Most of the research is wrong. It has been for decades often wrong.
Now someone proves it is wrong it is pulled. That is not a surprise to me. I was raised hearing just how the research shifted the results every other year.
You can categorize that however you like. Doesn’t change how wrong research often is.
The minor stuff is in that mix according to my BIL not me.
What am I supposed to care if research is wrong? As I said I expect research to be wrong.
My BIL is mostly just finding molecules and reporting they are in cells. He is not drawing conclusions. That is where a lot of researchers go wrong.
The conclusions remain unknown even when stated. Just grabbing at the first thing that seems to add up does not work well with the human body at times.
Certain pop-media science figures (and therefore influential) whenever they’re on TV seem to be the main demographic. Doctors “offering” a new drug are also in that Venn diagram a lot.
The examples I cited way back at the beginning of this thread had nothing to do with being WRONG, they were related to academic papers submitted to scientific publications (print or online) that were knowingly FRAUDULENT.
When hundreds of scientists are researching a topic attempting to explain a biological phenomena or devise some new material to increase the efficiency of batteries, etc., those scientists are ALWAYS going to get things WRONG. Their formulation of the phenomena they are trying to reproduce or create might be wrong, causing the structure of their experiment or study to be wrong and thus the results incorrect. Their understanding of the phenomena they are attempting to recreate or synthesize might be wrong so the instrumentation of their experiment might eventually prove to be flawed, causing their conclusions to be WRONG.
Devoting yourself to a scientific career essentially means devoting yourself to a career of being mostly wrong, interspersed with precious few occasions in which you are right and “stay right” for years / decades. Real scientists know that and accept it as part of the profession.
These people were perpetrating FRAUDS. They were using statistics to MANUFACTURE data they claimed they COLLECTED from experiments that “proved” amazing leaps forward in their area of research. In some cases, they were picking random microscopic scale photos of cells having NOTHING TO DO with their research and using those photos as “illustrations” of the processes and results they were studying. It’s just a bunch of random cells, what the hell do the rest of us non-science rubes looking at a story onine know? We’ll never know what the scientist saw looking into that electron microscope. Looks “sciencey” to me. I guess the results must be true.
That’s the reaction these fraudsters are counting on to provide their ego, their resume or their funding a temporary boost. Unfortunately for the people perpetrating these frauds, they did not anticipate how search technologies would simpify finding the fraud techniques they were using. Apparently, some still haven’t caught on because new frauds are being detected every week in new publications.
I love this - mind if I use this myself when I need to? And not only that; but you also have hundreds of other scientists spending their lives trying to prove you wrong!
I’d say a large part of it is the increased circulation of papers and the audience interested in researching/debunking them. Back in the day most would be seen by what, a dozen people? A couple dozen if the paper went into the frat house file cabinet and came back again a few years later?
Now they can be read by hundreds, thousands, and they have become fodder for discrediting people who, 30 or 40 career years later, are in a position where somebody wants to take them down.
I did a few things in college that would be fodder for enemies now, if I were in a position that anybody cared about.
One of the questions I’ve always asked myself (for as long as I’ve been following and highlighting fraud and/or misrepresentation in science, that is…Old Lags from the H&N board will know what I’m talking about) is “WHY?”. Obviously key motivators are things like personal gain, attention grabbing to attract grant funding, resume padding (so many labs, so many PhD candidates chasing “publications”) but my “why” revolves more around the inevitable exposure that should be expected when headline grabbing new discoveries cannot be reproduced by independent researchers. Thing is, though, given the apparent frequency that fraud or intentional manipulation of data appears to be perpetrated with subsequent retraction it’s pretty apparent that the Scientific Method…as it’s understood by folk who do understand science…simply isn’t powered to detect this. At least as far as the biomedical sciences are concerned. I’m sure it’s potentially the same with the “harder” sciences like physics and whatnot but that’s not where I focus my attention.
As you mention, though, fraudulent science is immeasurably different from wrong science (faulty hypothesis to start with, say) bad science (faulty/sloppy methodology)…or even dysfunctional science (where someone is so convinced they’re right that they massage statistics just a bit).
An example of fraudulent with big impact is Andrew Wakefield’s alleged discovery of a link between MMR vaccine and an inflammatory bowel condition (discovered by him, supposedly) leading to autism. The 11th anniversary of the paper’s retraction was this last Groundhog Day…ironically enough. Oddly, the Scientific Method did work in this instance in that it was potentially such a consequential finding…even as a case series report…and published in a high impact journal that folk immediately attempted to reproduce his findings and failed. What led to Wakefield’s ultimate demise was his outright refusal to a) make any attempt to reproduce his ownndata and b) keep his trap shut and accept the ignominy of irreproducible results and took to the popular press.
