Talk:John Brignell

From SourceWatch
Jump to: navigation, search

I reverted the changes made by

  • "However, According to a report in the New York Times in 1999, Materia cases do now stand at 2.7 Million. It is possible that his figures were calulated using the assumption that, if DDT had been used correctly and completely, malaria would no longer be in existance. [1]"
the slightly diff figure in the NYT from 1999 doesn't add much ; the rest is speculation. Aside from which I have never heard of public health community suggesting malaria could be eradicated but rather controlled.

It is possible to clarify his claims with the following analogy. You may see relative risk expressed as a ratio (ex: 1.5). This number 1.5 means, for instance,that a woman with a particular risk factor is 1.5 times more likely to develop breast cancer than someone without that risk factor.

  • "Another way to express this number is to say that the risk factor increases relative risk by 50%. However, it's important to ask... "50% increase relative to what?" If the baseline risk level we're comparing to is 1%, a 50% increase isn't as meaningful."
this is a common argument of risk managers but Lamberts quote takes issue with the suggestion that extra risk is claimed not to be statistically significant. It clearly is - we are not talking about something with a string of zeroes in front of it. And if you are one of the extra 0.5% who get the diseases then it is very statitically significant.
  • although his logic is based on the fact that scientists were not looking for it prior to the 1970's, hence his claims;
this addition implied this came from the Lambert citation itself; but rather than add the ref mid-sentence it is not clear that is what Brignell'a basis is. Even if it is then the absence of evidence hardly supports the claim that the ozone hole was always there.--Bob Burton 16:26, 7 Sep 2005 (EDT)

the following was inserted at the top of the article today; I've removed it to here:

What follows is the work of an individual known as The Adhominator. You can recognise his style, as he never attacks the argument, only the arguer. You can identify him, because he is the only authority he quotes. Enjoy!

Earlier this morning I received an email from Brignell complaining that a pre-amble warning readers about Lambert's blog posts kept being removed. (It is posted above). After Diane removed the latest addition of this, a few hours later removed the entire section on Brignell and his views on DDT, the ozone hole and a Lancet study. I reverted this and have rewritten the section. I'll post a note offline to Brignell explaining the reasons. --Bob Burton 01:07, 30 Sep 2005 (EDT)

I've rolled back the following change made by from the article as it was not written to meet SourceWatch editorial conventions:

Note: If you have a solid understanding of complex statistics, check out the explanation of relative risk on Brignall's site and you will see that he clearly demonstrates why a relative risk of 1.5 is not significant. He is not just quoting some book, he produces the mathematics. Tim Lambert clearly does not understand complex statistics. Either that, or he has a personal agenda. [Jim Smith, B.Maths (Newcastle)]

--Laura Miller, 02:31, 4 Oct 2005 (EDT)

recent edits

The statistics in these surveys are not drawn from hundreds of thousands of results, but from dozens, and the samples are not "perfect records", they are samples from populations, which means they are randomly distributed. So the "thought experiment" is obviously invalid.

This seems to be the living embodiment of the old statisticans' joke "Your honor, not only does this data come from a mere sample, but it's a random sample at that!"

A straw man argument is where a weaker version of a theory is criticised and the result presented as though it applied to the whole theory. It is a fact that increasing the sample size of a study always increases the reliability of the results obtained. A study based on "hundreds of thousands of results" must therefore attract a much higher degree of acceptance than one based on dozens. Similarly, a result based on "perfect records" must be more reliable than one incorporating normal levels of error due to random sampling. Being a straw man argument, the "thought experiment" is obviously invalid. (Jim Smith)

Publication, of course, is not the same as being significant. The above example supports Professor Brignell's demonstration that publication bias causes a relative risk that can be approximated to 1.6.

