Some celebrated studies did not hold up later, from fabricated data to over-claimed results, and a failure is not a reason to discard careful methods.
The ten cases below are Andrew Wakefield's 1998 MMR paper, the 1989 Fleischmann and Pons cold-fusion announcement, Diederik Stapel's fabricated social-psychology papers, Dean Hamer's 1993 linkage claim often nicknamed a "gay gene," the 2010 power-pose paper, the older idea of repression as a reliable vault for traumatic memory, William Moulton Marston's physiological theory of lying, the 1996 elderly-priming experiment, Jan Hendrik Schön's retracted physics papers, and the 2014 STAP-cell claims. A later unethical gene-editing episode is sometimes filed with failed studies; this list stays with work that was offered as a scientific result and then did not hold up. The tone is an account of error and fraud, not an argument against science. This article will not give medical advice or explain how a fraud was done.
Key Facts
| Case | What did not hold up | A limit |
|---|---|---|
| 1 | The 1998 Lancet paper claiming a link between the MMR vaccine and autism was retracted after ethical and data failures | Later large studies have not supported that link; this is not personal medical advice |
| 2 | Room-temperature "cold fusion" as announced in 1989 was not reproduced as a reliable, widely accepted result | The episode does not make all energy research a hoax |
| 3 | Stapel fabricated data for many papers, which were later retracted | A person's fraud is not a verdict on an entire field |
| 4 | A single "gay gene" on the X chromosome was an over-claim; later work points to a complex, many-factor picture | Failed replication of one linkage is not a license for stigma or for medical advice |
| 5 | Strong claims that a brief "power pose" reliably changed hormones and risk-taking did not survive later tests | A feeling of confidence after a pose is a smaller, separate question |
| 6 | Repression as a special mechanism that stores trauma intact and returns it reliably later is not well supported | Trauma is real; the dispute is about a particular theory of memory, and this article will not retry private cases |
| 7 | Lying does not leave a unique, dependable stress signature of the kind early polygraph promoters claimed | Fear and anxiety can move the same needles; this is not a guide to evading a test |
| 8 | A famous report that elderly-related words made people walk away more slowly has not proved robust in later replications | Priming research continues; this one headline result was a weak pillar |
| 9 | Schön's spectacular results in molecular electronics and related physics were found to be fabricated and were retracted | The cleanup was a community act, not a reason to discard condensed-matter research |
| 10 | Stimulus-triggered acquisition of pluripotency, announced in 2014, did not hold up and the papers were retracted | Stem-cell biology has other, better-documented methods; this article will not offer a protocol |
1. A Retracted Vaccine Paper, and What It Was Not
In 1998 The Lancet published a paper by Andrew Wakefield and co-authors that suggested a link between the measles, mumps, and rubella vaccine and a novel syndrome including autism [1]. The paper was small, and it was later gutted by investigations that found ethical breaches and data problems. The journal retracted it. Wakefield was later struck off the medical register in the United Kingdom. The scare, however, traveled farther than the retraction, and it still lives as folklore. Subsequent large epidemiological studies have not supported an MMR-autism link of the kind the paper implied. This article will not tell any reader what to do at a clinic. It will say what the scientific and editorial record did with that paper [1].
The unexpected fact is that a short, already fragile article can outlive its own retraction as a public rumor [1]. The careful fact is a documented publication, a documented retraction, and a documented later body of evidence that did not confirm the claimed link. Fear was the afterlife. Method was the autopsy.
A retracted paper can still do social work after it has stopped doing scientific work.
2. A 1989 Energy Claim That Could Not Be Stabilized
In 1989 Martin Fleischmann and Stanley Pons announced that they had produced nuclear fusion in a room-temperature electrochemical cell, a result soon labeled "cold fusion" [2]. The promise was extraordinary: abundant energy from a bench apparatus. Laboratories around the world tried to repeat the work. The claimed excess heat and nuclear signatures were not established as a reliable, independently reproduced effect in the form that had been announced, and the scientific mainstream did not accept the result. Some researchers continued to explore related electrochemical ideas under other names. Those later programs have not, so far, produced a generally accepted, practical fusion method of the 1989 kind. A failed headline is not the same as a secret conspiracy against a working device [2].
