Lewandowsky by way of Watts Up With That
So now there’s a conspiracy theory going around that I didn’t contact them.
Does this guy not know the definition of conspiracy?!
Given the lack of evidence that he tried to contact skeptic blogs, and his bizarre excuse for not reporting the blogs he tried to contact when describing his methodology, some people suspect he didn’t try very hard to contact skeptic blogs. But that suspicion is not a conspiracy theory.
The evidence that Deltoid posted the survey before Lewindowsky’s graduate student’s first contact with Climate Audit and that Lewindowsky may have distributed links to different surveys to different blogs is leading people to suspect him of incompetence. But suspecting Lewindowsky is incompetent is not the same thing as believing in a “conspiracy”.
Here’s why it’s not: An academic exhibiting incompetence when conducting, analyzing or describing a survey methodology does not meet the dictionary definition of “conspiracy”.
Now, I’d like to move on to the issue of releasing the emails. Lewindowsky now tells us,
I would love to be able to release those emails if given permission, because it means four more people will have egg on their faces
I’m glad you have joined me in wanting you to release the emails. But you are mistaken that even one blogger will have any egg on their face as a result of this. I certainly won’t have egg on my face. What I’ve said is this:
My answer to her [Jo Nova] was: No. He didn’t ask me.
Well…. Thinking a bit more: I think it’s more accurate to say don’t recall being approached. Mind you: I’m not sure that I would have remembered a request.
Lewandowsky hasn’t reported who he contacted, I don’t recall him contacting me- but I wouldn’t necessarily remember if he did. I’ve said this. It’s absolutely true and will remain true even if I turn out to be on the list.
I do not see how I would have “egg on my face” if it turned out he contacted me. Guess what? I didn’t think the revelation that I might be on the list would result on any egg on my face when I wrote:
If Lewandowksy did contact me in my capacity as operator of The Blackboard, I grant him permission to reveal this to the world. In fact, since I would prefer he reveal that I was one of the bloggers I do not grant permission for him to refuse to reveal I was contacted.
My motive in writing that was simple and can be inferred from plain meaning of my words. Given that Lewindowsky claimed to refuse to publicize the names of the blogs because he presumed the information was private, I recognized that if I wanted him to reveal the information about me, I would need to a) to grant him permission to reveal my so-called private information and b) possibly state that I refuse to grant him permission to keep it private.
So I did granted permission publicly. This has nothing to do with harboring any “conspiracy theory”. I requested him to release information because I want him to release it.
As for soliciting others to follow my lead: If the reason for keeping their identities secret is to maintain their privacy, I wanted to encourage them to grant permission for the information to be public.
I wanted them to request their information be released because I would like to know more about which blogs were solicited and Lewindowsky evidently cannot do it if he does not get permission to do so. This has nothing to do with harboring any “conspiracy theory”.
I have absolutely no idea where anyone would get it into his wool-filled brain cavity that giving him permission to release information he claims to wish to release is evidence that I or anyone else harbor a conspiracy theory. I also don’t know why he thinks anyone would have egg on our faces if it turns out we are on the list. We are asking precisely because we want to know. Moreover, we are asking the information be shared because we want others to know.
I would also like to respond to his insinuation that we haven’t some how looked hard enough for the emails. I can only speak for myself, but I am happy to reveal why I am not going to look harder.
Conducting his survey may have been important to him at the time but it’s really nothing to me. I do not think its importance to him compels me to maintain records of our email exchanges for his sake. I does not compel me to burn email exchanges with perfect strangers into my memory nor to resurrect the hard drive which died in 2011 so that I can search for any emails he might have sent me in 2010.
If it turns out that I am one of the bloggers on his list and yet he refused to tell me I was on of the bloggers on the grounds that he presumed privacy requires him to conceal my own communications from me, I will certainly conclude that he is either a) stark raving mad or b) was using privacy as a screen to not fully describe his methodology. Neither of these opinions is believing in a “conspiracy”.
I will also say that if I am one of the bloggers I would want to be provide copies of all emails so that I can learn the dates on which exchanges took place and read the wording of the invitation. If Lewindowsky thinks my wanting to know the extent of my own involvement in his project is a “conspiracy theory” then a) he is nuts or b) he doesn’t know the definition of the word conspiracy. I’m currently inclined to think both. My thinking this does not amount to harboring a “conspiracy theory”.
