Posted on 11/22/2010 1:45:21 PM PST by Casey Hendrickson KDOX
“As it turns out, you are far more likely to be given inaccurate (or fabricated) scientific data by American scientists, than any other group of scientists in the world.”
Is wrong.
Far more likely is a specific thing, and he is WRONG WRONG WRONG.
You are THREE times more likely to be given fabricated data by an Indian scientist and TWICE as likely to be given fabricated data by a Chinese scientist.
The following sentence is true no matter what you fill in the blank with.
Most (fill in the blank) scientific papers will be American (rather than any other nation).
Most GOOD scientific papers will be American.
Most BAD scientific papers will be American.
Most scientific papers will be American.
Saying that most scientific papers retracted will be American is futile because it is obvious.
Saying that most scientific papers with fraud will be American is futile because it is obvious.
Saying (as the blogger did) that you are more LIKELY to get fraudulent data from an American scientist is ABSOLUTELY WRONG. EVERYTIME.
That is not a “viewpoint”, that is the TRUTH.
It is also undeniably the truth, using your numbers, that Americans provide 29.8% of the papers written in English and provide 42.6% of the fraudulent papers written in English. If all things were equal Americans should only provide 29.8% of the fraudulent papers. The actual percentage is much more. That means something. If all things were equal, I would expect that 29.8% of all scientists who published a paper in English(pubmed etc.) and lived within 10 miles of my home would be American. Well I am pretty well convinced that nearly 100% would be American? And there is a reason for that.
I disagree with your wording. It is not obvious, it is expected. Given that Americans produce the most papers, at what point would you consider the fact that Americans produce the most fradulent papers as significant. 100 more papers? 1000? Or are statistics a fool's errand?
It means something, but not what they said it meant, what he said was absolutely 100% wrong, no matter how much you want to quibble over it and get postmodern about viewpoints.
And if you look at the rate of fraud, it is extremely low overall, and America is hardly a standout, especially compared to India and China.
5 papers out of 100,000 in the USA, 11 for China, and 18 for India? The per year homicide rate for 100,000 people in the USA is 16!
So a person you know here in the USA is three times more likely to get murdered during the year than a paper by a US scientist is likely to be fraudulent.
Wow. What a nest of vipers those science guys are!
That is your problem. You are taking this personally. Who really gives a rat's patootie? I know that American science for the vast part is on the up and up. BUT, the data does show that there is something to be concerned about. Despite your prejudices to the contrary. The 42% vs 29% disparity is unexplained.
So a person you know here in the USA is three times more likely to get murdered during the year than a paper by a US scientist is likely to be fraudulent.
Yep, and a fraudulent paper is more likely to be written by an American out of proportion to the expectation.
Yeah, and who cares if people then further MISREPRESENT what the mangled numbers mean because of total mathematical incompetence and an anti-science bias?
Well anyone who cares about truth, mathematical competence, American prestige, and science. That is who would care.
Now the numbers are mangled? You accuse a scientist who published a paper in a peer-reviewed journal as anti-science? That is ludicrous. And if stating that a certain group of scientists are more prone to "lie" than other groups of scientists is anti-science. Then you are doubly so since you accuse the Chinese and Indians of being more prone.
It is undeniably a fact that Americans produce 42% of the fraudulent papers when they should be expected to only produce 29% of that group. The Chinese produce ~10% of the fraudulent papers and are expected to produce 11% of that group.
Rates are simple divisions and have no meaning until comparisons are made. 60 MPH is neither fast nor slow until it is compared to something. 60MPH is fast for the speed of a land animal but is slow for the speed of a passenger jet.
The number you tout means that if you pick one out of papers produced by only Americans the probablility that it will be fraudulent is X. X happens to be smaller than the probability if you used the group of papers produced by only Chinese or likewise the Indians. But if you threw all of the papers into one group and did the same selection the X for Americans would be larger than for any other group. That is to be expected since Americans produce 29% of the total papers(in english etc.)BUT!!! The expectation would be for 29% of the fraudulent papers and not 42% as it is. The US produced 84 fraudulent papers. They shoud have produced only 57. Using 57 as the number of fraudulent papers the rate should have been .003% instead of .005% fraud.
And finally, the fact is, someone has to be the winner in any "contest" of statistics. By your admission, German scientists are more "honest" than American scientists.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.