The unexpected fact is that a press conference can move faster than a replication [2]. This article will not offer a cell design or an energy tip. The careful fact is a documented announcement, documented failed or inconsistent replications at the time, and a documented collapse of the original claim as textbook knowledge. Excitement was understandable. Agreement never arrived.
A clean-energy hope can be sincere and still not be, after the checks, a working discovery.
Related: Top 10 Why Introverts Are Attractive
3. Fabricated Data, Then a Stack of Retractions
Diederik Stapel was a prominent Dutch social psychologist whose papers offered tidy findings about how environments shape judgment and bias [3]. In 2011 an investigation concluded that he had fabricated data on a large scale, inventing results rather than collecting them. Journals retracted many of the papers. Students and co-authors who had trusted the data were among the people harmed. The case became a landmark in discussions of research integrity in psychology. It does not license a sneer at every social-psychology experiment. It does show what a community has to do when a trusted author stops being a source [3].
The unexpected fact is that a career can be built, for a time, on numbers that never came from a participant [3]. This article will not give a method for inventing a data file. The careful fact is a documented investigation, documented retractions, and a documented professional fall. Fraud is a person. The field is larger than the person.
A retraction list can be long and still be, correctly, an act of repair rather than a funeral for a discipline.
4. A Linkage Claim, Not a Single Gene for a Life
In 1993 Dean Hamer and colleagues reported a linkage between male sexual orientation and a region on the X chromosome, Xq28, a finding that newspapers quickly shortened to a "gay gene" [4]. The study was real work, not a hoax, and it asked a legitimate genetic question. Later attempts did not establish a single locus of that kind as the explanation, and more recent genomic studies point to a many-variant, small-effect picture mixed with non-genetic factors, not a switch. "Discredited" is too blunt if it sounds like fraud. "Over-claimed in public, and not confirmed as a single gene," is closer. This article will not offer medical advice, a test, or a theory of worth. Sexual orientation is not a laboratory prize [4].
The unexpected fact is that a careful-sounding chromosomal region can become, in headlines, a destiny [4]. The careful fact is a documented 1993 linkage report, documented trouble with a simple replication of that story, and documented later complexity. Biology can be involved. A single gene was the part that did not hold as a popular claim.
A study can open a conversation and still not have found a switch.
5. Power Poses, and the Claims That Shrank
A 2010 paper by Dana Carney, Amy Cuddy, and Andy Yap reported that brief "high-power" poses changed hormone levels and risk-taking in a small sample [5]. The idea traveled far, especially through a popular talk and a simple physical recipe. Later replications and reviews did not support the strong physiological and behavioral claims in the original form. One of the original authors later publicly distanced herself from the findings. Whether standing in a certain way can change a person's mood is a narrower question, and it is not the same as a dependable hormonal switch. The celebrated version did not hold up [5].
The unexpected fact is that a two-minute pose can become a global tip while the hormone numbers fail later checks [5]. This article will not offer a pose as advice. The careful fact is a documented original report, documented failed replications of the strong claims, and a documented shrinking of the story. Confidence is a feeling. Cortisol is a measurement, and it was the measurement that would not stay still.
Related: Top 10 Times Scientists Were Sure and Then Wrong
A popular talk can outrun a small sample, and the sample is what later work has to re-do.
