EEVblog Electronics Community Forum
General => General Technical Chat => Topic started by: dannyf on September 04, 2016, 04:34:00 pm
-
Nice read here: http://www.sciencemag.org/news/2016/09/whistleblower-sues-duke-claims-doctored-data-helped-win-200-million-grants (http://www.sciencemag.org/news/2016/09/whistleblower-sues-duke-claims-doctored-data-helped-win-200-million-grants)
Before that, they had Potti, and a couple others involved.
Given the saintly reputation of those "scientists", I doubt any of this fraud is motivated by money.
The same for other researchers in you-know-what fields, :)
BTW, FCA isn't applicable to public universities that collectively get the bulk of the research dollars. So the duke story is only a tip of an iceberg.
Your tax dollars at work, :)
-
The other day I watched this video, it deals with p-hacking. Very interesting. Once you cherry pick the hypothesis, the statistical assumptions are invalid.
https://www.youtube.com/watch?v=42QuXLucH3Q (https://www.youtube.com/watch?v=42QuXLucH3Q)
Related
https://en.wikipedia.org/wiki/Data_dredging (https://en.wikipedia.org/wiki/Data_dredging)
http://retractionwatch.com/ (http://retractionwatch.com/)
Science is a very useful methodology that is practiced by people with varying levels of adherence.
Once you have humans in the loop, anything is possible ;-)
-
https://en.wikipedia.org/wiki/N_ray
"Once you have humans in the loop, anything is possible ;-)"
Was, is, and will be. :)
-
it deals with p-hacking.
I had a guy telling me that he had consistently made money, if you took out the losses. :)
and he is absolutely right.
-
Some articles on inability to replicate research, whether due to fraud or errors.
Also, publishing of rubbish research papers.
20131018
http://www.economist.com/news/briefing/21588057-scientists-think-science-self-correcting-alarming-degree-it-not-trouble (http://www.economist.com/news/briefing/21588057-scientists-think-science-self-correcting-alarming-degree-it-not-trouble)
Scientists like to think of science as self-correcting. To an alarming degree, it is not
20140213
http://www.nature.com/news/scientific-method-statistical-errors-1.14700 (http://www.nature.com/news/scientific-method-statistical-errors-1.14700)
Scientific method: Statistical errors
P values, the 'gold standard' of statistical validity, are not as reliable as many scientists assume.
(Also, the cool term 'P-hacking')
20140303
http://www.libertynewsonline.com/article_330_34940.php (http://www.libertynewsonline.com/article_330_34940.php)
HOW COMPUTER GENERATED FAKE PAPERS ARE FLOODING ACADEMIA
http://www.theguardian.com/technology/shortcuts/2014/feb/26/how-computer-generated-fake-papers-flooding-academia#sthash.MVj6gy74.dpuf (http://www.theguardian.com/technology/shortcuts/2014/feb/26/how-computer-generated-fake-papers-flooding-academia#sthash.MVj6gy74.dpuf)
20140304
http://ethicalnag.org/2009/11/09/nejm-editor/ (http://ethicalnag.org/2009/11/09/nejm-editor/)
NEJM editor: “No longer possible to believe much of clinical research published”
“It is simply no longer possible to believe much of the clinical research that is published, or to rely on the judgment of trusted physicians or authoritative medical guidelines. I take no pleasure in this conclusion, which I reached slowly and reluctantly over my two decades as an editor of The New England Journal of Medicine.”
20140304
http://phys.org/news/2014-02-scientific-young.html#nwlt (http://phys.org/news/2014-02-scientific-young.html#nwlt)
To teach scientific reproducibility, start young
The ability to duplicate an experiment and its results is a central tenet of the scientific method, but recent research has shown an alarming number of peer-reviewed papers are irreproducible.
WRH: Including virtually all of the papers regarding human-caused global warming.
20140312
http://blogs.nature.com/news/2014/03/call-for-acid-bath-stem-cell-paper-to-be-retracted.html (http://blogs.nature.com/news/2014/03/call-for-acid-bath-stem-cell-paper-to-be-retracted.html)
Call for acid-bath stem-cell paper to be retracted
Less than 40 days after a team led by Haruko Obokata of the RIKEN Center for Developmental Biology in Kobe, Japan, presented two stunning papers claiming a method of using a simple acid-bath method to reprogramme mature mammalian cells back to an embryonic state – so called STAP cells – researchers in Japan, including one of the paper’s co-authors, are calling for them to be retracted.
