Presentation on theme: "And old and a recent example of scientific fraud Claus Emmeche Center for the Philosophy of Nature and Science Studies Faculty of Science University of."— Presentation transcript:
And old and a recent example of scientific fraud Claus Emmeche Center for the Philosophy of Nature and Science Studies Faculty of Science University of Copenhagen
Outline Some definitions An old example Phrenology Contemporary cases, Stem Cell Research, the case of Woo Suk Whang Evolution of symmetry, the case of Anders Pape Møller Older cases Mendel
Attempt at a taxonomy of fraud Fraud / Scientific misconduct - violation of the standard codes of scholarly conduct and ethical behavior in scientific research. The main forms are: Fabrication - the publication of deliberately false or misleading research, often subdivided in: Fabrication - the actual making up of data and publishing them; reporting experiments which were never conducted. Falsification - manipulation of research data and processes or omitting critical data or results ”Fudging", "massaging", or manufacture of experimental data. Inappropriate, and statistically invalid, "culling" of experimental data, such as the exclusion of experimental runs which contradict the hypothesis the scientist is trying to demonstrate, or excessive filtration of "noise" which suggests a correlation where none can be shown to exist. Intentional portrayal of interdependent events as independent. Plagiarism: the act of taking credit (or attempting to take credit) for the work of another. The appropriation of (1) experimental design or (2) data or (3) experimental material or (4) credit for work that belongs to others; e.g., ”Citation amnesia": failure to credit other/prior discoverers, to give an impression of priority. Maybe the most common type of misconduct. Improper credit can also be inadvertently- assigned, cf. Dulbecco's law = "Credit for a scientific discovery generally goes to the most famous, not the first, discoverer” “An acquaintance of mine once was refused a strain of virus possessed by a prominent investigator, so he cleverly soaked the letter of refusal in an appropriate liquid and recovered enough virus to start his own cultures.” (Lewontin)
another definition The Pedersen report [course material p ] Proposal for rewording Scientific dishonesty is present when a researcher or research group by means of falsification, plagiarisation, non-disclosure or in some similarly gross fashion misleads people about scientific efforts or results of research, and includes: Undeclared construction or fabrication of data or substitution with fictitious data. Selective and surreptitious discarding of undesirable results. Undeclared misleading use of statistical methods. Undeclared misinterpretation of results and distortion of conclusions. Plagiarisation of others’ results or publications. Grossly misrepresented rendering of others’ results. Inappropriate credit as author or misrepresentation of place of work. Applications containing deliberately incorrect information.
“Bad” science, pseudo-science, or just outdated? The case of phrenology: Main thesis (reconstructed) 1.The mind is modular 2.Personal characteristics varies according to the person’s individual module characteristics 3.Brain module strength or performance is correlated with size 4.Brain module size is correlated with cranial surface morphilogy (protusions) 5.Measuring cranial surface thus gives indirect clues to personality A definition of phrenology with chart from Webster's Academic Dictionary, circa 1895
“Bad” science, pseudo-science, or just outdated? The case of phrenology: Main thesis (reconstructed) 1.The mind is modular 2.Personal characteristics varies according to the person’s individual module characteristics 3.Brain module strength or performance is correlated with size 4.Brain module size is correlated with cranial surface morphilogy (protusions) 5.Measuring cranial surface thus gives indirect clues to personality
“Bad” science, pseudo-science, or just outdated? In the Victorian period, phrenology was often taken quite seriously. Many people consulted a phrenologist to get advice in matters like hiring personnel or finding a marriage partner. However, phenology was rejected by mainstream academia; the discipline was excluded from the British Association for the Advancement of Science. The popularity of phrenology varied throughout the 19th century, with some considering the field similar to astrology, chiromancy or merely a fairground attraction, while others published scientific books and journals on the subject. Phrenology was also very popular in the United States, where automatic devices for phrenological analysis were devised.
“Bad” science, pseudo-science, or just outdated? In the early 20th century however, phrenology benefited of a new interest, particularly in the viewpoint of evolutionism on one hand and of criminology and anthropology (as studied by Cesare Lombroso) on the other hand. The most important British phrenologist of this century was the famous London psychiatrist Bernard Hollander ( ). His main works are The Mental Function of the Brain (1901) and Scientific Phrenology (1902). Hollander introduced a quantitative approach to the phrenological diagnosis, defining a methodology for measuring the skull and comparing the measurements with statistical averages.
