circling the Square

Circling the square

I’ve been attending a two-day meeting at Nottingham, Circling the Square, organised by Reiner Grundmann and colleagues from the Institute for Science and Society at Nottingham. This follows a successful similar meeting last year. The conference has its own blog, where video, audio and text from the meeting are gradually being uploaded.

The themes of the meeting were science and its links with policy, with the public and with the media. Many aspects of science were discussed, including for example drugs policy, where ex-Government Advisor David Nutt gave a keynote talk that contained some interesting points but rather over-stated his case with some dubious claims. But I will concentrate here on the climate-related aspects of the meeting. Whereas some social science conferences on these topics are little more than echo chambers for activists, this is certainly not the case with this meeting, which included a broad range of views.

Mike Hulme – “Scientists speaking with one voice: panacea or pathology?”

Mike Hulme (UEA) is the author of (amongst many other publications) an interesting book Why we disagree about climate change.

On Tuesday morning, he gave an hour-long keynote talk on “Scientists speaking with one voice: panacea or pathology?”, discussing issues around the consensus debate on climate change. He asked whether consensus messaging is normal or useful in science, and whether figures such as 97% actually mean anything. He noted that the ‘high degree of consensus’ was used in the first IPCC assessment (1990), arising from group deliberations, and that more recently this had been picked up by what he called the ‘consensus entrepreneurs’ and ‘consensus defenders’.

He discussed possible motivations for the consensus, including shoring up authority, providing a foundation for policy, and closing down dissent, before describing the so-called linear model, or “the plan” (Sarewitz), or “gateway belief”, which states that establishing a scientific consensus with the public will lead to public support and policy action. He mentioned a recent paper by Van der Linden et al that claimed to find evidence to support this. But bizarrely, their own data in fact does not support the hypothesis at all – their measure of support for action only rose marginally from 75.2 to 76.9, as noted by Dan Kahan, who had previously pointed out that a decade or so of consensus messaging has had no impact.

He also quoted from a paper “Should we aim for consensus?” by Beatty and Moore (2010): “Requirement of unanimity is pernicious” and “the authority of a scientific body is not undermined by questioning, but depends on it”. A key point he made was that consensus is insufficient – far more important than the numerical value of any consensus is the character and integrity of the scientists.

Here is his conclusions slide (HT Mike Schafer)

The ensuing Q&A session discussed whether consensus messengers were mainly from other fields, whether climate scientists had overstated the case, and whether consensus messaging could fail because of the romantic appeal of the underdog.

“Science advice” panel

The remaining sessions were in the form of ‘panels’ where a group of three or four people give very short introductory talks around a theme, followed by a substantial discussion session involving audience and panel members.

The first panel session was on “Science advice” and opened with some words of wisdom from Miles Parker, who had worked at DEFRA. He said that scientific advisors needed to ‘understand their interlocutors’, something that scientists were not really trained to do. ‘If you want to advise, first listen’. Perhaps David Nutt could have benefited from this.
Also on this panel was climate scientist Hans von Storch. He emphasized that scientists have their own values and preferences, and tend to think of their own fields as being particularly significant, and that scientists should be more aware of their own cultural background and ways of thinking. His text has been posted on the conference website.

“Science, uncertainty and science advocacy” panel

On Monday afternoon, Judith Curry was on a panel on the topic of “Science, uncertainty and science advocacy”. Her introductory statement can be seen at her blog. She defined advocacy as “forceful persuasion”, making an analogy with lawyers. She said she was concerned that too many scientists, and too many professional societies, were indulging in advocacy. She also complained that she herself had been (falsely in her view) accused of advocacy. She felt that many climate scientists have done a poor job of assessing uncertainty, and understated it. She pointed out that there is really no code of conduct for scientists communicating with the public, and this had resulted in irresponsible behaviour by some.

On the same panel, Conrad Brunk made some similar points: science is increasingly being used for advocacy, and there is a tendency to downplay uncertainty. Scientists talk to the public in a different way from how they talk to each other, and this can make them vulnerable to criticism.