6. Repression Was Not a Vault, and Loftus Was Not Its Author
A theory of repression, with roots in older psychoanalytic writing, held that the mind can lock traumatic memories away and that they can later return intact, sometimes in therapy [6]. In late-twentieth-century practice, some recovered-memory methods treated that idea as a clinical engine. What did not hold up is the special mechanism: memory is reconstructive, suggestion can create confident false memories, and a buried, photographically accurate record is not a well-supported general fact. Elizabeth Loftus, often miscast in popular summaries as the person who "popularized repression" in the 1990s, was instead a leading critic of naive recovery claims. Trauma is real. The dispute is about a model of memory, not about whether harm happened. This article will not retry private legal cases or offer therapy advice [6].
The unexpected fact is that a famous critic of a theory can be remembered, wrongly, as its salesperson [6]. The careful fact is a documented older theory, documented later criticism, and documented experimental work on suggestion. Compassion is required. A vault in the mind is the part that science would not underwrite as a general law.
A memory can be painful and still not have been stored as a sealed file.
7. Lying Is Not a Unique Stress Signature
William Moulton Marston, early in the twentieth century, promoted the idea that deception produces a distinctive, measurable pattern of blood pressure and related changes, work that fed the later culture of the polygraph [7]. Police and other agencies used the machines for decades. Later reviews have not found a unique, dependable physiological signature of lying that can be separated, in a robust way, from fear, anger, or the stress of being tested. That is why many courts limit polygraph evidence, and why a "failed test" is not, by itself, a scientific proof. This article will not explain how to beat a machine or how to run one. It will say that the founding claim did not hold as a clean law of the body [7].
The unexpected fact is that a device can remain in use after its original theory has been narrowed almost to a habit [7]. The careful fact is a documented early claim, documented later skepticism, and documented legal caution. Stress is real. Its meaning is the part that would not stay unique.
A needle can move for many reasons, and "because the person lied" is only one of the stories that has been attached to the movement.
Related: Top 10 Most Bizarre Phobias: Strange Fears That Exist
8. A Famous Priming Effect That Would Not Stay Still
In 1996 John Bargh and colleagues reported that people exposed to words associated with older age later walked away from the experiment more slowly, a flagship example of "social priming" [8]. The study was widely cited and widely taught. Later replications, including well-powered attempts, did not reliably recover the walking-speed effect, and the episode became a symbol of psychology's replication crisis. Priming as a broad family of ideas is not therefore dead; this particular dramatic result was a weak pillar for the larger claim that a few words can reset a gait on demand. Method, sample size, and publication bias all became part of the later discussion [8].
The unexpected fact is that a short walk down a hallway can stand, for years, as a theory of the mind [8]. This article will not treat a failed replication as a prank. The careful fact is a documented original report, documented later non-replications, and a documented change in how the field talks about such effects. Subtle influence is still studied. The tidy hallway result was the part that did not hold up cleanly.
A citation count can rise while the walking times, in other laboratories, refuse to cooperate.
9. Spectacular Physics, Retracted
In the early 2000s Jan Hendrik Schön, working at Bell Labs, published a run of spectacular papers on organic electronics and related condensed-matter claims [9]. The results promised devices and states of matter that many groups wanted to see. Other laboratories could not reproduce them. An investigation found that data had been fabricated or duplicated in misleading ways. Journals retracted the papers, and Schön's career in that form ended. The case is now a standard warning in research ethics. It is not a reason to throw away the wider fields of organic semiconductors or molecular electronics, which continue with other people's data [9].
The unexpected fact is that a famous laboratory's letterhead can delay, but not prevent, a forensic reading of a figure [9]. This article will not reconstruct the fabrication. The careful fact is a documented investigation, documented retractions, and documented failed replications that came first as a smell. Ambition was visible. The spectra were not.
A breakthrough can look like a golden decade in print and still be, after the inquiry, a stack of withdrawn papers.