20140323
http://yaledailynews.com/blog/2014/03/12/former-postdoc-files-suit-against-university/ (http://yaledailynews.com/blog/2014/03/12/former-postdoc-files-suit-against-university/)
Former postdoc files suit against University
A former post-doctoral researcher at the Yale Medical School has filed a suit against the University, her supervisor and a former post-doctoral fellow over what she alleges was deliberate tampering with her research.
20140424
Blinded by scientific gobbledygook
http://www.ottawacitizen.com/technology/Blinded+scientific+gobbledygook/9757736/story.html (http://www.ottawacitizen.com/technology/Blinded+scientific+gobbledygook/9757736/story.html)
I have just written the world’s worst science research paper: More than incompetent, it’s a mess of plagiarism and meaningless garble.
Now science publishers around the world are clamouring to publish it.
They will distribute it globally and pretend it is real research, for a fee.
It’s untrue? And parts are plagiarized? They’re fine with that.
Welcome to the world of science scams, a fast-growing business that sucks money out of research, undermines genuine scientific knowledge, and provides fake credentials for the desperate.
And even veteran scientists and universities are unaware of how deep the problem runs.
20140710
http://www.uk.sagepub.com/aboutus/press/2014/jul/7.htm (http://www.uk.sagepub.com/aboutus/press/2014/jul/7.htm)
Retraction of 60 articles implicated in a peer review and citation ring (sagepub.com)
https://news.ycombinator.com/item?id=8009333 (https://news.ycombinator.com/item?id=8009333)
20160904
http://www.publications.parliament.uk/pa/cm201012/cmselect/cmsctech/856/856vw.pdf (http://www.publications.parliament.uk/pa/cm201012/cmselect/cmsctech/856/856vw.pdf)
House of Commons - Science and Technology Committee
Peer review in scientific publications
Eighth Report of Session 2010–12
http://www.collective-evolution.com/2015/05/16/editor-in-chief-of-worlds-best-known-medical-journal-half-of-all-the-literature-is-false/ (http://www.collective-evolution.com/2015/05/16/editor-in-chief-of-worlds-best-known-medical-journal-half-of-all-the-literature-is-false/)
-
I am yet to get a good grasp of that p-hacking thing. For example, they tell me that if I comes with thousands of hypothesis, test all of them against a data set and pick the low P ones I will have a large error rate.
For example, how this is different from thousands of researches coming with one hypothesis each and testing against a respective data set and publish if it has a low P? I would think that the error rate in both cases should be about the same and it is derived from the probability of one hypothesis to give a false low P against a single data set.
Any, I find this to be fascinating. The science of science or meta science if you will.
-
I am yet to get a good grasp of that p-hacking thing.
it is from hypothesis testing: you are looking for the probability that the null hypothesis is true.
What people mistake most here is that even if rejected / accepted, the hypothesis could still be true / untrue (ie. type 1/2 errors). ie, the conclusion is never absolute.
-
Publishing succesful "scientists" from our country often get 2x in the Mainstream Media too.
The ones with the most spectacular breakthrough get the second time in the news for the most spectacular fraud.
There are complete departments at our universities that got sucked into groupthink, it's impossible to call a "researcher" in those "fields" fraudulent, because he/she/it complies completely.
Complete departments have to be shut down, start over again, or stop and call their "domain" a fraud.
-
I find the complete lack of understanding of the scientific process in the media distressing. They don't know, or perhaps care enough to question the 'research' they report on. Several months ago there was a piece on the morning news about a study of 'law enforcement' committing certain crimes. Of course what was reported was a total number of cops charged over a number of years. The first things I thought of was how do those numbers compare to the general population and why those particular years were chosen.
It took perhaps 15 minutes of research to find the per capita occurrence of this type of crime was several times (about 10x as I recall) in the general population. It is also common for researchers to cherry pick date ranges that support their hypothesis.
-
Did someone mention Batteriser, Batteroo, Roohparvar?
