5 ms·

>> In fact the "95% of scientists agree" is just such an example - if you look into it, the number becomes very questionable. And once people see one of your ar

>> In fact the "95% of scientists agree" is just such an example - if you look into it, the number becomes very questionable. And once people see one of your arguments is shoddy, why should they give your other arguments the benefit of the doubt? > In fact the "95% of scientists agree" is just such an example - if you look into it, the number becomes very questionable. <------------- > A very poor example on many levels. I'd like to draw your attention to something. Notice how you've dropped a key idea that @adventskalender pointed out: "And once people see one of your arguments is shoddy, why should they give your other arguments the benefit of the doubt?" I think this type of behavior is interesting, common, and potentially consequential. > 1) This statement is fundamentally irrelevant to the (scientific) discussion, because it does not contain a scientific argument. And yet, it is commonly repeated on a regular basis in the "scientific" discussion, despite not being technically true. > It could be 0%, it could be a 100%, that wouldn't change the validity of any of the scientific arguments presented. It changes the rhetorical/persuasive value of it though. > Truth isn't measured in mass appeal. Perception of truth is. > the point that a consensus broadly exists in climate science is correct Then say that, rather then falsely report the statistically tortured "97%". > and there is support for the precise number of 97%. [0] I see only more rhetorical support for the number, not mathematical support. Not that it matters really, just sayin'. > But your point is also wrong: I am not a climate scientist, and the whole discussion can be followed with a bit of high school physics, chemistry and biology, plus Wikipedia to fill in the gaps. Your next point shows that you were unwilling to put in that effort, but it isn't actually that hard ... It may not be hard to "follow it", but what does that expression mean, and what does it not mean. I'm fairly certain it doesn't mean understood at a level sufficient to form epistemically sound conclusions. Again, high marks for persuasiveness, but I don't find this line of reasoning convincing. But that's me. To be clear: I say this not in dispute of the actual science of climate change, the axe I am grinding is related to the nature of the conversation around it. I am hoping some people might become more willing to consider the possible detrimental consequences of the way this topic is discussed.

svara 7y ago
I feel like I have to address this "97% of scientists" point now, since this is coming up again and again. I did not do this so far because it's ultimately beside the point. Climate science studies the climate, and not climate scientists, and I don't find the exact percentage of climate scientists agreeing or disagreeing with something at all interesting or relevant. Worse, harping on about that point makes it sound as if this was ultimately about some sort of popularity contest. Well this is not a popularity contest, it's a scientific issue where there is a well-known method for making progress: You present logical arguments and counter-arguments, based on facts from credible sources (or original experiments). That's the only thing I'm interested in, and the he said / she said style of arguing that may get you points on reddit and twitter is super boring to me. I've tried to make this point in various ways already, but it seems to be, possibly willfully, misunderstood, so I'm going to actually comment in detail on "refutations" of the Cook et al. 2013 paper [1]. The criticisms that I could find in the discussion section are that 1. paper abstracts are manually interpreted 2. there might be selection bias 3. while scientists might agree, they might disagree on the degree to which warming is occuring or to which it is a problem Regarding 1.: Partly correct, misleading. Manual interpretation is a perfectly fine way to deal with natural language data. However, Cook et al. did not rely on this method alone. They also asked authors to self-assess. They compare their manual abstract rating to the self-assessment, and find the self-assessment to lead to substantially less "No position" ratings and more "Endorsment" ratings, while barely changing the "Reject" ratings. Regarding 2.: Obviously correct. This is a potential issue in any empirical sociological study. In this case, however, the issue is limited, since 1) the total pool of climate scientists is limited, 2) the sample of papers (11944) and of scientists asked to self-assess (8547 asked, 1200 responses) is high and 3) the result is consistent with that of other similar studies [2]. Regarding 3.: While I agree that this would be more interesting than just distinguishing between rejection / endorsment / no position, it simply wasn't the question that Cook et al. were looking at. Importantly, none of these points are particularly insightful, yet they were presented as if they constituted some sort of hidden knowledge uncovering a vast conspiracy. Points 1 and 2 are discussed by Cook et al. in their Discussion section, right there in the original paper. Point 3 isn't discussed because it's outside the scope of the paper. So sure, you can now go and say "they used a sampling strategy, so there is the potential issue of sampling bias" and that is technically the truth. It is also incredibly disingenuous, since every single scientific work ever published has limitations. These limitations are understood and discussed by the scientific community. As long as no major flaw is found that invalidates the findings, the result stands. This is just how science works. What else are you asking for? A mandatory self-assessment by every climate scientist in the world? Taking a sample and reporting statistics on the sample (while discussing any potential limitations to your sampling strategy) is the correct way to do it. It is therefore totally fine to say that 97% of scientists agree with artificial global warming, based on the study by Cook et al. Depending on how exactly you ask the question, who you ask, and when you ask, you may get other results than 97% (Cook et al. have reviewed this in detail in [2]). That is not a contradiction, it is totally expected. [1] https://iopscience.iop.org/article/10.1088/1748-9326/8/2/024024 https://iopscience.iop.org/article/10.1088/1748-9326/8/2/024... [2] This is reviewed in Cook et al. 2016, https://iopscience.iop.org/article/10.1088/1748-9326/11/4/048002 https://iopscience.iop.org/article/10.1088/1748-9326/11/4/04...

