November 19, 2017, 09:13:30 AM

Author Topic: Statistics show Scientific Fraud more common than we think  (Read 950 times)

Offline Taal

  • Adventurer
  • *
  • Posts: 65
  • Gender: Male
Statistics show Scientific Fraud more common than we think
« on: July 05, 2017, 10:54:11 PM »

Quote
Statistics Hint There is Much More Scientific Fraud than Reported

Fraud, errors, and falsified data winds its way into the smallest niche publications, as well the most prestigious journals with the greatest impact factors. But the ones who get caught are generally the most egregious offenders, like a young researcher in Japan who published “game changing” results on stem cells that didn’t exist.

But how many studies with less bold claims fly under the radar, and are never caught by an editor or a peer reviewer?

Many, answers a new statistical analysis of thousands of studies in eight prestigious journals, published this week.

The statistical analysis of 5,087 randomized, controlled trials turned up dozens that data had to be skewed for the end result, according to John Carlisle, the author of the paper in the journal Anaesthesia.

“Fraud, unintentional error, correlation, stratified allocation and poor methodology might have contributed to the excess of randomized, controlled trials with similar or dissimilar means, a pattern that was common to all the surveyed journals,” writes Carlisle.

“It is likely that this work will lead to the identification, correction and retraction of hitherto unretracted randomized, controlled trials,” he adds.


Of the 5,087 studies, 72 had already been retracted. But Carlisle’s statistical analysis found that 43 more almost certainly contained problematic data and results.

https://www.laboratoryequipment.com/news/2017/06/statistics-hint-there-much-more-scientific-fraud-reported

science is controlled by humans with all their agendas, biases and politics.

Offline JamSiManYo

  • Adventurer
  • *
  • Posts: 64
  • Gender: Male
Re: Statistics show Scientific Fraud more common than we think
« Reply #1 on: July 06, 2017, 12:10:52 AM »
Yup. It's been an ongoing problem for at least a couple of decades because the entire system practically encourages this sort of behavior. There is a lot of pressure to publish, especially for new scientists. This leads to studies about really obscure things just to get published.  And because funding is finite, there's fierce competition for authorship. This means fudging the data so that you can present the conclusion you hoped to find. Positive results lead to higher impact which in turn brings you more funding for other projects. And the people who fund the new projects tell the scientists what sort of results they expect to see.

ETA: I'm suddenly reminded of the two biggest controversies I heard growing up. First eggs were bad, then they were okay, then they were bad for you again, and now they're good for you. Same thing happened with milk.
« Last Edit: July 06, 2017, 12:12:58 AM by JamSiManYo »

Offline Rumbledy Hump

  • Super Waygook
  • ***
  • Posts: 360
  • Gender: Male
Re: Statistics show Scientific Fraud more common than we think
« Reply #2 on: July 06, 2017, 02:17:42 AM »
This analysis found 2% of studies with fudged results. That's hardly an indictment of science as a whole. In fact, researchers don't make conclusions off of single studies in case there is a flaw in its design, intended or otherwise.

But herein lies the problem brought up by JamSiManYo. Because academics want to make a name for themselves, there is a lack of incentive to conduct studies that repeat original research, making it less likely these errors will be uncovered.
But how do I know what to lick and what not to lick?

Offline JNM

  • The Legend
  • *****
  • Posts: 2442
  • Gender: Male
Re: Statistics show Scientific Fraud more common than we think
« Reply #3 on: July 06, 2017, 05:44:01 AM »
This analysis found 2% of studies with fudged results. That's hardly an indictment of science as a whole. In fact, researchers don't make conclusions off of single studies in case there is a flaw in its design, intended or otherwise.

But herein lies the problem brought up by JamSiManYo. Because academics want to make a name for themselves, there is a lack of incentive to conduct studies that repeat original research, making it less likely these errors will be uncovered.
There is a movement afoot in the scientific community to do more replicate studies. I am not sure how they will incentivize this.

Offline KimDuHan

  • Hero of Waygookistan
  • *****
  • Posts: 1158
  • Gender: Male
Re: Statistics show Scientific Fraud more common than we think
« Reply #4 on: July 06, 2017, 06:30:34 AM »
Science is updated all the time look at science books from 30 years ago lots of the information is outdated or plain out wrong!!

Especially with the earth's crust and core! Lots of early texts were way off on dinosaurs and other early lifeforms.

Offline Arabin

  • Super Waygook
  • ***
  • Posts: 259
Re: Statistics show Scientific Fraud more common than we think
« Reply #5 on: July 06, 2017, 07:23:20 AM »
72 out of 5000 odd? Whoah, there Woodward. Are you sure you should post such a devastating indictment of science as a whole? The scientific-industrial complex might come after you.

I look forward to your flat earth theories, your anti-vaxxer diatribes and of course, your elaborate explanations of why Climate Change is a conspiracy by Freemasons/reptilians/globalists/Jews/liberals/the Chinese. I will naturally treat your words as equal to a PhD with 25 years of research experience. After all, you have access to the internet. That makes your views valid. Interesting too.

