The Dunning-Kruger effect: deniers may “take down” what they don’t understand but at heart they are curious

Abstract: The Dunning-Kruger effect describes a cognitive bias, whereby individuals with limited skills – but with an overconfidence in their skill set – overestimate their abilities. I would contend the denial movement is rich with such individuals, who may be highly competent in other areas, but display a remarkable lack of understanding of science and the scientific method. I tested one such “denier community” for signs of Dunning-Kruger with positive results. While not an actual piece of research, I believe the responses I gathered tend to support the idea that many in the denial movement exhibit cognitive biases such as the Dunning-Kruger effect.However I also learnt that many of the “deniers” are articulate, curious about the world and actively engaged in trying to understand some of the science.

I have a confession to make: I am not qualified to discuss the intricate, technical details of climate science.

It’s beyond my capability.I can grasp the essentials, and even make sense of (some) the actual peer-reviewed research that I read. However I am very conscious that I have large gaps in my knowledge, and that crucially I am not qualified to critique the work of science.In order to have a real understanding I’d need to pursue a Bachelor of Science and post-graduate degrees to be able to speak authoritatively on climate science.

Like an avid footy fan, I understand many details of the game, but I’ve never played. However, what I’ve noticed about many deniers is their incredible confidence in being able to not only understand the science but critique it authoritatively.

Indeed, you’ll often find in both forum discussions and in person they can state with complete confidence that they have read the research, understand it and see “several crucial flaws” in it. Do they have a special insight that many of us lack? I suspect not, however there is a very good explanation for their apparent confidence.

Dunning-Kruger: how people overestimate their own abilities

The “Dunning-Kruger effect” is a form of cognitive bias which – put simply – states that unskilled people overestimate their own abilities. First published in the Journal of Personality and Social Psychology in December 1999, the author’s hypothesis is best summed up on Wikipedia:

• Incompetent individuals tend to overestimate their own level of skill.
• Incompetent individuals fail to recognize genuine skill in others.
• Incompetent individuals fail to recognize the extremity of their inadequacy.
• If they can be trained to substantially improve their own skill level, these individuals can recognize and acknowledge their own previous lack of skill.

The paper can be found here (the abstract is free, though you will have to pay for the article in full):

People tend to hold overly favorable views of their abilities in many social and intellectual domains. The authors suggest that this overestimation occurs, in part, because people who are unskilled in these domains suffer a dual burden: Not only do these people reach erroneous conclusions and make unfortunate choices, but their incompetence robs them of the metacognitive ability to realize it. Across 4 studies, the authors found that participants scoring in the bottom quartile on tests of humor, grammar, and logic grossly overestimated their test performance and ability.People tend to hold overly favorable views of their abilities in many social and intellectual domains. The authors suggest that this overestimation occurs, in part, because people who are unskilled in these domains suffer a dual burden: Not only do these people reach erroneous conclusions and make unfortunate choices, but their incompetence robs them of the metacognitive ability to realize it. Across 4 studies, the authors found that participants scoring in the bottom quartile on tests of humor, grammar, and logic grossly overestimated their test performance and ability

I believe that is sufficient to explain what Dunning-Kruger effect is. The question is, do we have examples of it from climate change deniers?

The answer is, we do.

My own experiment: Jo Nova’s blog

Jo Nova, the Perth based denier has quite active discussions on her blog site. They provide rich examples of conspiracy theories, vicious attacks on climate scientists and technical discussions on climate science.

I entered these boards to see how readily the denier community centered around this board potentially exhibited the Dunning-Kruger effect.

For my own variation of testing the Dunning-Kruger effect I engaged several posters in a discussion on the “evidence” for the relationship between the concentration of CO2 in the atmosphere and temperature. The science itself is well established, and can be easily found on the Internet.

I posted links to several peer-reviewed papers as “evidence”: in fact, this was demanded by Jo Nova and several posters.  The papers I provided this community with were sourced from AGW Observer, and relate specifically to “the correlation between carbon dioxide concentration and temperature.”

Very quickly, the forum was alive with comments. Indeed, many posters felt they could dismiss the research in a quick and easy manner:

Mike #21, I can’t look at all your articles but this one caught my eye:

A Bayesian Statistical Analysis of the Enhanced Greenhouse Effect – Tol & De Vos (1998)

In particular they provide a regression equation (2) and a graph Figure 1 that goes up to 1990 and would be interesting to apply the same equation to the years from 1990 to 2009 and see how badly it diverges.