“Bad” science, pseudo-science, or just outdated? Phrenology as Protoscience (?) Many of the basic premises of phrenology have been confirmed by medical science. Phrenologists were the first to claim that all mental activity, including emotion, takes place in the brain. Furthermore, modern methods such as fMRI demonstrate that different types of activities take place in different regions of the brain. Recent studies also support the related claim that neural centers can become enlarged with use (J. Neuroscience, 17, 1997). Thus, many of the claims of phrenology have been absorbed into modern scientific thinking. Most of the practical implications of phrenology are now generally considered false. In particular, the ability to judge character based on skull shape has no modern scientific basis. The basic concept of a mental "organ" developed into the modern theory of functional neural centers, i.e., regions of the brain which perform a particular mental task. However, fMRI studies demonstrate that the actual function (or faculty) of the 27 organs described by phrenologists is incorrect; the only exception is Gall's "language faculty", which is somewhat close to Broca's area, a brain region associated with language processing.
Evolution of symmetry: the case of APM Anders Pape Møller A case of ”mode 1” fraud.
Stem cell research: the case of Whang Woo-Suk Hwang (born 1953), South Korean biomedical scientist. A case of ”mode 2” fraud. Woo-Suk Hwang He was a professor of biotechnology at Seoul National University (dismissed on March 20, 2006) who rose to fame after claiming a series of remarkable breakthroughs in the field of stem cell research. Until November 2005, he was considered one of the pioneering experts in the field of stem cell research, best known for two articles published in journal Science in 2004 and 2005 where he (fraudulently) reported to have succeeded in creating human embryonic stem cell lines by cloning. Both papers have been editorially retracted after being found to contain a large amount of fabricated data. He has admitted to various lies and frauds.
Stem cell research: the case of Whang In February 2004, Hwang and his team announced that they had successfully created an embryonic stem cell with the somatic cell nuclear transfer method. Until Hwang's claim, it was generally agreed that creating a human stem cell by cloning was next to impossible due to the complexity of primates. Hwang explained that his team used 242 eggs to create a single cell line. Hwang WS, et al. (2004). "Evidence of a pluripotent human embryonic stem cell line derived from a cloned blastocyst". Science 303 (5664): Hwang's team announced an even greater achievement a year later in May 2005, and claimed they had created 11 human embryonic stem cells using 185 eggs. Hwang WS, et al. (2005). "Patient-specific embryonic stem cells derived from human SCNT blastocysts". Science 308 (5729): This was instantly hailed as a breakthrough in biotechnology: the cells were allegedly created with somatic cells from patients of different age and gender (the stem cell of 2004 was created with cells from a single female donor). This meant every patient could receive custom-made treatment with no immune reactions; his team had boosted their success rate by 14 times; and this technology could be medically viable.
Stem cell research: the case of Whang Originally published in Science Express on 19 May 2005Science 17 June 2005:Vol no. 5729, pp DOI: /science This article has been retracted Patient-Specific Embryonic Stem Cells Derived from Human SCNT Blastocysts Woo Suk Hwang, 1,2 * [et al] Patient-specific, immune-matched human embryonic stem cells (hESCs) are anticipated to be of great biomedical importance for studies of disease and development and to advance clinical deliberations regarding stem cell transplantation. Eleven hESC lines were established by somatic cell nuclear transfer (SCNT) of skin cells from patients with disease or injury into donated oocytes. These lines, nuclear transfer (NT)–hESCs, grown on human feeders from the same NT donor or from genetically unrelated individuals, were established at high rates, regardless of NT donor sex or age. NT-hESCs were pluripotent, chromosomally normal, and matched the NT patient's DNA. The major histocompatibility complex identity of each NT-hESC when compared to the patient's own showed immunological compatibility, which is important for eventual transplantation. With the generation of these NT-hESCs, evaluations of genetic and epigenetic stability can be made. Additional work remains to be done regarding the development of reliable directed differentiation and the elimination of remaining animal components. Before clinical use of these cells can occur, preclinical evidence is required to prove that transplantation of differentiated NT-hESCs can be safe, effective, and tolerated.
