If the result is important, people will try to do work based on it and sooner or later figure out that the foundation is rotten and why.
Most of the time the work isn't important, except to the "scientist" himself in accumulating published papers ("publish or perish").
This is one of the reasons there's no need to directly double check every scientific conclusion, instead you tend to work on good faith and then backtrack when things get screwy.
Exactly. And that's often how fraud gets found. You are trying to build on someones research and some of the assumptions aren't working or you can't baseline the work. That usually leads to trying to reproduce the original work and voila ... fraud at the worst (I'd say most of the time it's just sloppy work)
I disagree with your assertion, due to the consequence of a similar series of faked results first reported on earlier this week.
Essentially, people wasted 2 months of distributed computing time trying to use the data from a fake crystal structure. Unsurprisingly, these were unsuccessful. Hundreds of thousands of compute hours were wasted and people were diverted from making real discoveries because the structure was already 'solved'. This is a huge setback, in my opinion.
A crystallographic paper is more or less just a structure determination in a common file format (CIF - the article talks about it). There's very little in the way of interpretation.
As such, they're way easier to falsify than papers in even really closely related fields. I've spent plenty of time pulling structures from ICSD and other databases populated off Acta Cryst; my PhD was in computational mineralogy - but a raw simulation isn't a publishable result in the way data hot off the diffractometer can be.
By the way, if anyone here gets the chance to visit a neutron source (eg Isis, isis.rl.ac.uk), take it! Amaaaazing machines.
I've wasted 2 months of distributed computing time on just bad assumptions made by mistakes I made, so as noted below, in the big picture, that's a relatively minor loss. Does it set back science. Of course, but is it catastrophic, no.
As someone pointed out above, in fields where you collection and analysis times are measured in years if not decades you could run into trouble, but for chemistry and biology where you are at significantly shorter collection timescales, it isn't that big of an issie
You are probably right about these fake crystal structures. The moment someone tries to do some chemistry based on these structures, they will fail and eventually figures out that the structure is wrong.
However, your conclusion doesn't follow for fields where experimental results are few and far between. Fraud would be much harder to detect in a field like astrophysics or econometrics. There simply isn't enough data, so rather than contradicting the fraudulent results, you will probably just average truth with lies.
What the fuck is wrong with the scientific community? First Elsevier (who faked an entire journal), then the East Angela university climate group and now this?
Prominent? Based on what? I've never heard of Jinggangshan University before. It's not even top 5 in Jiangxi province, where there ain't many universities.
When I was in university, I had a friend who had faked her thesis to get a Master's degree from a university in Beijing. She admitted to me that she had simply found an obscure English journal paper and translated it into Chinese.
I wouldn't be surprised if this incident happened in the same atmosphere that existed at her school: new grad students who have grown accustomed to cheating on their way through undergrad, and professors who don't check their students' work.
Anybody else make it to Paragraph Five before they were 100% certain they weren't talking about a "Crystal Structure" file system? This line, in particular, had me puzzled: "Testing is routinely carried out using cifs"
This is why I couldn't handle working in the hard sciences. While there are breakthroughs every day, there is little time for institutions to go through and verify these findings before they become commonly accepted.
It is astonishingly easy to find examples of people basing years of research off of faulty premises from other papers who wind up losing those years of research.
Shit so this is a second one already?
these were Dr H. Zhong and Professor T. Liu and the previous one http://news.ycombinator.com/item?id=1006478H.M. Krishna Murthy
I start to feel good about going to the industry!
21 comments
[ 6.2 ms ] story [ 53.6 ms ] threadIf the result is important, people will try to do work based on it and sooner or later figure out that the foundation is rotten and why.
Most of the time the work isn't important, except to the "scientist" himself in accumulating published papers ("publish or perish").
This is one of the reasons there's no need to directly double check every scientific conclusion, instead you tend to work on good faith and then backtrack when things get screwy.
Essentially, people wasted 2 months of distributed computing time trying to use the data from a fake crystal structure. Unsurprisingly, these were unsuccessful. Hundreds of thousands of compute hours were wasted and people were diverted from making real discoveries because the structure was already 'solved'. This is a huge setback, in my opinion.
A setback, yes. Huge for those who wasted their time? Maybe.
But compare it to, say, Lysenkoism. That was a huge setback for Soviet biology.
Wait, is this a separate scandal? What's with all this bogus crystallography?
The results were certainly different but perhaps the lab is the same. I'll check when I'm not on my iPhone.
As such, they're way easier to falsify than papers in even really closely related fields. I've spent plenty of time pulling structures from ICSD and other databases populated off Acta Cryst; my PhD was in computational mineralogy - but a raw simulation isn't a publishable result in the way data hot off the diffractometer can be.
By the way, if anyone here gets the chance to visit a neutron source (eg Isis, isis.rl.ac.uk), take it! Amaaaazing machines.
As someone pointed out above, in fields where you collection and analysis times are measured in years if not decades you could run into trouble, but for chemistry and biology where you are at significantly shorter collection timescales, it isn't that big of an issie
However, your conclusion doesn't follow for fields where experimental results are few and far between. Fraud would be much harder to detect in a field like astrophysics or econometrics. There simply isn't enough data, so rather than contradicting the fraudulent results, you will probably just average truth with lies.
Google says http://www.google.com/search?q=%E4%BA%95%E5%86%88%E5%B1%B1%E...
Results 1 - 10 of about 367,000 for 井冈山大学
It's not established until 2003.
http://ycombinator.com/newsguidelines.html
"You can make up a new title if you want, but if you put gratuitous editorial spin on it, the editors may rewrite it."
The original title was just "editorial," but adding the word "prominent" does sound like spin to me.
I wouldn't be surprised if this incident happened in the same atmosphere that existed at her school: new grad students who have grown accustomed to cheating on their way through undergrad, and professors who don't check their students' work.