-
In a world were everything is about money, money more money, status and power, unfortuantely even "science" follows the same path.
Therefore you see a lot of "fraude" in some degree.
Sometimes not as obvious, but it's not difficult to bias the results.
Moreover, just by the fact that you need weird (expensive) memberships to even freaking read the papers is just to stupid for words. (IEEE & AES).
Some fields are really only about publishing and professors get a pretty nice bonus for every published paper.
I feel heavily ashamed as a scientist/engineer
-
I constantly get salespeople trying to sell me some drug or therapy in my industry and throw scientific articles my way. I scrutinize them carefully because I know they are biased or conclusions stretched. There is a difference between statistical significance and clinical significance. Sure you can show drug X improves things with a 95% confidence interval by some amount Y, but if Y doesn't amount to a hill of beans, is it worth using? Or when charts aren't zero scaled, they start at 0.8 and end at 1.2, then the bars look vastly different until you really look carefully.
Another example, when no drug leads to improvement by 4.0 simply through other traditional therapy, placebo drug by 4.2 due to additional psychological effects, drug A is improving by 4.4 and drug Z which is being advertised is resulting in 4.8. So they claim 2x better than drug A for drug Z because over traditional therapy alone it was 0.8 more and drug A was 0.4 more. Meanwhile, the significance of 0.8 is not worth the cost or complications or risks of taking the drug. But all the marketing says Z is 2x better than A at improving whatever.
-
Moreover, just by the fact that you need weird (expensive) memberships to even freaking read the papers is just to stupid for words. (IEEE & AES).
a AES member still needs to pay for each paper, only a bit less. ^-^
-
This fellow named Aaron Schwartz was hoping to change all that when the powers that be landed on him so hard he took his own life.
https://en.wikipedia.org/wiki/Aaron_Swartz (https://en.wikipedia.org/wiki/Aaron_Swartz)
Science publications and especially all studies funded by the public through government grants should be freely available to the public. Let Aaron's hacktivism not be forgotten or in vain.
And I quote from Wikipedia JSTOR article:
The articles that were downloaded by Aaron were from scholarly journals which published scientific papers largely funded by public universities and taxpayer money.
The following month, federal authorities charged Swartz with several "data theft"-related crimes, including wire fraud, computer fraud, unlawfully obtaining information from a protected computer, and recklessly damaging a protected computer.[21][22] Prosecutors in the case claimed that Swartz acted with the intention of making the papers available on P2P file-sharing sites.[20][23]
Swartz surrendered to authorities, pleaded not guilty to all counts, and was released on $100,000 bail. In September 2012, U.S. attorneys increased the number of charges against Swartz from four to thirteen, with a possible penalty of 35 years in prison and $1 million in fines.[24][25] The case still was pending when Swartz committed suicide in January 2013.[26] Prosecutors subsequently dropped the charges after his death.[27]
-
For example, how this is different from thousands of researches coming with one hypothesis each and testing against a respective data set and publish if it has a low P? I would think that the error rate in both cases should be about the same and it is derived from the probability of one hypothesis to give a false low P against a single data set.
This is exactly the Monty Hall problem. The key insight is that the two probabilities are not the same because the knowledge of the participants is different.
-
Moreover, just by the fact that you need weird (expensive) memberships to even freaking read the papers is just to stupid for words. (IEEE & AES).
a AES member still needs to pay for each paper, only a bit less. ^-^
Yes, I know.
Absolutely ridiculous, paying for a membership and than STILL need to pay for papers. :palm:
(one of the reason I canceled mine)
-
One of the major problems driving junk science is the institutional demand for large quantities of publications by faculty. Quality is rarely a consideration.
-
One of the major problems driving junk science is the institutional demand for large quantities of publications by faculty. Quality is rarely a consideration.
Well they were told to "publish or perish" so what do you expect? This is not so much about science but rather the human condition. This is more about the human need to succeed, to compete, and to try and be a winner. Corruption can and does encompass all human activities, even religion so science is not the cause. It can only be expected and guarded against.
-
One of the major problems driving junk science is the institutional demand for large quantities of publications by faculty. Quality is rarely a consideration.
