Beesource Beekeeping Forums banner

1 - 10 of 10 Posts

·
Registered
Joined
·
2,049 Posts
Discussion Starter #1

·
Registered
Joined
·
286 Posts
Unfortunately, these things occur with some regularity in science. The good news is that on the scale of science as a whole (e.g. over a hundred thousand published articles each year) these cases of fraud are very rare. The bad news is that their existence undercuts confidence in the work of those not engaged in fraud.

The best news is that we (scientists/journals) are getting much better at detecting fraud. Many journals now use machine-learning algorithms to check all submitted works for evidence of data manipulation, and there is a large movement towards open data (i.e. you don't just publish the paper, but also release all of the raw data to the public) which makes it near-impossible to get away with fraud.

It sucks, but it is a limited issue.
 

·
Registered
Joined
·
2,049 Posts
Discussion Starter #3
Thanks for your reply.

I agree with all of your points.

I don't understand the people that perpetrate these frauds. They are not idiots that somehow were allowed to run amuck in a lab. They are very intelligent people and I assume well paid. I wonder if it they are responding to pressure from above. If so, then I am suspect of a lot of what I have read on their bee research. When I do an internet search and someone from a respected University is the author I tend to choose those.

We all benefit overall from science. I don't want to go back to the good old days. :)

I believe it is probably limited in scale, but dang it sure does cause a lot of doubt.

Alex
 

·
Registered
Joined
·
286 Posts
There are two main reasons that seem to drive most of this:

1) Science is extremely competitive - the first person to publish a discovery gets all the credit; anyone who follows (even by days) looses the credit. Careers and made (and lost) in single studies. That puts a lot of pressure on us, and to some, drives them to cheat in order to get an 'edge' on their competition. I'm fortunate enough to be in a tenured position, so I have a bit of a safety net - but for many it is literally publish (and publish well) or perish.

2) Fossilisation. We are still human, and some scientists have trouble accepting that the pet hypothesis/model they've been working on for decades may be incorrect. So results are reported incorrectly, to keep their treasured ideas alive. That appears to be what occurred with the researcher whose mentioned in your article.
 

·
Registered
Joined
·
2,049 Posts
Discussion Starter #5
Thank you for your candor and patience.

I know very little about the world of academia.

Thanks again,
Alex
 

·
Registered
Joined
·
5,870 Posts
Here is an explanation that may explain one aspect sometimes seen in latter parts of a career.

"I know that most men, including those at ease with problems of the greatest complexity, can seldom accept even the simplest and most obvious truth if it be such as would oblige them to admit the falsity of conclusions which they have delighted in explaining to colleagues, which they have proudly taught to others, and which they have woven, thread by thread, into the fabric of their lives." - Tolstoy."
 

·
Registered
Joined
·
2,313 Posts
Suggest you look-up Sir Peter Medawar's 1964 BBC talk entitled, " Is the Scientific Paper a Fraud ? " - in which he concludes that it is, albeit not fraudulent in the conventional sense as used in this thread.

With regard to biological experiments I would take his criticism one stage further, and point out that a) biological experiments are invariably based on the manipulation of a single variable, whereas biological systems themselves are multi-variable and infinitely more complex as a result of this, and b) biological experiments are frequently conducted as Tabula Rasa experiments, even when the subjects under examination are known to have memory.

Not so much 'fraud' perhaps, but a massive self-deception regarding the validity of findings ...
LJ
 

·
Registered
Joined
·
286 Posts
little_john, I work in the biological sciences and you're not really correct. Some experiments are setup to manipulate single variables - in many cases that is the only way you can be certain that the response you observe is due to that manipulation - but it is rare in the extreme to find a study which only investigates/manipulates a single variable across all experiments in the study. That era died in the 1980's.

Moreover, most studies these days (at least in biomedical-orientated work) incorporate methodologies which are explicitly designed to provide multivariate outputs that give you a broad sense of the changes occurring in your system. As one example, in my lab we routinely use methods which tell us how every single gene in the human genome changes in our patient samples or in our model systems. We also use approaches such as mass spectometry and CHiP to get unbiased data on how the protein and gene-regulatory networks change in our patients and models.

And my lab is far from special in our use of those approaches - those are the modern standard.
 

·
Registered
Joined
·
2,049 Posts
Discussion Starter #9
LittleJohn,

Thanks for the link.

I will have to read that at least one more time as it is over my head. :)

Alex
 

·
Registered
Joined
·
567 Posts
In fact one variable can wreck a theory. Maybe I should have said one overlooked fact. The ability to figure out what questions to ask can be more important than years of work and experimentation. You can think of it as "I am going to find out if A or B is the cause of something" when it is actually C or D etc. Looking for proof that something is either A or B can be a narrow limiting factor.

Then there are facts that upset a whole lot of research and theories. When industries and a ton of money depend on some falsely well held opinions and practices, the threat of exposure can cause a lot of unethical obfuscation.

https://www.huffpost.com/entry/of-mice-and-men-unseen-da_b_1352201
 
1 - 10 of 10 Posts
Top