I don’t know about any conspiracy theory, but if you click on Mosher’s link, it takes you to Breitbart. ??
jim2–
Yeah. He used to blog there. . .
I think it’s a conspiracy that you are denying there’s a conspiracy!
Ray– Opps! I forgot to send emails to my conspirators to describing my “plan” to deny Lewindowsky’s accusation that we all believe he conspired with himself to publish methodologically flawed paper! I wish I could have been a fly in the wall eavesdropping on the conversations wherein Lewindowsky described his plot to himself and got Lewindowsky to agree to use an online survey that permits people to fill it out using anonymous proxies like “hidemyass” and then write whoppers like this
Uhmm… given that only 10 answers indicated the respondent believed the NASA landing was faked, it wouldn’t take any substantial subset of the 1,000 to skew the result highlighted in the freakin title of the paper. It would take at most 10 faked survey results. That’s 0.1%.
Moreover, given that you can use “hidemyass” or any proxy at kwiksurveys, and the surveys were supposedly up for more than a month it wouldn’t take a “substantial subset” of the 1000 to conspire to subvert the survey. It would take 1 count ’em 1 person who had inclination, some time and some scripting skill to skew any result they wished to skew. A commenter at Deltoid said he could easily write that script and you know what he probably could have done so.
I bet Brandon Sholleberger could write the script in an afternoon– and he’s not unique in this regard. (It would take me a week…. I’d have to look lots of stuff up and the exercise would involve lots of cursing.)
I have been following the exchanges on various blogs and the comments from the “pro science” side have been strangely contorted by the need to use the words “conspiracy theory” in ways that make no sense. The “deniers”/sceptics have basically asked a couple of pointed questions – who did you send the survey to, can you explain more about what seems to be a flawed survey design? etc – and back come screeds of ad hominem abuse referrring to conspiracies and gish gallops.
The increasingly baffling Eli Rabett even says “Conspiracy theories are important parts of the denialists’ world view.”
It is clearly a concerted attempt to paint people who do not share their alarmism as whackos. However they seem unaware that they are coming across as kids in a playground.
Speaking of deceitful behavior and conspiracies..
This one’s a real whopper.
Here we have Michael Mann orchestrating with John Cook and SkS to get favorable opinions of his book reviewed. Some effort appears to have been made to remove negative reviews. Perhaps Amazon should explain to us their policy on critical reviews of books, and why negative comments were deleted, and positive solicited comments–against Amazon’s own policy–have allowed to remain in place.
Perhaps willard can explain how what Mann and Cook have is not in any way engaged in deceitful behavior. It’s just another version of the same class of behavior as sock puppetting (this one may be more closely related to astroturfing), namely deceitful manipulations to make it appear like there is more positive support for an idea (or book) than really exists.
I bring it up here because this was an organized effort to deceive, and hence a real conspiracy.
I can imagine the howling and screeches of indignation from the usual suspects, were McIntyre or Watts were to engage in the same behavior.
Lucia:
Hm… forget demographic adjustments, this makes you wonder what the p value was for Lewendowsky’s inference given in the title of his paper.
Uhmm… given that only 10 answers indicated the respondent believed the NASA landing was faked, it wouldn’t take any substantial subset of the 1,000 to skew the result highlighted in the freakin title of the paper. It would take at most 10 faked survey results. That’s 0.1%.
Beyond that, only three of those ten were from “skeptics.” Three, out 1,145 responses support the title of his paper, and he thinks it would take:
To cause this? A substantial subset conspiring together to generate three responses? Seriously?
If you want every survey response to show up from a different IP, writing a script would be a bit of work. But even then, it wouldn’t be that hard or time-consuming. I’m sure there are (at least) hundreds of people across climate blogs who could do it in a day if they wanted to.
Carrick:
The trick, I imagine, is Lewendowsky grouped conspiracy theories together and reached a conclusion. After that, he naturally extrapolated that conclusion to every conspiracy, or at least to the moon landing conspiracy.
Because actually looking at how many responses showed skeptics claiming to believe the moon landing was faked (3 out of 1,145 responses) isn’t necessary when you already know the answer.
Bradon
Notice the paper focuses on the number of respondents (~1000) rather than the number of positive responses which support the claim in the paper title (3).