Well they were told to "publish or perish" so what do you expect? This is not so much about science but rather the human condition. This is more about the human need to succeed, to compete, and to try and be a winner.
That's reasonable to set a metric and evaluate people by that metric, but the trick is to define the metric such that maximizing it will indeed achieve the bottom line goal.
When you incentives people, they will find creative ways to increase their score.
-
One of the major problems driving junk science is the institutional demand for large quantities of publications by faculty. Quality is rarely a consideration.
Well they were told to "publish or perish" so what do you expect? This is not so much about science but rather the human condition. This is more about the human need to succeed, to compete, and to try and be a winner.
That's reasonable to set a metric and evaluate people by that metric, but the trick is to define the metric such that maximizing it will indeed achieve the bottom line goal.
When you incentives people, they will find creative ways to increase their score.
I agree totaly. Hence my signature "measurement changes behavior".
-
When you incentives people, they will find creative ways to increase their score.
Yeah. Lots of inefficiencies. A typical 4-yr education in a reasonably good college costs $250K, or $500K pre-tax. It is getting more and more difficult to justify that kind of investment. That's a huge business that has defied the productivity gains over the last 30 - 40 years. The first person cracking it will make a killing.
-
This fellow named Aaron Schwartz was hoping to change all that when the powers that be landed on him so hard he took his own life.
....
Never mentioned is that he had threatened suicide years earlier, when he wasn't happy at his job. The police were called and broke into his apartment. For some people it takes very little provocation for them to kill themselves.
-
Even this guy had something to say on the subject https://youtu.be/0Rnq1NpHdmw
-
Yeah. Lots of inefficiencies. A typical 4-yr education in a reasonably good college costs $250K, or $500K pre-tax. It is getting more and more difficult to justify that kind of investment. That's a huge business that has defied the productivity gains over the last 30 - 40 years. The first person cracking it will make a killing.
You can just go to Germany and finish your studies for a fraction of that cost.
-
"For some people it takes very little provocation for them to kill themselves."
Whose fault is it in those cases? His job, his boss, his employer, society at large, his parents, his teachers, or himself? Or something else.....
-
Those amounts of money or beyond sick.
People expect that 17-18 year olds magically can plan their entire life.
When switching from one study to another, or having one or two years delay (because of important personal reason or whatever) and it's not worth the investment at all.
Than you're better off working your way up.
If you can't find work or have to quit for (health) reasons, you're in debt the rest of your life.
The whole thought makes me angry and sick.
Second is that you see more and more companies not even bothering about a degree because the level of knowledge and skill is just to sad. EE students who don't even know how to solder. Are you kidding me?
The sad part is that I see a lot of other (EU) countries following the same path. Apperently even just decent education is a luxury. Don't be shocked in a few years if we run out of good employers, violent rates go up and a lot of people (and therefore banks) will be in debt.
-
Those amounts of money or beyond sick.
The cost of education went up because the easy to get students loan flooded the market with money.
I would tie the loan amount to the expected income at the discipline the student choses. Also, I would drastically cut down all the general education classes, here for example, if you learn engineering they also force you to take a very significant amount of non related classes.
-
EE students who don't even know how to solder. Are you kidding me?
As an interviewer for electronic and software engineers over several decades in several companies, the ability to solder is very low on my list of priorities. The only useful thing it indicates is whether or not a candidate is interested in electronics. For technicians soldering skills are more important. Soldering can be taught in a week. The topics below take years to begin to understand.
So, what kind of things have been interesting/useful to my employers? Knowing:
- how to elicit requirements from a customer
- how to choose between MCU / FPGA / software for a particular design
- how to use an IQ modulator in a radio circuit
- which modulation scheme and protocol to use
- why there's no such thing as "global time" nor "ground"
- realising that with high voltages/currents you will need to develop good models and simulations to check the energy doesn't end up somewhere unwanted
- under what circumstances digital signals must be terminated, and how to choose the termination method
- what type of analogue filter to choose, and why
I was once criticised in a performance evaluation for an adequate but untidy modification I made to a prototype. My response, "agreed; I'm not a technician", was instantly accepted as being sufficient.
N.B. a competent technician is invaluable, just as are competent nurses. But you also need doctors and engineers. Vive la difference.