Could somebody elaborate on the above paragraph, as I admit to not seeing what is being asserted. Gzuckier 00:50, 5 Oct 2005 (EDT)

Whenever you perform a statistical test at a significance level of .05, due to the inherant nature of the test there is a 1 in 20 chance that you will get a result that seems significant but isn't. Put another way, 5% of all studies are bogus. When people publish their results, they tend to publish only significant ones; this is called publication bias. If 4 in 20 studies produced significant results, then 1 would be the bogus one, and so 25% of all published studies would be bogus. Of course, 4 in 20 is a guess; no one knows how many studies are not published; but it is possible to do some mathematics based on reasonable assumptions to estimate the effect of the missing studies. That effect equates to a relative risk of about 1.6; if you ignore the studies that produce a relative risk of less than 1.6 then your chance of accepting a bogus study as real goes back down to 1 in 20. (Jim Smith)

The argument that bias from publishing only significant results and not not significant results is involved, is clearly irrelevant to a paper where the finding is a count of something. Of course, the critics may be proposing that in fact numerous such counts of the dead in Iraq have been done, and only the ones which show a significant increase have been published. If that's the case, I strongly urge them or anyone else who has knowledge of these studies to come forward and identify them. Please. It's important. You could be doing the War on Terror a great service. Gzuckier 14:21, 12 Oct 2005 (EDT)

You are conflating two separate points. :-) (Jim Smith)

Relocated from the article page. --Bob Burton 16:04, 5 Oct 2005 (EDT)

Why does sourcewatch not allow Mr Brignell to defend himself against these charges below? How can Sourcewatch maintian credibility if you refuse to allow the Professor to rebutt the charges below. You have blocked him from this site and you refuse to respond to his letters. You refuse to allow him to add a disclaimer to an entry about him! It is baffling. Why you allow this Lambert to make all this stuff up and then you refuse to allow Brginell to respond. That is not fair. You are clearly very very biased against this man; so much so that you refuse to allow him to deefend himself. This does not speak well about materials here. Is every thing published here not to be trusted? Do the people you have written about even agree with what you have written about them?

It seems like you are taking what Brignell claims on his website at face value. John Brignell is welcome to post material on SourceWatch but, like all other contributors, it is expected that he abide by some basic protocols. These are:
  • where major changes are made a brief explanatory note is posted to the talk page (accessed by clicking the 'discussion' tab).
  • Individuals editing a page on themselves are expected to either register under their own name or indicate clearly on the talk page that they were responsible for making certain edits to their ownpage. That way other contributors can make their own decisions as to whether to let the changes stand.
  • Article pages are for articles. Comments and opinions are welcome but only on the talk page.

Despite these protocols being communicated to John Brignell he chose to ignore them. As such his edits were treated the same as we would treat any other contributor or vandal and were reverted to the preceding version of the article.

If Brignell opts to barr himself by refusing to observe the rather basic protocols that all other contributors abide by, then that's his choice.

All of this was communicated with Brignell on the talk page and in direct email with him. --Bob Burton 16:04, 5 Oct 2005 (EDT)

Lambert writes: "It is perfectly clear that the hole (in the ozone layer) was not always there", on the basis of data going back to 1950, and a ~2/3-fold variation in ozone levels.

In fact, there is no good data provided to support any comment before 1950. The moderator thinks that absence of evidence is not proof, and I agree. Will he can Lambert's assertion, on the basis that the facts do not support it ?

yours per

"Lamberts quote takes issue with the suggestion that extra risk is claimed not to be statistically significant."

this is a straw man argument. Brignell says that the risk is not significant, and does not claim that it is not statistically significant. Brignell points out that there are many variables and possible confounders in this epidemiology study, and that the possible effects of these are not estimated, or allowed for, in the formal statistical test. These possible effects could easily overwhelm the small increase in relative risk. yours per

In an attempt to balance the view of John Brignell presented here I have added a section in which I have tried to present his approach to his subject, as would be apparent to anyone reading his website. I expect that the editors will remove it; they don't seem interested in the presentation of a balanced view. (Jim Smith)

Malaria discussion

I've made some changes to the following paragraph. The reasons for the changes are mostly self-evident, but some further clarification is in order here:

It is noteworthy that the use of DDT in Sri Lanka, by reducing the number of deaths per year from malaria from 2.8 million in 1948 to 17 in 1963, had by 1970 saved about 56 million lives. It must be assumed that if it hadn't been banned it could have resulted in similar savings of life in other countries where malaria is a problem. Brignell states that the number of deaths due to malaria "makes The Holocaust look like a dress rehearsal," and one is obliged to agree with him.
  1. I don't know where the author of this passages gets a figure of 2.8 million malaria deaths in Sri Lanka in 1948. According to Laurie Garrett, whose work I referenced in my changes, when the international effort to eradicate malaria began, in 1958 (not 1948), "it was estimated that, for example, about 1 million people had malaria that year in Sri Lanka, some 100 million in India, and untold numbers, roughly estimated in the 'hundreds of millions,' in Africa." Note that this is the number of people who had malaria, not the number of people who died from it.
  2. The author of the above paragraph seems also to be under the incorrect impression that malaria control efforts from 1958-63 were limited to Sri Lanka alone, as well as under the incorrect impression that DDT was "banned." Finally, he seems unaware that there were other aspects to the malaria eradication effort. In addition to DDT, choroquine was used as a medication, and of course a number of traditional, non-chemical techniques were also used, such as mosquito screens, flyswatters and flypaper. By making these errors, the author of this passage (1) exaggerates the significance of DDT to malaria control, (2) misrepresents the number of malaria deaths in Sri Lanka, (3) extrapolates from those errors to the rest of the world, and (4) invents a false cause for the failure of malaria eradication (DDT wasn't banned; the effort ran out of money).

Based on these errors, the author of this passage seems to think he has the right to give orders about what everyone else has to believe: "it must be assumed"; "one is obliged to agree with him." Reality check: No one here is obligated to assume or believe anything, particularly when it is based on as many errors as we've seen presented here.

--Sheldon Rampton 19:18, 27 Nov 2005 (EST)

Is it appropriate to discuss theories about the use of DDT to treat malaria on a page dedicated to John Brignell? Surely the discussion should be limited to his personal views and why he holds them. If you wish to discuss DDT, create a DDT page and put a link to it. I have removed your discussion of the matter.

Your use of the phrase "and his ideological supporters" is surely not appropriate to an unbiased discussion.

"I don't know where the author of this passages gets a figure of 2.8 million" -- from Brignell's website. This is, after all, a web page about John Brignell. Perhaps you should read what he has to say before you attack him for saying it.

"The author of the above paragraph seems also to be under the incorrect impression that malaria control efforts from 1958-63 were limited to Sri Lanka alone, as well as under the incorrect impression that DDT was banned." -- I have read little about this and care less. I was under the impression that DDT had been banned in most countries world wide. I will modify the entry appropriately.

You have complained about my language; I will modify it.

(Jim Smith) --- I'm not sure whether it is the intention of 134*, or just the effect, but all those recent postings make this article virtually unreadable and unrelated to a profile piece on Bignell himself. If people want to publish reams of statistics I'd prefer they did it at Bignell's site - though I wonder if 134* is Bignell himself.--Bob Burton 21:07, 30 Nov 2005 (EST)

John Brignell is an ex-professor who publishes a website about the misuse of statistics and science in the media. His method is to take articles from news sources, examine the underlying science or statistics, and then expose those that are dodgy on his site. He apparently set the site up to promote his book "Sorry Wrong number!". That is a profile of John Brignell.

This page was not set up as a profile on John Brignell. It was set up as an attack on him. It has for some time consisted of a handful of entries in which people draw views from his site and attack them. These are not representative views. No effort is made to examine his reasons for holding them. The thrust of these entries is "He says this, he is wrong, he is a crank." The aim of the page is simply to undermine his credibility.

I have tried to respond on his behalf with no success; my comments were repeatedly discarded. Consequently, I have spent some time doing what his detractors have done, drawing views from his site and discussing them. If the views I have chosen are statistical in nature then that is because he argues statistics, and you need to understand the statistics to understand his arguments, and so I can't avoid it. For example, how can you understand his views on data dredges if you don't understand what one is? If he dismisses a study as a data dredge, and someone attacks his dismissal because it is politically incorrect, and claims that he is a stooge of some organisation, how can you see the claim for the nonsense that it is if you don't understand his reasons for dismissing the study?

I can say that in every case what I have presented is more representative of his viewpoint than anything that was there before. I can also say that it is more appropriate to the topic of the wiki, in that discussion of how people go about fooling the media into publishing nonsense seems more appropriate to the business of PR than discussion of environmental issues like passive smoking or the hole in the ozone layer.

I am not John Brignell. I have never met the man. I don't even believe everything he puts on his site. I am not the stooge of some organisation. I do hold a degree in mathematics. I do actually understand the arguments he makes, which is perhaps more than his detractors do. And I do have an overdeveloped sense of fair play. --Jim Smith