Of course it might seem that you need a lot of responses to skew a sample of 1000. For example, if you are trying to discover whether a majority believe A or B and you conclude the majority believe A because 90% said they need A, then you would need a lot of cheating for the correct answer to be B.
But… well…it’s not hard for 3 drunks at happy hour to decide to fill out a survey. They don’t even need to be 3 drunks at the same gathering. It could… well… let see…. Remember TCO? You think he might not spoof a survey? Or how about “Fitzcarraldo/Andrew/ Cassandra/ Marian/ Andrea/, Marie deschamps/ Albert/ Laura Gonzales/ Rebecca/ Edward/ Stephan/ Reader”, you think he wouldn’t spoof a survey. Maybe each would spoof in a different direction. But … still..
Heck. I might spoof an online survey. Depends.
“So now there’s a conspiracy theory going around that I didn’t contact them.”
Of course, that is something one would say if he were intent on showing that the opposition to his views on AGW were from those associated with some crackpot thinking on other issues and doing it in the manner not of science but pure politics. This is a favorite ploy of politicians and some of those who might comment in the MSM in defense of a partisan point of view.
.
Alternative rendition:
.
Given, but failling to see, the very strong evidence that he actually contacted said skeptic blogs (as in, the actual contacting emails themselves), and the very good reason why he couldn’t release the names before asking his board (which he duly did), “some people” immediately fired up the customary suggestions of fraud, misconduct, deceit, fabrication, etc. (“whole cloth” anyone?)
.
Which has nothing to do with conspiracy theorizing. At all. Not a sausage.
.
Or, alternatively, which might actually provide the strongest evidence for the (admittedly unduly emphasized) minor conclusion of the paper. 🙂
.
Kenneth:
.
Yes. And that is also something one would say if it were true.
“Which has nothing to do with conspiracy theorizing. At all. Not a sausage.”
Correct – just the mendacity and lack of awareness of spam filters of the person conducting the research.
“So now there’s a conspiracy theory going around that I didn’t contact them”
Sounds like a conspiracy.
Andrew
toto. just how would you characterise the foaming at the mouth responses of someone called Barry Bickmore at this reference…
http://www.shapingtomorrowsworld.org/lewandowskyVersionGate.html
he is said to be a scientist but he all I can see are the scary qaulities of a manic raving lunatic.
lucia:
Indeed. What this entire saga suggests to me is Lewandowsky either did a terrible job and didn’t even bother to check a basic and important fact about his data, or he intentionally decided to deceptive. Either way, it’s a horrible sign.
Incidentally, the link you provide for the quote at the top of your post is wrong. Anthony got that quote from a Desmog piece (which quoted Lewandowsky), not from the piece written by Lewandowsky.
By the way, it may just be that Lewandowsky was taking some creative liberty, but… I’m disturbed by this quote from him:
Even if such an idea was going around, it wouldn’t be a conspiracy theory. At the very worst, it’s a theory that one guy lied. How does that amount to claiming there was a conspiracy?
Also, for those still interested in actual data, here’s what I get when I remove all the lines in which a 4 appeared anywhere in the “conspiracy” sections:
.
– Correlation b/w “mean conspiracy” (excluding conspiracy-climate) and “causeCO2”: -0.1246 (p ~= 4e-5)
.
– Partial correlation b/w “mean conspiracy” (excluding conspiracy-climate) and “causeCO2”, controlling for “mean free market”: -0.1844 (p ~= 1e-9)
.
– Correlation b/w “Moon conspiracy” and “causeCO2”: -0.0830 (p ~= 0.006)
.
– Partial correlation b/w “Moon conspiracy” and “causeCO2”, controlling for “mean free market”: -0.0979 (p ~= 0.0013)
.
Most of the results vanish if I also also remove the lines with conspiracy 3’s, however.
.
tl;dr: Lewandowsky’s results don’t depend on the “4” responses.
toto (Comment #102762)
And you think he really attempted to contact them.
Either the guy does very sloppy work (and that is something not beyond the realm) or he was in such a big hurry to make some preconceived political point that he failed to check whether he actually had made contact (my favorite) or he consciously neglected blogs that might negate the point he was attempting to make or he did contact skeptical blogs and they conspired not to publicize his requests because they were afraid of the responses that might put them in a bad light or he tantalized the skeptic blogs with a half-hearted effort to contact them knowing full well that they would see it as something that he could build into a conspiracy response.