Daniel Sarewitz made some interesting points about uncertainty. He said that uncertainty is easy to define if you have a large sample of data points, but this is not the case for ‘wicked problems’ such as GM and climate. For example, he said that statements such as “there’s a 10% chance of 6C of warming this century” really have no meaning. He suggested that in such cases we should not use the word “uncertainty”, but should talk about “disagreement” instead.

“Science and the media” panel

This was followed by a panel session on “Science and the media” including Leo Hickman, formerly of the Guardian, now at Carbon Brief, who discussed the role of the advocating journalist, saying that this was valid as long as it was transparent. He also addressed the issue of poorly worded press releases (this was also discussed last year, particularly by David Colquhoun). He proposed that it was important for the scientists themselves to take control over press releases and sign them off personally, rather than relying on University press offices.

Mike Schäfer (Zurich) also spoke about the problems with press releases – according to one investigation, 40% of university press releases contain exaggeration. He also raised concern about possible increasing segmentation, with people selecting their own preferred media source now that so many different sources are available.

“Science and the civil society” panel

On Tuesday after Mike Hulme’s talk there was a session on “Science and civil society” that included Ben Pile. He described a ‘bubble’ of interactions between governments, academia, NGOs and business, from which he said the public was excluded. He criticised NGOs, saying that they were sometimes incorrectly seen as representing the public, and that this amounted to a ‘democratic deficit’. He showed data indicating decreasing hunger and mortality, together with increasing food production and GDP in developing countries such as Bangladesh, and said that statistics showed that climate was well down on the list of risk factors for developing countries – thus challenging the “climate change predominantly affects the poor” narrative.

In the ensuing discussion, Peter Webster mentioned the substantial drop in deaths from storms, which can be attributed to greatly improved weather forecasts. Ben’s controversial remarks about NGOs went unchallenged until, near the end of the discussion, panel member Maurice Frankel (famous for his role in the FoI campaign) said that he didn’t see a problem with NGOs, as they countered the view of the industry lobby.

It was interesting and encouraging to see Ben Pile and co-panellist Stevianna de Saille chatting amicably afterwards, despite being at opposite ends of the spectrum in terms of their views on environmental activism and green NGOs.

Thanks to Reiner Grundmann and colleagues for organising a very interesting meeting (and apologies to all those panellists I haven’t mentioned).

Paper on climate scepticism published

A little over a year ago I noticed a call from the journal Environmental Communication for a special issue of articles on “Climate change communication and the internet”.  For some time I had been thinking vaguely about writing something about the interesting Reader Background thread at Jeff Condon’s Air Vent blog, so I wrote a paper on this and sent it to the journal.

The paper was handled very well by the journal.  It was reviewed ‘double-blind’, i.e. both the author and reviewers were anonymous. This isn’t usually done in my field but apparently it’s quite common in social science.  The reviewers were clearly experts in the field; they read the paper very carefully and made a lot of comments, criticisms and constructive suggestions for improvement. I revised the paper and then there was another round of quite detailed comments, but after that the paper was accepted.

The published paper is here, paywalled because I did not pay Taylor & Francis the Open Access fee.  However, apparently I can give 50 people access to the paper – let me know if you’d like this (in the olden days, you used to get a bundle of 50 paper reprints to mail out).  Alternatively there is a draft preprint version of the paper, before the review comments, with a few very minor changes. It contains a bit more discussion and opinion, fewer up-to-date references, and more mistakes, than the final version.

Below is a brief summary of the main points of the paper.

Why are people sceptical about climate change?

Surveys show that a significant minority of the population are sceptical about climate change and that there has been a modest increase in scepticism over the last few years. Some research has considered why this may be, but surveys do not usually ask people exactly what it is that makes them sceptical. There is some useful information on this question in comments on climate sceptic blogs, but this has not been studied in the literature. However, it should be kept in mind that these views expressed in blog comments are probably not representative of the general public.