-
Whose fault is it in those cases? His job, his boss, his employer, society at large, his parents, his teachers, or himself? Or something else.....
Well, the blame is to whoever you decide to assign it to. None of those actors is going to take responsibility in any meaningful way, so I consider it a moot point. What caused it is a different question, and you could delve into the proximal and ultimate causes of what happened.
Ultimately, my fear is that society has a dangerous tendency to tell people that they're "gifted" and so the rules that apply to everyone else don't apply to them.
-
"The cost of education went up because the easy to get students loan flooded the market with money. "
Absolutely. The more a good is subsidized, the more people will consume it. Dumb politicians don't understand that.
What we observed over the last twenty years in higher education is nothing but a giant wealth transfer, a reverse robbinhood, from the poorest of the poor (students) to the top of the well offs : Warren was making 750k over 9 months for example. This government facilitated transfer created millions of young slaves: students paying off their debt that cannot be discharged through the bankruptcy process.
"I would tie the loan amount to the expected income at the discipline the student choses. "
I woukd take a different approach: the schools will need tobhave a skin in the game. I would introduce a clawback to the schools and adminstrators personally on losses on student loans.
Let the school admit whoever they want, into whatever hopeless majors and do a terrible job graduating and placing their students. In the end, they are on the hooks.
Otherwise, it is going to be a huge bailout.
-
"Also, I would drastically cut down all the general education classes, here for example, if you learn engineering they also force you to take a very significant amount of non related classes."
It depends on your objective for education: to learn specific knowledge, and or for betterment of you as a human being.
In my case, it isn't that important for my kids to pick up specific knowledge - they could have done that though a vocational school at a fraction of the costs.
I would much prefer them learn more soft skills, interacting with others, learn to analyze unstructured problems and think critically. To develop their own networks.
So taking on those liberal arts or non technical classes is why I send them there.
Anything else I can do for them, far better than the schools can ever do.
-
These things are also the things you need at work.
Much more usefull than being bothered for endless semesters about xth order differential equations.
99.8% of the people don't need and use them anymore ever. Complete waste of time.
Best thing I learned at school is writing reports, doing homework and projects as efficient as possible. It also teaches you on a very efficient way how to remember things for exactly two days, because of exams (not to understand things).
-
It also teaches you on a very efficient way how to remember things for exactly two days, because of exams (not to understand things).
Well, either it wasn't a very good course, or you chose to only remember and not to understand.
Personally I chose to understand, and that made remembering much easier - and more useful in the long run. I still refer to my textbooks that I bought in the 70s, because some of the information in them is still relevant and useful.
So maybe that makes me one of your "0.2%". Good. That means I don't have to be bored doing crap jobs that can be done by the "lesser" "99.8%".
Which category would you rather be in?
-
Oh, no that's not what I mean.
I like the things I do.
I am purely talking about how the examens are given. The examens only care about remembering a few basic (non important) things. So you'll end with a lot of students who have great marks because they're good in making examens and even graduate cum laude. But actually they don't really understand what it's all about (at all!)
You can do both, but we simply didn't always had time for it because less important subject took all our time.
-
So taking on those liberal arts or non technical classes is why I send them there.
You are free to choose for your daughters double major in liberals at and engineering. No need to force it on everybody else that wants to learn engineering, math, physics or any other academic discipline.
-
Oh, no that's not what I mean.
I like the things I do.
I am purely talking about how the examens are given. The examens only care about remembering a few basic (non important) things. So you'll end with a lot of students who have great marks because they're good in making examens and even graduate cum laude. But actually they don't really understand what it's all about (at all!)
You can do both, but we simply didn't always had time for it because less important subject took all our time.
It takes any remotely competent interviewer less than five minutes to work out whether a candidate is a time server or is doing something because they like it. No competent interviewer relies on exam results.
Poor exam results are, however, something that needs to be investigated and explained, because they can indicate serious gaps in knowledge or ability.
-
Oh, no that's not what I mean.
I like the things I do.
I am purely talking about how the examens are given. The examens only care about remembering a few basic (non important) things. So you'll end with a lot of students who have great marks because they're good in making examens and even graduate cum laude. But actually they don't really understand what it's all about (at all!)