You could pick any of the above or none because what is really obvious is that the work is incredibly bad science, and further, regardless of its conclusions, legitimate or otherwise, it has no bearing on the AGW debate and finally shines a bad light on the peer review system in his field – if this paper were published.
toto:
So you’re endorsing that there is a conspiracy theory, when there obvious is not one? What’s your position on moon landings?
BTW, you don’t see anything humorous about Lewindowsky finding conspiracies under every rock and accusing everybody who disagrees with him of being conspiracy nuts?
toto (Comment #102769)
Another favorite tactic I have seen with mushy science is a crappy paper gets published and then after due time that paper is used for references in another paper and evidently with the hope that the reviewers and readers forget how crappy that reference was or worse taking the papers worth based on conclusions that might agree with the paper now under review. These papers also are great fodder for political debates where a politician wants to back a claim with something that sounds legitimate to the uninitiated.
OK, time to come clean. I filled out the survey on a non-sceptic blog saying I was a sceptic and thought the landings were faked, and got two sceptic friends to do the same. It was only a joke! Sheesh – everyone making a federal case out of it, articles in the bloody Guardian, etc. What’s the world coming to?
I would put ‘/sarc’, but Lewindowsky seems so stupid, what would be the point?
Dear lord, I just followed the link posted by diogenes, and… Wow! Lewandowsky says:
Now then, I’ve left off a lot of pathetic drivel from Lewandowsky, but even so, this is hard to swallow. He said the reason different surveys were given to different groups of people was “counterbalancing.” That’s a method used to try to prevent the order of questions from influencing the results.
This means his plan was to give different groups different surveys because he worried each individual survey would generate different, inaccurate results. He then planned to look for differences amongst various groups.
His plan was to introduce artificial differences between groups then look for differences between those groups.
Brandon
So it appears.
I posted comment 29 here: http://www.shapingtomorrowsworld.org/lewandowskyVersionGate.html#comments
lucia, nice comment. I think it pretty well summarizes the problems with that paper.
By the way, it made me finally realize the surveys were done two years ago. I assumed they were done recently because of how little effort the paper shows having been done. I think I even disagreed with Steve McIntyre on a point because of it (sorry about that, it looks like you were right).
I wonder if maybe there was difficulty getting the paper published, and that’s why there was such a delay. If so, it would restore some faith in the peer review system.
By the way lucia, you’re dumber than a first grader. And you have poor memory because you may not be able to remember some random e-mail from two years ago. Or at least, so say Tyger, two comments after yours:
I’m not sure how it’s a logical fallacy to point out you cannot (practically) search for e-mails on a harddrive that no longer works, but it’s good to know a first grader could manage it!
And remember, even if you gave your permission to an author to identify you are someone he contacted, he couldn’t:
Because it’s only okay to quote “private” e-mails, not say who was sent such e-mails after they’ve given you permission to do so. The only appropriate conclusion I can come up with for this comment is the same one Tyger used:
Carrick:
.
The post above yours has Kenneth insisting that SL’s team didn’t contact skeptic blogs, despite evidence to the contrary. A post below yours has Brandon insisting that their randomizing question order (aka “duh!”) is some kind of evil plot to “introduce artificial differences between groups then look for differences between those groups” (because random differences are so helpful in supporting directional effects).
.
If you think that this has nothing to do with conspirational thinking, then I guess we just have a vocabulary mismatch.
.
Seriously, it’s as if you guys had suddenly decided to to prove his point better than his paper does. It’s so weird I’m beginning to suspect some kind of hyper-elaborate meta-conspiracy! 🙂
toto, I would appreciate it if you didn’t make things up about me:
Failing that, I’d appreciate it if you did so in less idiotically obvious ways. There is no way anyone* would read my comment and believe I was “insisting… [there] is some kind of evil plot.” If I must defend myself against fabrications, I’d like those fabrications to at least be more than someone wildly waving their hands in the air calling me a loon.
*Well, anyone who wasn’t incredibly bad at reading.
toto
Brandon didn’t assert a conspiracy theory. He simply observes the effect Lewindowsky’s non random distribution would have on the results. If there is an question ordering effect on answers, distributing different surveys would have the effect of introducing artificial difference between groups. The reason is that the different surveys weren’t distributed randomly.