Stem cell research: the case of Whang Official probe by Seoul National University and the confirmation of fraud An internal panel was set up in Seoul National University to investigate the allegation. On December 23, 2005, the panel announced its initial finding that Hwang had intentionally fabricated stem cell research results creating nine fake cell lines out of eleven, and added that the validity of two remaining cell lines is yet to be confirmed. The panel stated that Hwang's misconduct is "a grave act damaging the foundation of science.” In its final report of January 10, 2006, the panel found out that, contrary to Hwang's claim of having used 185 eggs for his team's 2005 paper, at least 273 eggs were shown to have been used. In addition, the panel discovered that Hwang's team was supplied with 2,061 eggs in the period of 2002 to Hwang's claim of not having known about the donation of eggs by his own female researchers was also denied by the panel; in fact, it was discovered that Hwang himself had distributed egg donation consent forms to his researchers and personally escorted one to the MizMedi Hospital to take the egg extraction procedure. The panel stated that Hwang's 2004 Science paper was also fabricated [...] Although Hwang's team didn't rule out the possibility of parthenogenetic process in the paper, the panel said, his team didn't make any conscientious effort to probe the possibility through the tests available.
the case of Whang - moral reactions… [from the blog ] Although I have no vested interest in seeing Hwang severely punished, there is a case to be made that there should be some, and that it should be swift and severe. [...] No, the case is a pragmatic one, one that might help us all avoid an even greater evil. As I wrote last month in Roe v. Hwang : Roe v. Hwang "My (prediction) second is that Hwang will be highly sought-after as a private researcher, i.e. doing research in this area for interests who have no interest in peer-reviewed publication. I hope my second prediction is wrong.” What I was hinting at there is so much money to be made and so many unscrupulous being willing to fund research that could earn a part of it, that if Hwang is not dealt with harshly, and soon, he could very well find himself in the employ of just such people. This would be a tragedy. Without peer review (which in this case was duped) and professional ethics (which is this case were violated) and the lure of scientific acceptance (which is something Hwang will now never, ever be able to achieve), there would be no constraints on the development of technologies and procedures and life forms that, if seen in the light of day, would horrify the vast majority of the civilized world even though many of the same might eventually be brought around to accepting the results, i.e. "clone my dead dog or my lost child, just don't tell me how you did it.) There would be a situation whereby Hwang and others would be able to strive only for ends while thinking nothing of means- the exact same state of affairs that, by his own admission, got him into all this trouble in the first place.
then, a brief discussion experiment … What went wrong? Hwang started out (presumable) as a good scientist doing good science. Mention three personal characteristics that you think is important for becoming a good researcher e.g.: Creative Persistence (stædighed) Coorperative (samarbejdsevne) or: Intelligent [multiple!] Organization talent Diligent (flittig) or: Curious Keen (skarpsindig) Focused/nerdish (målrettet)
“He did good science”. But what is quality? ”research” here is academic science (cf. Ziman) Good vs. “Bad”: refer primarily to quality Method? Can it be learned? Is it something individual or social? What is “bad”? Pseudo-science & Fraud OBS: Bad ≠ “Not-elite”; “Bad” = Mediocre? Does evaluation of quality have to be coupled to funding? Who decides? (& govern) Peers Funding agencies Students Where does quality assessments take place?
Are ’good’ vs. ’bad’ and ‘honest’ vs. ‘dishonest’ two orthogonal dimensions? Parameter Quality: Honesty: Good scienceMediocre science science fraud
Are ’good’ vs. ’bad’ and ‘honest’ vs. ‘dishonest’ two orthogonal dimensions? Parameter Quality: Honesty: Good scienceMediocre science science5 % Science95 % Science fraud Mendel ? Pape Møller? Woo Suk Whang ? [ Lomborg? ] What extra dimensions are missing?