You can do both, but we simply didn't always had time for it because less important subject took all our time.
I only remember that when i studied , only saw numbers morning ,afternoon and evening ,on resume i passed of the 90% from the career as human calculator(prestudies lab,project labs and exams ).
These amount of time lost for nothing because a computer can calculated on less of the second(Here are Charlie Brown, square heads).
EE students who don't even know how to solder. Are you kidding me?
As an interviewer for electronic and software engineers over several decades in several companies, the ability to solder is very low on my list of priorities. The only useful thing it indicates is whether or not a candidate is interested in electronics. For technicians soldering skills are more important. Soldering can be taught in a week. The topics below take years to begin to understand.
So, what kind of things have been interesting/useful to my employers? Knowing:
- how to elicit requirements from a customer
- how to choose between MCU / FPGA / software for a particular design
- how to use an IQ modulator in a radio circuit
- which modulation scheme and protocol to use
- why there's no such thing as "global time" nor "ground"
- realising that with high voltages/currents you will need to develop good models and simulations to check the energy doesn't end up somewhere unwanted
- under what circumstances digital signals must be terminated, and how to choose the termination method
- what type of analogue filter to choose, and why
I was once criticised in a performance evaluation for an adequate but untidy modification I made to a prototype. My response, "agreed; I'm not a technician", was instantly accepted as being sufficient.
N.B. a competent technician is invaluable, just as are competent nurses. But you also need doctors and engineers. Vive la difference.
As you asked it to junior engeenering you will take a depression.
-
EE students who don't even know how to solder. Are you kidding me?
As an interviewer for electronic and software engineers over several decades in several companies, the ability to solder is very low on my list of priorities. The only useful thing it indicates is whether or not a candidate is interested in electronics. For technicians soldering skills are more important. Soldering can be taught in a week. The topics below take years to begin to understand.
So, what kind of things have been interesting/useful to my employers? Knowing:
- how to elicit requirements from a customer
- how to choose between MCU / FPGA / software for a particular design
- how to use an IQ modulator in a radio circuit
- which modulation scheme and protocol to use
- why there's no such thing as "global time" nor "ground"
- realising that with high voltages/currents you will need to develop good models and simulations to check the energy doesn't end up somewhere unwanted
- under what circumstances digital signals must be terminated, and how to choose the termination method
- what type of analogue filter to choose, and why
I was once criticised in a performance evaluation for an adequate but untidy modification I made to a prototype. My response, "agreed; I'm not a technician", was instantly accepted as being sufficient.
N.B. a competent technician is invaluable, just as are competent nurses. But you also need doctors and engineers. Vive la difference.
As you asked it to junior engeenering you will take a depression.
I'm not sure what you are trying to say, but any competent interviewer mutates the questions according to the candidate's background and experience.
There's little point in trying to trap a candidate into giving a "wrong" answer, or "answering" a quiestion that's outside their expertise. There's a lot of point in allowing a candidate sufficient latitude that they can show what they can do and show how they approach problem identification and solving, especially with new problems that are relevant to the job opening.
It is trivially easy for a good candidate to train themselves in the technology that your company is using today. The interviewer's job is to identify such candidates.
To illustrate the point, a simple example from thirty years ago... "A client comes and, during the first meeting, says they make toy cars and road tracks for young children, and they want to add a new working traffic lights product. What do you say and do?". Clearly there's no "correct answer" to that, but the way they approach the problem is very revealing.
I would expect a candidate to ask what is and isn't meant by "working", the production costs and timescales. Then they should be able to instantly invent and explore a range of implementation options (bonus points for considering purely mechanical solutions), using those to tease out requirements the client hasn't stated. OTOH if they can only say "I would use a microprocessor", then they have too limited imagination or experience, or are only used to "doing what they are told".
-
Man did this topic get pulled off subject. |O
Fraud in science posting started with it is or was a thing apparently and it WAS interesting to read peoples ideas, experiences, and opinions on the subject. I guess the subject was exhaustive and people jumped to a shinny new subject(s). :--
-
...Much more usefull than being bothered for endless semesters about xth order differential equations.
99.8% of the people don't need and use them anymore ever. Complete waste of time.