Everyone at blog A got blog A’s version.
Everyone at blog B got bog B’s version.
Everyone at blog C got blog C’s version.
So, in this case, if the answers were affected by the version and blogs A, B and C have different demographics, you could no longer determine whether one demographic groups answers were different from anothers because that demographic group got a different survey or if it happened because those two groups really thought something different.
If you wanted to randomize what you need to do is set up a script that ensures that a person coming from blog A has a 1/3rd change of getting survey A, 1/3rd chance of getting B and 1/3rd chance of getting C and so on.
This is especially important in an online survey where L can’t know which blogs will end up participating nor which blog readers will be most enthusiastic in responding.
Brandon, it’s worse than that. Lucia is now “some obscure person†and “an extreme right winger†to boot. It certainly made me giggle. There are some people on all sides of this issue who seem to never venture out of their safe little echo chambers. I think Tyger and Sou might be prime examples.
Laurie–
Well, if I’m “some obscure person” then we can presume I won’t be on Stephan’s list of five. Right?
toto:
Didn’t he use the words “very sloppy work”? I think that’s a reasonable interpretation of his comments.
Your’s not so much.
Your defense of the use of a sow’s ear to make a silk purse is…illuminating.
Lucia
I think it’s far more likely you won’t be on SL’s list precisely because you’re not “some obscure personâ€. In the minds of SL and his SkS advisors, Steve Mc and his denizens are out and out deniers. You and your denizens, on the other hand, are known as “lukewarmers†and, therefore, much more likely not to give the kind of answers SL was looking for 😉
Claiming the side one wishes to denigrate is caught up in a wacky conspiracy is the lazy person’s way to avoid a discussion.
Reading Lewandowsky’s increasingly shrill and cowardly dodges over at his blog makes it clear that he is a very lazy intellect getting more and more frustrated at having to actually answer questions and provide evidence.
His brain muscles, so to speak, are not in very good condition to actually engage intellectually. He has to resort to phony surveys, sham results, pal review of his work and hiding out behind name calling.
He is actually going farther than Gleick- Gleick, after getting caught, had the shame to make at least a partial admission and then simply go quiet. Lewandowsky is more like the naked emperor, wanting to put that noisy kid in jail for laughing.
toto,
Sending some e-mails to a blog owner about a possible survey is not really contacting a blog owner about participating in a survey.
Your assertion otherwise makes you look really dim.
SL is obviously dodging the issue and hoping he can tease and belittle all questioners he dislikes.
The results of the survey are nonsense, and I would suggest it is because SL’s survey was at best nonsense.
“And you think he really attempted to contact them.”
Obviously the meaning of my comment is that: Do you think based on the evidence now available he made a concerted effort to make sure the “skeptic” blogs he or someone else contacted acknowledged they were contacted and either provided the readers with the details of the survey or officially declined by way of email? I really do not see how a survey could be conducted with any scientific import that would not have made those efforts. Any scientist worth his salt would show that that effort was made – if indeed it were. If you wish toto, you can answer my question.
A complete analysis would in my mind have to consider and publish at least, if not in name, a category for the declining blog (a polite and scientific version of crackpot to thoughtful) and a reason given for declining. Surely it would be important for the evaluating the results to know if there were perhaps more thoughtful blog posters who might say I did not respond because I think your survey is unscientific or has no meaning in the discussion of the scientific evidence for AGW and its consequences or in the policy debate on mitigation..
And by the way, I am personally conspiracy free in matters political and AGW science. I think most of the errors and weaknesses in AGW related papers, and on all sides of the issue, are related to a very human weakness that allows a person intent on finding evidence for a theory to conclude the work at the point that the evidence favors the theory. I think that tendency explains the lack of sensitivity testing one sees in climate science papers. It might also explain the peer review process where reviewers have strong positions on the matter allowing papers to be published that thoughtful analysis later finds wholly lacking in supporting evidence or applied methodologies.
The other general tendency that I see and I see again as a human weakness: is the inability for some authors of these papers to admit even small mistakes have been made and evidently for fear that it might effect, not the scientific endeavor, but an advocacy stand on the matter.
A conspiracy would require more than one person anyway. Lewindowsky is apparently schizophrenic, since he is under the impression he is more than one person.