Unfortunately, they don’t actually specify the coefficients of the FIR filter that they use for CO2 values (see Appendix B), I could take a reasonable guess at that one. They also attribute the Dust Veil Index to Lamb (from the CRU at East Anglia) and the Lamb data file is here:

ftp://cdiac.esd.ornl.gov/pub/ndp013/ndp013.dat

This is not in a suitable format (not yearly samples) and many values are missing, does not even go up to 1990. Mike Mann provides his version of the DVI here:

http://www.ncdc.noaa.gov/paleo/ei/ei_data/volcanic.dat

Which does go past 1990, but not as far as 2009. Maybe this was the file the Bayesian analysis actually used? Note that the Mann file contains values for earlier years that Lamb had no values for. Also I stumbled onto this:

http://bobtisdale.blogspot.com/2008/04/mann-et-al-weighted-dust-veil-index.html

So other people have already noted the strangeness of DVI values.

This is a typical problem with trying to reproduce even basic equations and plots in these papers — chasing up the many bits of data that are not tightly referenced and difficult to get hold of. Their conclusions were very strong probabilities of several degrees of temperature sensitivity, however their main cooling factors were DVI and ENSO and in recent 15 years I doubt that would account for the actual observed global “mean” temp, ENSO has not been spectacular in any direction and not a whole lot of volcanic dust in the air either (but there has been some).

Also, their regression between 1870 and 1990 has bigger problems if you look at the Greenland ice cores that show a brief warm period approx once every 1000 years (e.g. Roman warm period, Medieval warm period). How good is a regression that takes a little sample of time in a longer and far more complex time series? The regression can only use what it has, it can’t know there is a peak every 1000 years.

My gut feeling is that this article is flawed, but with up to date data series I could solidly prove it was flawed by observing whether their regression equation predicted the recent cooling. Maybe you can help find the data and we can nail this one good.

There’s allot to take in here, and the author of this post sounds authoritative. However, what is worth noting is how confident he is in proving how flawed a highly technical piece of research is.

I will pay that poster credit: he/she is trying to find other sources to confirm or deny the conclusions of the paper.

Most comments where more akin to this:

geronimo

mike at 9.15pm:You haven’t read the papers your referred to have you? None of them have empirical evidence of CO2 causing temperature increase. It’s too difficult to provide empirical data, which usually means it’s the wrong theory. We have plenty of ice core data showing no relationship between temperature and CO2 except for the CO2 rising in response to a temperature rise.

In this instance no sources or counter evidence is provided, just a confident dismissal of the science.

These people are not stupid, they’re curious

The first poster accepts the papers are talking about CO2 concentrations and temperature rises: they simply reject the papers conclusions. The second poster does not think contain any evidence, and easily dismisses them.

Do I regard these individuals as “stupid”, “incompetent” or “completely unskilled”? No, not at all! Actually, I did not expect them to be any of those things. The heart of the matter is this: too many people think they are qualified in areas they are not.

I came away with the conclusion that many members of this community are articulate, engaged with the debate and intellectually curious.However, like me they lack a full understanding of the science.

Untested assumptions

I will admit that I ignore several key variables in my little “experiment ” which I readily admit:

• I do not know the educational or professional background of the posters, however the likelihood that they lack expertise in climate science is very high
• It’s a very random sample, and in no way qualifies as actual research

In doing this have I also exhibited the Dunning-Kruger effect? Well. there’s a good question!

I will say that as a piece of research, what I’ve done is basically meaningless.Perhaps it was no more than a quasi-scientific excuse to act as a forum troll.

However, I was curious to see how this particular denier community would respond when exposed to genuine science. The results where as I expected: denial.

A long time ago I recognised the limitations of my own knowledge and understanding. I will defer to experts on highly technical subjects. What I do see in many deniers is an overconfidence in their own abilities and a complete dismissal of the abilities of climate scientists.

So, should we give up on deniers?

No we shouldn’t! I don’t believe deniers are “stupid” or “ignorant”. Many display a keen desire to understand the science and engage in discussion. Most are polite, and will happily chat to those “on the other side” of the debate.

Nor should we say to them  “just trust the scientists”. They don’t.

Conclusions

At heart many “deniers” claim to be curious individuals. I think there is some truth to that.

Helping them understand just how fiendishly complicated the science that supports climate change actually is may engender more respect for the work scientist do. I also think those in the denier community might enjoy the opportunity.

Perhaps we should be less concerned with bombarding the deniers with the results of research, but engaging them with how the science works. I actually think many of them would be fascinated.

Otherwise many of these individuals are left to the mercy of the peddlers of conspiracy theories and pseudo-science.

As Dunning-Kruger suggests:

…If they can be trained to substantially improve their own skill level, these individuals can recognize and acknowledge their own previous lack of skill

Perhaps scientists should be reaching out to the denier community and giving them an intimate insight into the scientific method and how they arrive at their conclusions.