In April 2010, Jeff Condon launched a “Reader Background” post, proposing “a discussion of our various backgrounds and how we came to be interested in climate science”. I count that there are 154 people on that thread who express some degree of scepticism about climate change. 17 of these explicitly call themselves lukewarmers, and at least 8 others express lukewarm opinions, so about 1/6 can be regarded as lukewarmers – though this is probably an underestimate. At the other end of the spectrum, about 1 in 10 are what might be called “hardcore sceptics”, using language like “scam” in relation to climate change.  Over 1/4 of the responders say they have a PhD, and a further 1/3 have some form of degree.  More than 1/4 say they switched from being concerned about climate change to being sceptical.

Reasons given for scepticism include

  • Hype and alarmism, either in the media or from climate scientists. 32 people give this as a reason.
  • Memories of previous scares, such as the 1970s ice age scare, mentioned by 15.
  • Politics – some say  the climate story seems politically motivated, others say  it does not fit with their own views, which often lean more towards libertarianism than conservatism.
  • Climategate is mentioned by 30 people, but only seems to have been a major influencing factor for 13. The survey was only 5 months after climategate, so most had probably already formed their view.
  • Poor science is mentioned by about 60, with the hockeystick most common issue.
  • Blogs – Climate Audit is most cited (57 times), followed by RealClimate (42 times) with many negative comments about their attitude and apparent failure to answer questions satisfactorily.
  • Other minor factors include books, newspapers and films.

Two other blogs have carried out similar exercises, see Judith Curry’s Denizens thread (Nov 2010) and My personal path to Catastrophic AGW skepticism at WUWT (July 2013). These show a similar picture – though there is some overlap in the contributors.

Another dumb climate psychology paper

A paper Public division about climate change rooted in conflicting socio-political identities was published yesterday in Nature Climate Change. The six authors, Bliuc et al, come from Social Science and Psychology departments in Australia. There is also an associated News & Views article by Tom Postmes, Psychology: Climate change and group dynamics.

The paper is truly awful, in so many ways, and effectively illustrates the blinkered thinking that is endemic in the field, nicely summarised by Andy West in his recent post at Climate etc. The following paragraphs raise some of the problems with the paper.

It starts badly, with unquestioning assertion of the Cook et al 97% consensus paper. As usual this is stated in an unclear way, “97% agree that climate change is caused by humans” (what, some of it? Most of it? All of it?), a feature that Ben Pile refers to as consensus without an object.

Then the paper simply declares that
“The public is divided between climate change ‘believers’ (whose views align with those of the scientific community) and ‘sceptics’ (whose views are in disagreement with those of the scientific community)”
It’s as simple as that. It’s us v them. The goodies and the baddies. Although there is a brief mention later on of multiple shades of opinion, this is dismissed with the statement “we argue that there is value in seeing climate change believers and sceptics as conflicting opinion-based groups.” Yet the paper and the News & Views piece talk about overcoming divisions, and conflict reduction – reducing a division that they themselves have artificially created. This is the same criticism I raised regarding Homer-Dixon et al (I have just noticed that the journal has now published my comment on that paper).
This isn’t just stupid – it’s really damaging.

The main “finding” of the paper is, as suggested in the title, that it’s all to do with social group identity. They say “we argue that people come to see climate change beliefs and scepticism not just as an opinion on an issue, but as an aspect of self that defines who they are, what they stand for, and who they stand with (and against)” and “the results support the contention that cultural polarization and political mobilization are at the core of the climate change divide.” This is the main idea of Dan Kahan’s work, even down to the exact wording used, “who they are”. But, amazingly, none of Kahan’s papers are cited (again, this was one of my criticisms of Homer-Dixon et al). As previously mentioned, I think there’s an element of truth in this argument, but it’s usually overstated.

As usual with such papers, although the text is nonsense, there are some interesting points in the data obtained. One of the issues they asked ‘sceptics’ and ‘believers’ about is their “anger at the opposing group”. The numbers came out (on some scale) at 2.84 for sceptics, but 4.10 for believers. We’ve seen recently several examples of the obscene vitriol directed towards even those who are moderately sceptical about climate change.
But the way this data is reported by Bliuc et al is astounding: “We note, in particular, that part of the sceptic group consciousness is anger at climate change believers”. The smaller number (sceptic anger at believers) is highlighted, while the considerably larger number (believer anger towards sceptics) is ignored.