Kenneth–
I would add that since the time from survey to publish is roughly 2 years, it makes very little sense to act with ridiculous haste in the early stages. If he really wanted skeptic blogs to participate he ought to have
1) continued to invite skeptic blogs until he got some who were willing to post the survey.
2) requested that the willing alarmist blogs hold off publication of the link until he’s found a reasonable collection of blogs to post links. That is: The appearance of the link at deltoid should not have predated the email invitation to CA– and should in fact not have appeared until he had found some skeptic bloggers.
I can’t begin to guess which 5 skeptic blogs were emailed. I can think of hundreds of reasons why they might not have participated– but the fact that the survey was already “out there” at deltoid would certainly constitute a valid reason if that were the case.
If nothing else, many bloggers like “scoops”. The chance to discuss a stale story is, from a blogger perspective, an opportunity one might decline to accept. If the story is stale, and of no particular interest to the blogger, the blogger is likely to put it on his “look at this later” pile, and then…well.. never look at it.
It may be that people who are not bloggers don’t know this. But if someone is conducting a survey that will be distributed on blogs, it is bizarre not to educate themselves about factors that affect a bloggers interest in a story.
Unless you start from the assumption that climate sceptics ONLY read climate sceptical blogs (which would be a bad thing), it really doesn’t matter much on which site(s) the survey was posted.
You’ve spelled Lewandosky incorrectly at the title of this post. If that was intentional given the dictionary thing then the joke was weak IMO. That you’ve misspelled it again but differently in the next post is serial bad form.
Jules, that’s silly. To demonstrate, I’ll offer the simplest counterexample possible. We could assume skeptics don’t read the handful of blogs the survey was posted on. We could assume they read dozens of non-sceptical blogs, but not those particular ones.
When such an obvious refutation exists to a claim, it makes me wonder how that claim gets made. It certainly doesn’t make me feel inclined to point out the dozen or so other refutations that exist for the claim.
Jules–
I don’t think that’s true. I don’t think “ONLY” is the issue. The differential distribution or different surveys at different blogs can result in bias if skeptics tend to read skeptical blogs and warmists tend to read warmist blog. Because in this case, skeptics and warmists will tend to fill out a different survey. The degree to which the effect matters will depends on how strong the tendency is and how much the order effect matters in answering the question. But if it’s strong enough to bother to create different surveys, one can’t then simply decide it’s not strong enough when considering the effect of giving different groups different surveys.
I would also suggest the that the tendency for different groups to encounter different surveys would be stronger that you might think. Bear in mind that to find that survey a person has to see that a specific post. If a link is posted on Deltoid a regular reader of Deltoid is likely to see it and moreover, they are likely to see that version first and fill it out. If they happen to see it again at WUWT they are likely to not take the survey again.
But, in fact, a regularly visitor of Deltoid is not particularly likely to be a regular visitors of any particular skeptic blog. That is: They probably don’t visit WUWT every single time that blog updates and read every single article. So, even if they sometimes visit WUWT, they aren’t likely to see the post discussing the survey. Mostly likely, most of the time, they will see posts at WUWT that someone linked from Deltoid. And that’s not likely to be the post announcing a survey. It’s more likely to be a post someone at Deltoid was outraged about.
So, I don’t think ONLY is the operating principle.
Quite honestly, I think back in 2010, my blog would have been a rational one to put on the list. Maybe it was. Maybe it wasn’t. I’d like to know– but I don’t. 🙂
My apologies and red face…Lewandowsky.
FWIW: I rarely read any of the blogs on the list of 5 warmist blogs. The only reason I would visit is if someone else linked them. I did read Lamberts around the time he debated Monckton. But the fact is, my impression is he often blogs things I have absolutely no interest in. Ok… August is an open thread. September is an open thread. Sorry.. but why would I have visited at all?
I’m wondering how any useful results can be gleaned from any online survey when you don’t know if the respondents are being honest or not in their responses.
How would one verify responses without, say, knowing who responded and then having them take the test at a later date to check for consistency?
This ‘survey’ can be likened to the surface temperature record debacle. Take some figures that cannot be validated, pound them into a desired shape with statistics, adjustments, homogenisation and fudge factors then build a ‘theory’ around them that is consistent with one’s beliefs.
I’m pretty close to calling it. This survey is starting to look like a How To Screw Up an Online Survey.
Believe it or not, it is possible to do valid research with online data collection. It’s a lot of work but it can be done.