Yes, he held a patent on separate formulations of individual vaccines (personal gain) was being bankrolled by a law firm seeking to file the equivalent of a class action law suit (personal gain again)…but he knew that the individual elements of his fraud were being perpetrated in a way that made his results irreproducible almost before he even published. I cannot fathom the gall.
This particular example captured my attention back in the early aughts because this “research” was performed in the very department where my husband was looking for that origin of circulating serum trypsin in man mentioned upstream some 15 years or so earlier. Neither of us can answer the “why” question to this day.
P.S…and the actual fraud was highlighted in almost a serendipitous way. An investigative journalist was looking into the love affair the press was having with Wakefield when his results didn’t appear to warrant any attention (and I believe he’d been “invited to resign” from the Royal Free Hospital by this time) Wakefield made a big error in trying to use the UK’s quirky libel laws to shut him up…which gave Brian Deer the opportunity to access the full medical records of the children in the research study.
Here’s a little background on Sholto David, the guy who highlighted this fraud…
…and from the early (but not earliest) days of the Fakefield debacle, another Grauniad article from one of my fave medical writers.
Ben Goldacre is the writer who first put the Wakefield scandal on my radar screen back in the early aughts…along with other equally bad, but not quite so far reaching, examples of Science By Press Release in his Bad Science column in the Guardian. He was fairly fresh out of his specialty training in psychiatry at the time and used his column as a sounding board to try to combat the increasing amount of misinformation reaching the media outlets (and subsequently the general public) from glorified bird cage liners such as the Daily Mail to seemingly more trustworthy rags like the NYT. He also wrote a book (one of many on critical thinking)
My guess, and it is only that, is hubris. I’ve had numerous experiences with researchers who based on what I would consider only preliminary evidence are convinced that their model for how something is working is correct.
If one truly believes one has the right answer it is easy to see the motivation to cut corners and publish first. This is particularly the case if one believes it is only technical difficulties or an incompetent grad student that is delaying the production of the conclusive piece of evidence. Since the answer is right then attempts by other labs to replicate the “results” will in fact do so.
Suppose you are in a highly competitive field with multiple labs doing similar experiments. You come up with a model that beautifully explains your past data and you are convinced of its correctness, but one more experiment is needed to conclusively nail it down. Unfortunately that experiment will take a year and cost $50,000. You can do things the right way and risk being scooped by the lab at Cal Tech or Stanford. Alternatively, there is Photoshop…
Well, if this is your understanding of the Scientific Method, you should question whether you were raised within a background of scientific thinking.
In the context of new research/novel ideas, a new hypothesis (not theory BTW but hypothesis an important difference in science) always has the potential to be wrong. If it’s “new”, how would one know for sure ahead of time…even with prior plausibility…that it’s right. With that sort of foresight, there’d be no need to test the hypothesis in the first place (i.e. do a study) … and one could be confident of winning the LOTTO every week.
There’s a really, REALLY big difference between fraud and wrongness…which is why the former gets retracted and being wrong does not. For no other reason than a wrong hypothesis does warrant a place in the scientific archive, so that researchers have a database of seemingly promising ideas that didn’t pan out in the past…and possibly testing said “failed” ideas with developing technology and demonstrating that they did, in fact, hold water.
And by coincidence (it really is…I was walking through our living room and caught a small relevant snippet of this podcast as opposed to parking my arse to look for more Google stuff to validate a claim)
Buried in this (not sure of the minute mark) was a very appropriate but of conversation with now Professor Ben Goldacre about the development of the alltrials database. All about understanding the statistics (including those that appear to be right)
Look what happened to the president of Harvard… ex-president. She did next to zero wrong in her papers.
You might remember I qualified my comments on theories as conclusions. The problem in my view in practical medicine the conclusions are so wrong.
If you want the scientific method then the jump to a conclusion should happen a lot less. That is not the case. Very often the happy ending to research is wrong.
I do not care how you latch onto fraud, which is the topic. Wrong is the bigger problem. The more common problem that bedevils medicine.
If you think the press wants to report the difference to lynch mobs you are wrong.
The problem with what you are saying is the baby is being thrown out with the bathwater. The baby was often a miscarriage anyway. Many in the public know there is a problem. Fertile grounds for the anti-vaxxers.
I am pretty sure you cannot back that up with evidence, unless your definition of “mountains” is more like “occasional”. Depending on how one defines “Medicine” there are about 1-2 million peer-reviewed medical papers published every year. What percentage of these do you believe are wrong and what is the basis for that belief?
Perhaps you can provide an example of a paper that you believe was wrong.