Home » Science » Significance is not what it used to be …

Significance is not what it used to be …

Interesting article in PLoS Medicine (source):

Why Most Published Research Findings Are False
By John P. A. Ioannidis

Abstract: There is increasing concern that most current published research findings are false. The probability that a research claim is true may depend on study power and bias, the number of other studies on the same question, and, importantly, the ratio of true to no relationships among the relationships probed in each scientific field. In this framework, a research finding is less likely to be true when the studies conducted in a field are smaller; when effect sizes are smaller; when there is a greater number and lesser preselection of tested relationships; where there is greater flexibility in designs, definitions, outcomes, and analytical modes; when there is greater financial and other interest and prejudice; and when more teams are involved in a scientific field in chase of statistical significance. Simulations show that for most study designs and settings, it is more likely for a research claim to be false than true. Moreover, for many current scientific fields, claimed research findings may often be simply accurate measures of the prevailing bias. In this essay, I discuss the implications of these problems for the conduct and interpretation of research.

  • Delicious
  • Facebook
  • Reddit
  • StumbleUpon
  • Twitter
  • RSS Feed

7 Responses to Significance is not what it used to be …

  1. Why Most Published Research Findings Are False (Or severely misleading!)

    The ‘literature dump’ at Dover is a prime example of the severe bias inherent in many avenues of peer-review

    Five Years Later, Evolutionary Immunology and other Icons of Kitzmiller v. Dover Not Holding Up Well
    Excerpt: For those who don’t recall, during the Kitzmiller trial the plaintiffs’ attorneys performed a literature dump on Michael Behe which was claimed to show Darwinian explanations for the origin of the immune system.

    This episode of courtroom theatrics impressed Judge Jones so much that he wrote that “Between 1996 and 2002, various studies confirmed each element of the evolutionary hypothesis explaining the origin of the immune system. ” (Actually, it’s arguable that Judge Jones didn’t originally write those words–he essentially copied them from an ACLU brief which stated: “between 1996 and 2005, various studies confirmed each element of the evolutionary hypothesis explaining the origin of the immune system.”)

    Don Ewert has read the papers from the original literature dump, as well as more recent papers on the evolution of the immune system, and he finds that Behe was correct to argue that these papers don’t provide a Darwinian explanation for the origin of the immune system.
    http://www.evolutionnews.org/2.....42001.html

    “A Masterful Feat of Courtroom Deception”: Immunologist Donald Ewert on Dover Trial – audio
    http://intelligentdesign.podom.....1_03-08_00

    Generation of Antibody Diversity is Unlike Darwinian Evolution – microbiologist Don Ewert – November 2010
    Excerpt: The evidence from decades of research reveals a complex network of highly regulated processes of gene expression that leave very little to chance, but permit the generation of receptor diversity without damaging the function of the immunoglobulin protein or doing damage to other sites in the genome.
    http://www.evolutionnews.org/2.....40661.html

  2. Odds are…John P. A. Ioannidis’ article is false too.

  3. grizzfan:

    Can you prove beyond doubt that Ioannidis’ article is false???

  4. Welcome to 2005.

  5. 5

    In geomorphology papers dealing with numerous studies I often find they contradict each others confident conclusions as to make it almost useless to place any confidence in any researcher.
    They come to conclusions, reluctant to drop them, that are opposed by other researchers. I noticed this in the ideas on the creation of step-pools in rivers.
    I can easily imagine most ideas are just not true. yet it requires other people to show this.
    This is why in evolution stuff over the last century there is slow recognition of its flaws.
    Few people do or can apply methods of investigation that test evolutionism.

  6. I’m a little surprised the peepers haven’t made more of a chorus over this warming trand. This is, in one sense, the biggest story of all, since the Age of (Big) Science imposed its dry hegemony through the incantation “studies demonstrate that…”

    So, like, if Ioannidis is onto to something, the studies demonstrate…nothing.

    Very liberating, if you ask me.

  7. As to the lack of integrity for the peer-review of evolutionary papers, In this podcast which was just recorded recently, Casey Luskin interviews microbiologist and immunologist Donald Ewert about his previous work as associate editor for the journal Development and Comparitive Immunology, where he realized that the papers published were comparative studies that had nothing to do with evolution at all.
    http://intelligentdesign.podom.....9_03-07_00

Leave a Reply