Lewandoski didn’t do any of the work. He’s bringing an entire sector of research into disrepute.
Lucia wrote: “The differential distribution or different surveys at different blogs can result in bias if skeptics tend to read skeptical blogs and warmists tend to read warmist blog. Because in this case, skeptics and warmists will tend to fill out a different survey”
i’m not sure about this one Lucia. Unless people don’t answer honestly (which btw isn’t just a problem with online polls), it really doesn’t matter much.
A skeptic will not suddenly become pro-science if he follows a link on a pro-science blog. They will *always* tend to fill out a different survey. The question is if this affects the results of the poll…
Don’t forget to bear in mind it is *not* a poll trying to figure out absolute numbers, but a poll trying to find correlations (of course there need to be enough answers to be statiscally significant, but that’s another question).
I don’t think anyone reads every article on a blog, no matter what side the blog is. Life’s too short. I can only hope people are aware of the risk of a confirming bias, and are aware only reading one side of a debate never is good thing.
Btw, not that it’s that important, but it’s not fair to mention the lack of posts on the 2012 Deltoid as an (supportive) argument to claim there was no reason to visit Lamberts blog back then. In February 2010 (the only month i checked) Lambert blogged on daily basis.
jules
i’m not sure about this one Lucia. Unless people don’t answer honestly (which btw isn’t just a problem with online polls), it really doesn’t matter much.
No. This isn’t right. When handing out different forms of surveys, the intention is to correct that fact that the order of the questions is known to affect the answers when people answer honestly. So, even if people answer honestly, handing out different surveys to groups with different demographics will make it impossible to untangle the “question order effect” and the effect of demographics.
If the order of questions doesn’t matter then there is absolutely no good reason to hand out different surveys. So, the fact that different surveys are created is based on the concept that the order or questions does matter.
Sure. But it happens that I do not regularly read a single one of the blogs on that list. This is hardly surprising. Most are very, very low traffic. Almost no one reads them. Two or three have high traffic. I tend to avoid blogs where commenting is difficult. I’m banned from Tamino’s so I rarely visit it. Deltoid and SkS rarely have anything interesting on them.
I do read Annan’s blog, Real Climate, Rabett Run, and Planet 3.0. (The latter is to be sure I get informed on the current alarmist memes!)
I click links from those to other places. I read Neven’s blog during melt season. I read Nick Stokes blog from time to time. I read collideascape. I don’t know which of the warmist blogs other lukewarmers read, but my impression based on blogs people link in comments their choices tend to overlap with mine.
I’m just trying to explain the fact that I don’t visit it. I don’t recollect ever visiting if. I do read it sometimes. But it’s not a weekly thing. Certainly not daily. It seems to me in the past there were lots of posts on mosquito nets and other topics that didn’t attract me (thought I commented once or twice).
I did read it during the period when Tim was debating Monckton.
But it’s never been on my “must read this this week list”. I suspect my blog isn’t on his list either. He probably visits The Blackboard on more or less the same basis. That’s the way of blogs.
Similarly, I rarely visit Stoat because it seems like 3/4 or more of his posts are about running, rowing or other cruft and I don’t find the few that have anything to do with climate particularly interesting. But I have visited it when people mentioned his discussions of ice. My path to visiting that blog is: someone mentions it, if I think the content is interesting, I visit it.
So, I visit other warmist blogs– some regularly whether or not anyone links others mostly only when someone links.
I cannot agree with Jules on this one. There is a diversity in populations between skeptics, agnostics and warmers, otherwise there wouldn’t be a differentiation among the populations in terms of blogs they frequent.
As a general rule I don’t read blogs where the author can’t think clearly or there are extreme moderation rules in place (these are mostly warmer blogs btw).
If you are trying to study the attitude of skeptics you need to study them at the blogs they frequent and not just hope you get a “drive by” and get lucky. Moreover, experimental design is an important thing in social science, because it is how you avoid the trap of the “false positive.”
If your design is to have five skeptic blogs, you need to go with your goal of five skeptic blogs, not just pick five at random and accept silence for “no”.
From the referenced paper (Table 2):
Lewandowsky’s study fails at step one.
But not only at step one.
It does appear that the paper’s reviewers were indeed his peers, at least with respect to their understanding of experimental design.
Once you label me you negate me.
-Soren Kierkegaard