svara 7y ago

I feel like I have to address this "97% of scientists" point now, since this is coming up again and again. I did not do this so far because it's ultimately beside the point. Climate science studies the climate, and not climate scientists, and I don't find the exact percentage of climate scientists agreeing or disagreeing with something at all interesting or relevant. Worse, harping on about that point makes it sound as if this was ultimately about some sort of popularity contest. Well this is not a popularity contest, it's a scientific issue where there is a well-known method for making progress: You present logical arguments and counter-arguments, based on facts from credible sources (or original experiments). That's the only thing I'm interested in, and the he said / she said style of arguing that may get you points on reddit and twitter is super boring to me. I've tried to make this point in various ways already, but it seems to be, possibly willfully, misunderstood, so I'm going to actually comment in detail on "refutations" of the Cook et al. 2013 paper [1]. The criticisms that I could find in the discussion section are that 1. paper abstracts are manually interpreted 2. there might be selection bias 3. while scientists might agree, they might disagree on the degree to which warming is occuring or to which it is a problem Regarding 1.: Partly correct, misleading. Manual interpretation is a perfectly fine way to deal with natural language data. However, Cook et al. did not rely on this method alone. They also asked authors to self-assess. They compare their manual abstract rating to the self-assessment, and find the self-assessment to lead to substantially less "No position" ratings and more "Endorsment" ratings, while barely changing the "Reject" ratings. Regarding 2.: Obviously correct. This is a potential issue in any empirical sociological study. In this case, however, the issue is limited, since 1) the total pool of climate scientists is limited, 2) the sample of papers (11944) and of scientists asked to self-assess (8547 asked, 1200 responses) is high and 3) the result is consistent with that of other similar studies [2]. Regarding 3.: While I agree that this would be more interesting than just distinguishing between rejection / endorsment / no position, it simply wasn't the question that Cook et al. were looking at. Importantly, none of these points are particularly insightful, yet they were presented as if they constituted some sort of hidden knowledge uncovering a vast conspiracy. Points 1 and 2 are discussed by Cook et al. in their Discussion section, right there in the original paper. Point 3 isn't discussed because it's outside the scope of the paper. So sure, you can now go and say "they used a sampling strategy, so there is the potential issue of sampling bias" and that is technically the truth. It is also incredibly disingenuous, since every single scientific work ever published has limitations. These limitations are understood and discussed by the scientific community. As long as no major flaw is found that invalidates the findings, the result stands. This is just how science works. What else are you asking for? A mandatory self-assessment by every climate scientist in the world? Taking a sample and reporting statistics on the sample (while discussing any potential limitations to your sampling strategy) is the correct way to do it. It is therefore totally fine to say that 97% of scientists agree with artificial global warming, based on the study by Cook et al. Depending on how exactly you ask the question, who you ask, and when you ask, you may get other results than 97% (Cook et al. have reviewed this in detail in [2]). That is not a contradiction, it is totally expected. [1] https://iopscience.iop.org/article/10.1088/1748-9326/8/2/024024 https://iopscience.iop.org/article/10.1088/1748-9326/8/2/024... [2] This is reviewed in Cook et al. 2016, https://iopscience.iop.org/article/10.1088/1748-9326/11/4/048002 https://iopscience.iop.org/article/10.1088/1748-9326/11/4/04...