Good quality — of what? social system What is quality in research? Quality of what? — Try a comparative sociological approach: Each social system has its own basic parameter of “relevant quality” (cf. N. Luhmann). As an institution with a Symbolically Generalized Medium (SGM), Economy relates to the profitable/non-profitable or +/- payment distinction (SGM = money) Politics relates to the +/- power distinction (SGM = power) Religion relates to the (existential) +/- faith & immanence/transcendence distinctions (SGM = faith) The judicial system relates to the legal/unlegal & just/unjust distinctions (SGM = law) Ethics relates to the +/- respect distinction (SGM = morality) Love or Intimate relations relate to the +/- loving distinction (SGM = love) Art relates to the beautiful/ugly & +/- sublimity distinctions (SGM = art) Science relates to (some version of) the truth/false distinction (SGM = truth)
Good quality — of what? True knowledge! At the basis of any Symbolically Generalized Medium (SGM) lies a binary code that is partly a division of the contingent and partly a division of an area-delimiting applicability defining what can be articulated and what cannot within that social system. Economy relates to the profitable/non-profitable or +/- payment distinction (SGM = money) Politics relates to the +/- power distinction (SGM = power) Religion relates to the (existential) +/- faith & immanence/transcendence distinctions (SGM = faith) The judicial system relates to the legal/unlegal & just/unjust distinctions (SGM = law) Ethics relates to the +/- respect distinction (SGM = morality) Love or Intimate relations relate to the +/- loving distinction (SGM = love) Art relates to the beautiful/ugly & +/- sublimity distinctions (SGM = art) Science relates to (some version of) the truth/false distinction (SGM = truth)
Good quality — of what? True knowledge! Science relates to (some version of) the truth/false distinction In what sense is science about the general code of truth/false? Much philosophy of science - especially at first half of the 20th Century - has tried to answer this question in an abstract, a priori manner, while contemporary commentators more look at the social and historical dimensions of “truth”.
The qualitative qualities of good research 1.Erfaringsbasering (empiri, iagttagelse, måling, virkelighedsforankring) 2.Offentlighed (gennemsigtighed, reproducerbarhed, publicerbarhed). 3.Systematicitet (indre systematik og sammenhængsmulighed med anden viden). 4.Indsigtsgivning (forståelse, teorirelatering, generalitet, også ved enkelthændelser). 5.Forklaringsmulighed (fx mht. historie, struktur, dynamik, eller mening). 6.Intern og ekstern konsistens (argumentatorisk og mht. anden viden). 7.Reviderbarhed, åbenhed om mulige fejl, kritikparathed, uafhængighed
Some general qualities of good research 1.Experience-based (empirical observations, measurements, reality-rootedness) 2.Public access (transparancy, reproducibility, ‘publishability’). 3.Systematicity (inner systematics as well as connectedness with other knowledge). 4.Insight-bringing (understanding, teory-relatedness, generality, also for case studies of single events). 5.Explanation-offering (e.g., regarding history, structure, dynamics, or meaning). 6.Internal and external consistency (regarding arguments and other knowledge). 7.Revisability, opneness regarding possible errors, readiness for critique, independence.
Towards an anatomy of criteria for ‘good’ science Good vs. Mediocre/Bad General Internal criteria: Focus Scope Parsimony, Consistency Originality Precision Problem solving capacity Explanatory force Testability Validity & Reliability ‘Exemplary potential’ & fruitfulness for further research Communicability for dissemination Specific Internal criteria Numeric Precision Simplicity Reproducibility … etc. External criteria Relevance, usability Sustainability Ethical acceptability Honest vs. Dishonest Thrustworthyness / Authenticity / reliability
Towards an anatomy of dimensions of ‘good’ science Good vs. Mediocre/Bad General Internal criteria: Focus Scope Parsimony, Consistency Originality Precision Problem solving capacity Explanatory force Testability Validity & Reliability ‘Exemplary potential’ & Fruitfulness for further research Communicability for dissemination Specific Internal criteria Numeric Precision Simplicity Reproducibility … etc. External criteria Relevance, usability Sustainability Ethical acceptability Honest vs. Dishonest Thrustworthyness / Authenticity / reliability * * * Criteria for evaluators? Rules for scientists? Values in science?
conclusion The secret of life is honesty and fair dealing. If you can fake that, you've got it made. Groucho Marx
Questions for discussion: Discuss if dishonesty and fraud are always clearcut or if there are borderline cases. Compare the Pedersen report p. 10 § 3.1: ”the working group does not find that the concept of dishonesty can be graduated”. Discuss under what conceivable circumstances your own or colleagues’ research projects might be accused of fraught. How to get rid of fraud? Could specific changes in the normal scientific process (including training, experimenting, evaluation of scientific claims, testing, publishing) help bringing down the number of cases of dishonesty?