Offline sonofthedude777

  • Veteran
  • **
  • Posts: 172
  • Gender: Male
Re: Statistics show Scientific Fraud more common than we think
« Reply #6 on: July 06, 2017, 01:14:44 PM »
Science is always right. NDT said!
I am not a number. I am a free man.

Offline jddavis7

  • Super Waygook
  • ***
  • Posts: 413
  • Gender: Female
Re: Statistics show Scientific Fraud more common than we think
« Reply #7 on: July 06, 2017, 01:25:14 PM »
Oh look! Another discussion where no one read the actual study. If I'm correct in my reading, this study shows that outlier studies (ones that go against what is usually accepted in the scientific community) tend to be where issues of data fabrication (about participants weight, age, etc) lead to later retraction.

Offline Taal

  • Adventurer
  • *
  • Posts: 65
  • Gender: Male
Re: Statistics show Scientific Fraud more common than we think
« Reply #8 on: July 06, 2017, 07:56:02 PM »
Science is updated all the time look at science books from 30 years ago lots of the information is outdated or plain out wrong!!

Especially with the earth's crust and core! Lots of early texts were way off on dinosaurs and other early lifeforms.

If science and the popular definition of what is true is constantly changing, then what makes you think its correct now?

Offline hippo

  • Veteran
  • **
  • Posts: 131
  • Gender: Male
Re: Statistics show Scientific Fraud more common than we think
« Reply #9 on: July 06, 2017, 10:24:51 PM »
Science seeks to create the best understanding of the way things are based on the limited information people have at any given time.  The great thing about science is that one of its primary goals is to prove itself wrong to continually improve understanding of how things are. 

There are biases that can impede and/or distort scientific discovery.  But is there anything more effective than science in achieving an understanding of how the world works?  When science is distorted or there is insufficient evidence, science itself is the best weapon against misunderstandings.  Just this month I read about several studies saying humans (Homo sapiens) have existed much longer than previous thought.

Offline donovan

  • The Legend
  • *****
  • Posts: 2143
  • 후레쉬
Re: Statistics show Scientific Fraud more common than we think
« Reply #10 on: July 06, 2017, 11:16:16 PM »
There's a catch 22 here: Statistics bear out that science is fake, while statistics are scientifically proven to be dubious.

I always teach from the gut~


Offline JNM

  • The Legend
  • *****
  • Posts: 2442
  • Gender: Male
Re: Statistics show Scientific Fraud more common than we think
« Reply #11 on: July 07, 2017, 07:22:23 AM »
There's a catch 22 here: Statistics bear out that science is fake, while statistics are scientifically proven to be dubious.

I always teach from the gut~

It isn't the science of statistics that is dubious, but rather the misunderstanding and deceptions while communicating results.

"On average, Germans are taller than Filipinos" gets shortened to "Germans are taller than Filipinos", then some tall Filipino (or more likely, someone who once met a tall Filipino) gets his nose out of joint and seeks to discredit the original researcher, without having read the paper. 

But he has a blog... so it's legit, right?


Offline donovan

  • The Legend
  • *****
  • Posts: 2143
  • 후레쉬
Re: Statistics show Scientific Fraud more common than we think
« Reply #12 on: July 07, 2017, 07:28:13 AM »
There's a catch 22 here: Statistics bear out that science is fake, while statistics are scientifically proven to be dubious.

I always teach from the gut~

It isn't the science of statistics that is dubious, but rather the misunderstanding and deceptions while communicating results.

"On average, Germans are taller than Filipinos" gets shortened to "Germans are taller than Filipinos", then some tall Filipino (or more likely, someone who once met a tall Filipino) gets his nose out of joint and seeks to discredit the original researcher, without having read the paper. 

But he has a blog... so it's legit, right?

There is a positive correlation between misunderstanding and misinterpretation. Causation still unclear.

Offline Mr C

  • Hero of Waygookistan
  • *****
  • Posts: 1211
  • Gender: Male
Re: Statistics show Scientific Fraud more common than we think
« Reply #13 on: July 07, 2017, 10:12:02 AM »
Science is updated all the time look at science books from 30 years ago lots of the information is outdated or plain out wrong!!

Especially with the earth's crust and core! Lots of early texts were way off on dinosaurs and other early lifeforms.

If science and the popular definition of what is true is constantly changing, then what makes you think its correct now?

It's more correct.  If our knowledge of everything was perfect and complete, science would just stop.

The thing some people seem to be missing is that the study referred in the OP is itself science.  This is part of how science works.

And there are no statistics that bear out that science is fake.  Just imperfect.  If anyone has a better system for arriving at truth, we're all ears.


Offline Arabin

  • Super Waygook
  • ***
  • Posts: 259
Re: Statistics show Scientific Fraud more common than we think
« Reply #14 on: July 07, 2017, 01:29:47 PM »