Another common theme in these papers is that the main aim is how to change the minds of the sceptics: “strategies for building support for mitigation policies should go beyond attempts to improve the public’s understanding of science, to include approaches that transform intergroup relations”. This seems particularly devious: “efforts to undermine group efficacy, for example, by convincing sceptics that their actions are unlikely to prevent action on climate change, represent a more plausible path.” They are recommending that people should try to undermine sceptical groups – presumably they would condone the hounding of Lennart Bengtsson when he dared to join the GWPF.

The paper has a fair amount of jargon-speak. I liked this sentence: “Given that there are different causal orders proposed by existing models we conceptualize the antecedents of action as an integrated cluster of variables that represent a distinct group consciousness for believers and for sceptics, each of which predicts commitment to action to support the cause they each support.” I have no idea what it means – perhaps “we drew some pictures”.

In summary, we have a biased paper promoting political activism, exacerbating division and with a main conclusion that has already been stated many times in the literature. How did this rubbish get published? Oh, it’s in Nature.

There are articles about the paper at The Register, at The Conversation and at WUWT.

Lewandowsky’s Loopy Logic

As I’ve mentioned before, I really try to ignore the Lewandowsky nonsense, but  occasionally  an opportunity comes up that’s too good to miss. In The Conversation this week there’s an article “Are you a poor logician? Logically, you might never know” by Stephan Lewandowsky and Richard Pancost. Yes, Lewandowsky, the Chair of Cognitive Psychology at University of Bristol, who writes papers on his pre-determined conclusions without apparently noticing that his data doesn’t support them, has written an article about people who aren’t very good at thinking logically, and who over-rate their own competence; the article goes on to stress the importance of “introducing accurate scientific knowledge into public debates”. Honestly, I’m not making this up.

So I posted the following comment at The Conversation (and knowing the tendency of The Conversation to delete comments that don’t support their agenda, promptly took a screen shot).

Lewandowsky’s logical blunder has been reported on numerous occasions, by Steve McIntyre here, here and here, by ManicBeanCounter, by Jose Duarte and by Brandon Shollenberger, who showed how Lewandowskyan logic can be used to show that people who are concerned about climate change are pedophiles. Here is Brandon’s plot of Lewandowsky’s data, which nicely illustrates the error (small random numbers have been added to the responses so that they show up individually):

This is the data that Lewandowsky used to justify his notorious paper “NASA faked the moon landing—therefore, (climate) science is a hoax”. Despite the obvious errors, and calls for retraction, the journal Psychological Science has so far stubbornly refused to take any action.

After the section on logical thinking and the Dunning-Kruger effect, of which Lewandowsky himself is such a fine example, the remainder of the Lewandowsky-Pancost article sinks further. There is an unfair personal attack on Anthony Watts, saying that he thinks hot buildings contribute to warming, when in fact, as the authors are well aware, his concern has been with the poor siting of weather stations. Then there is the false analogy of smoking causing lung cancer – at the risk of stating the obvious, tens of thousands of people die every year from lung cancer, almost all of them smokers, so the link is perfectly clear, unlike the claims about future warming based on speculative computer models that are increasingly failing to match reality.

My comment was deleted by the Conversation’s moderators, and I received an email giving their guidelines – none of which were broken by comment. It can’t really be seen as off-topic, since it’s about logical thinking, and there are other comments that are much more off-topic that survived. Presumably the excuse would be that pointing out Lewandowsky’s errors counts as a “personal attack” – though he himself is allowed to attack Anthony Watts. (My comment survives at this blog where the article is copied). Several other comments were deleted and the comment thread was rapidly closed.

A future post will deal more specifically with the so-called Conversation, its censorship of comments, its abuse of public funding, and its bogus claims of “Academic rigour” and being “free of political bias”.


Update 10 Nov:

Some more links:

Ben Pile has an article on Lewandowsky’s Logic and what he aptly terms the “Nonversation”. He also notes the failure of the academic community to act as a check, and how this then reflects badly on the entire field.