Related: Top 10 Oldest Buildings in the World
10. A 2014 Stem-Cell Claim That Did Not Hold Up
In 2014 Nature published papers from a group led by Haruko Obokata, at Japan's RIKEN institute, claiming that a simple stress could reprogram ordinary cells into a pluripotent state called STAP, or stimulus-triggered acquisition of pluripotency [10]. The promise was enormous: an easy route to powerful cells without the usual reprogramming tools. Other groups could not repeat the result. Investigations found serious problems in the work, the papers were retracted, and the claim did not enter the toolkit of stem-cell biology. Better-documented methods for reprogramming exist; they are not this one. This article will not describe a protocol, and it will not treat the episode as a reason to dismiss the rest of the field [10].
The unexpected fact is that a simple-sounding recipe can pass, briefly, the gate of a leading journal and still die in other people's hands [10]. The careful fact is a documented publication, a documented retraction, and documented failures to reproduce the state as claimed. Hope was the fuel. Replication was the test.
A cell can be famous for a season and still not be, when the methods are checked, a new kind of life in a dish.
Conclusion
The ten cases above include a retracted vaccine paper, an energy announcement that would not reproduce, a fabricated social-psychology career, an over-claimed gene story, a pose whose hormone claims shrank, a memory theory that was not a vault, a lie detector's founding signature, a priming effect that later walks would not confirm, a stack of withdrawn physics papers, and a stem-cell shortcut that was retracted. Error, over-claim, and fraud are different diseases. They are all reasons to keep the checks.
A study can fail and still leave the method, if the method is honest, in better shape than it was before the failure was named.
- ^ [1] Accounts of the 1998 Lancet MMR paper, its retraction, the later professional findings against Andrew Wakefield, and subsequent epidemiological work that has not supported the claimed vaccine-autism link, presented without personal medical advice. https://www.britannica.com/science/MMR-vaccine retraction-related materials
- ^ [2] Accounts of the 1989 Fleischmann and Pons announcement and of the subsequent failure to establish cold fusion as a reliable, independently reproduced effect. https://www.britannica.com/science/cold-fusion https://www.britannica.com/science/nuclear-fusion
- ^ [3] Reports of the 2011 investigations into Diederik Stapel, the fabrication of data, and the later retractions, treated as a research-integrity case rather than a verdict on all social psychology. https://www.nature.com/articles/479015a related research-integrity reporting
- ^ [4] Accounts of the 1993 Hamer linkage report on Xq28 and of later genomic work that does not support a single "gay gene," written without medical advice, testing tips, or claims about anyone's worth. https://www.britannica.com/science/human-genetics related later genomic discussion
- ^ [5] Accounts of the 2010 power-pose paper and of later replications and reviews that did not support the original strong hormonal and risk-taking claims. Association for Psychological Science materials https://www.britannica.com/science/social-psychology
- ^ [6] Historical and scientific accounts of repression and recovered-memory claims, of later work on the reconstructive character of memory, and of Elizabeth Loftus's role as a critic rather than as the theory's 1990s popularizer, without retrying private cases or offering therapy advice. https://www.britannica.com/science/memory-psychology https://www.britannica.com/biography/Elizabeth-F-Loftus
- ^ [7] Historical accounts of William Moulton Marston's physiological theories of deception and of later scientific and legal skepticism about the polygraph as a unique lie detector, without instructions for running or evading a test. https://www.britannica.com/topic/polygraph https://www.britannica.com/biography/William-Moulton-Marston
- ^ [8] Accounts of the 1996 Bargh elderly-priming report and of later replication attempts that did not reliably recover the walking-speed effect, placed in the wider discussion of psychology's replication crisis. https://www.britannica.com/science/social-psychology https://www.psychologicalscience.org/
- ^ [9] Accounts of the investigation into Jan Hendrik Schön's work at Bell Labs, the retractions, and the failure of other groups to reproduce the spectacular results. https://www.nature.com/articles/419419a https://www.britannica.com/topic/Bell-Labs
- ^ [10] Accounts of the 2014 STAP-cell papers, the failure of independent groups to reproduce the claim, and the subsequent retractions, without any reprogramming protocol. https://www.nature.com/articles/nature.2014.15393 https://www.britannica.com/science/stem-cell
Discussion
Log in with