Ben links to a series of three recent posts at by Andy West at WUWT. One of Andy’s main points is that Lewandowsky, and the climate movement generally is a prime example of the cognitive bias failings that he accuses others of – similar to the point I am making here.

Brandon Shollenberger reminds me that he wrote a longer document on the statistical error.

Finally, the LiveFromGolgafrincham blog has been fortunate to procure a special guest post from the man himself, in which he explains that the latest Conversation article was in fact intended to be a humorous parody, which makes a lot of sense.


Update 13 Nov: Latest Lewpy Logic

Another own goal by Lewandowsky appeared yesterday. A short, content-free opinion piece by him was published by IOP (have they forgotten what the P stands for?). It is the latest instalment in his “every-one-who-disagrees-with-me-is-a-conspiracy-theorist” canon. His paper draws attention to climategate, by claiming that sceptic blogs show a “continued and growing fascination” with it.
Yet again, the man exposes his clueless lack of self-awareness.
Needless to say, Lewandowsky’s claim of “continued and growing” sceptic fascination is nonsense. The above link shows that Climate Audit has had one article on Climategate in 2014, none in 2013 , and one in 2012. The much more active WUWT blog has had 4 in 2014, 2 in 2013, 16 in 2012, and 22 in Nov-Dec alone in 2011, following Climategate 2. Maybe someone can do a more thorough investigation?
ATROSTO again, it is he, by publishing this inane article, who exhibits “continued and growing fascination”, and “continued conspiratory obsession”.

As Judith Curry wrote on twitter, “New paper by Lewandowsky once again projects his own conspiracy ideation onto skeptics”.

In her latest blog post, We are all confident idiots, she writes “Lew is so busy dissecting the ‘bias’ of climate change skeptics that he misses his own rather glaring biases.”

WUWT writes “More insane conspiracy theory from Dr. Stephan Lewandowsky of Bristol University”

Fooling themselves

Adam Corner’s Talking Climate blog provides another example of the activist wing within social sciences fooling themselves.

Victoria Wibeck believes that there is a “persistent and problematic paradox”. Despite the overwhelming evidence for climate change and the high levels of concern people express, we still are not taking action on climate change. We are concerned about climate change, but not engaged. Previously I called this The Norgaard Delusion, after Kari Norgaard’s puzzlement over the fact that people said, when asked, that they were concerned about climate change, but didn’t talk about it all the time and continued their normal lives. The blog post goes on to consider “How can cli­mate change be made to feel mean­ingful in people’s everyday lives and how can they be encour­aged towards col­lective, affirm­ative action?”

Wibeck’s post provides a good example of confirmation bias – she selects survey results that tell the story she wants to tell, and ignores things that don’t. She cites a Special Eurobarometer survey that was all about climate change, to support her belief that climate change is a high priority (circular reasoning). If she had looked at the Standard Eurobarometer 80 survey she might have seen this:

And it’s a similar story in the US.

She also quotes from the “Climate Change in the American Mind” survey, picking out the third bullet point but ignoring the first one that says “There has been an increase in the proportion of Americans who believe global warming is not happening (23%, up 7 percentage points since April 2013)”.

There is no paradox. Wibeck and others in the groupthink-circle are fooling themselves. Time for another Feynman quote:

The first principle is that you must not fool yourself–and you are the easiest person to fool. So you have to be very careful about that. After you’ve not fooled yourself, it’s easy not to fool other scientists.

“Conceptual Structure of Social Disputes” – worse than we thought

In the previous post I commented on a paper “The Conceptual Structure of Social Disputes”, by Thomas Homer-Dixon, Manjana Milkoreit, Steven J. Mock, Tobias Schröder and Paul Thagard. The paper claimed to present an understanding of climate skepticism (among other topics) but seemed instead to show the prejudiced and ill-informed view that is wide-spread among social scientists. One further misrepresentation and smear that I didn’t notice yesterday is the claim of “absence of concerns about environmental issues”, dropped into the paper with no evidence.

It turns out that things are worse than we thought. Yesterday I said that I had not had time to look at the Milkoreit thesis which apparently forms the basis for the statements in the paper and the CAM diagram. Here is what the paper says:

“The maps are derived from Milkoreit’s extensive research on attitudes toward climate change. Using a variety of primary text sources, including newspaper articles, blogs, and transcripts of speeches of presidential candidates, interview data collected in 2012, and secondary literature on the role of ideology, media, and business actors in climate politics, Milkoreit selected concepts and conceptual links that various authors or interview participants had used or referred to most frequently.”

Having looked at Milkoreit’s thesis, this statement seems to be untrue. CAMs are discussed in chapter 3, p77-158, and that chapter contains no mentions of newspapers or blogs. In fact newspapers and blogs are only mentioned once in the entire thesis, on p 224 in chapter 5 on the “Q” method (with no indication of which newspapers or blogs) and these were only used to formulate a questionnaire. Responders to this questionnaire were divided into 6 factor groups A-F, but none of these groups express skeptical views – they are all minor variations within the climate-concerned group. Appendix 5-2 gives the political views of the participants, which are quite illuminating:
28 Left/Liberal
1 Center
2 Right/Conservative.

In summary, the thesis of Milkoreit provides no evidence to support the claims made in the published paper or press release.

Update: The journal, Sage Open, allows reader comments on articles, so I have submitted a comment, here in pdf form.

What’s going on in the minds of social scientists?

A provocatively titled press release from the University of Waterloo, What’s going on inside the minds of climate change skeptics? has led to some discussion on twitter, so I am setting up a blog post to allow more detailed discussion (and returning the compliment).

The article is based on a paper The Conceptual Structure of Social Disputes which looks at four disputed areas, one of which is climate change. The paper produces “cognitive-affective maps” (CAMs) that are diagrams linking together positive (green), negative (red), ambivalent (purple) and neutral (yellow) thought processes. Here is the CAM for climate skeptics:

The obvious question is where does all this come from? Is there any data or evidence behind it? Does it really tell us anything about the thinking of climate skeptics, or is it telling us more about the thinking of social scientists? Apparently it comes from the PhD thesis of Manjana Milkoreit, but this has 469 pages so I haven’t quite finished reading it yet. One flaw, pointed out by Ronan Connolly, is that the paper twice describes the liberal person in favour of climate action as ‘well-informed’, with the implication that skeptics are not. In fact, Dan Kahan’s work has shown (see last graph here) that skeptics are just as knowledgeable about climate science. The failure of the paper to cite Kahan is quite remarkable given the close relevance of his work. Another criticism made is that it presents the issue as two distinct groups, without acknowledging the reality of a continuous spectrum of views. The paper simply declares that to a climate skeptic, “anthropogenic climate change is not real”. This unhelpful attitude contributes to polarisation and antagonism. The authors then declare that “Figures 8 and 9 show that the conflict over climate-change policy is rooted in profound ideological differences”, when in fact these figures only show the prejudiced opinions of the authors. The chart presents the trendy view that skepticism is all about politics and policy; it is stated that blogs were one of the inputs used to determine this, but Amelia Sharman has shown that skeptical blogs are very much focussed on the science.

Another recent example of this sort of thing is poster 2 at the ecolabs blog, which was presented at a recent climate communication conference and supposedly plots the position and influence of various people in the climate debate. Reading the small print on the poster, the whole thing is “speculative and subjective”. The poster is quite informative, not of the state of the climate debate, but of the prejudiced groupthink that permeates much of the social science field. In the mind of Dr Boehner who produced the poster, climate contrarianism is all about the Koch brothers, Exxon, Fox and Murdoch. Sceptical bloggers such as Anthony Watts (who sometimes gets 1/4 million page views per day), Steve McIntyre, Bishop Hill and Jo Nova simply do not exist – Barry and I have asked her if she is aware of their existence. However, this poster of “prominent actors participating in climate communication” includes at least three who are no longer alive. This remarkable shoddy work reflects badly on Roger Pielke Jr’s group at the University of Colorado.

Of course, there is some good objective social science research being done in this field (Kahan, Sharman, Pearce, Grundmann…) but unfortunately the legacy of some awful earlier stuff lives on.