That's why fraud in science is so easy. Too many people are ignorant about the basic stuff, don't understand the assumptions that have been made in their fancy "simulation" and don't know how "the software" is limited or wrong in what situations.
-
That's why fraud in science is so easy. Too many people are ignorant about the basic stuff, don't understand the assumptions that have been made in their fancy "simulation" and don't know how "the software" is limited or wrong in what situations.
Yes, of course. Statistical rigor is just a annoying barrier that is oppressing them :(
Every time some muppet says "the studies say this" I wonder if they could tell a probability distribution function from a pancake.
-
Most of what these people are doing is not fraud. Fraud requires you to understand what you are doing and do it wrong intentionally. While there are people doing that, two other things explain most of these papers. Both bad, but not criminal. The first is simple ignorance. Not understanding the math well enough to realize it doesn't mean anything. The second is a common human fault, allowing pre-conceived notions to affect your understanding of results. It is very hard to avoid, even with the best of intentions.
Many universities even train for this bias towards a pre-conceived notion (mostly unwittingly). How many of you were taught a model of transistor behavior, and then sent into the lab to do experiments to confirm this result? How often did a lab supervisor force you to a write up confirming the presented model, even if bad equipment, bad implementation of the tests, sloppy record keeping, noise or other problems made that confirmation problematic?
-
Most of what these people are doing is not fraud. Fraud requires you to understand what you are doing and do it wrong intentionally. While there are people doing that, two other things explain most of these papers. Both bad, but not criminal. The first is simple ignorance. Not understanding the math well enough to realize it doesn't mean anything. The second is a common human fault, allowing pre-conceived notions to affect your understanding of results. It is very hard to avoid, even with the best of intentions.
Many universities even train for this bias towards a pre-conceived notion (mostly unwittingly). How many of you were taught a model of transistor behavior, and then sent into the lab to do experiments to confirm this result? How often did a lab supervisor force you to a write up confirming the presented model, even if bad equipment, bad implementation of the tests, sloppy record keeping, noise or other problems made that confirmation problematic?
Careful, the ignorance isn't exculpatory and less for people with superior studies, because it has a short way to manifested negligence with all responsability civil and penalty.
For example ,i am Technical engineering specialist on Electronic Industrial but i can sign projects of others specialties . For example I can design and sizing a beam for a structure,
and by ignorance i choose the first profile beam that i saw at warehouse and above the iron(instead of steel) :clap: .
Is this ignorance or negligence? Negligence, because i have been too lazy on studying the several profile beams and its materials for the project.
-
Is this ignorance or negligence? Negligence,...
gross negligence, because with your certification and background, you should be reasonably expected to anticipate the consequences of your actions.
-
I will agree on the legal ramifications, but still think some sort of distinction needs to be made between intentionally providing wrong results and being lazy or ignorant. If for no other reason than that the cures for the conditions are very different.
To put it another way, I might well be willing to have a friendly beer with a lazy or ignorant person. Much less likely for the intentional fraud.
-
I will agree on the legal ramifications, but still think some sort of distinction needs to be made between intentionally providing wrong results and being lazy or ignorant.
Sometimes the laziness or ignorance intersect with one's interests so it stays that way intentionally. Kind of a plausible deniability.
-
Most of what these people are doing is not fraud. Fraud requires you to understand what you are doing and do it wrong intentionally. While there are people doing that, two other things explain most of these papers... The first is simple ignorance....
You are inventing excuses for these establihment professors? Their fraud is intentional, from the beginning, you can´t introduce unawareness when the subject is a professor in his field.
Some blame it on publishing pressure, others on BIGmoney, I see a normal human ratrace. With a lack of de-facto correction.
Both bad, but not criminal.
Goverment decides what is fraud. Not in Laws, but in how these laws are applied.
They decide who´s jokes are seen as criminal facts, and who´s criminal facts are seen as a joke.
How many of you were taught a model of transistor behavior, and then sent into the lab to do experiments to confirm this result? How often did a lab supervisor force you to a write up confirming the presented model, even if bad equipment, bad implementation of the tests, sloppy record keeping, noise or other problems made that confirmation problematic?
I didn´t experience that in my west-european university.
-
Galenbo is right.
If someone just pretends he is ignorant, or just says certain things in a certain way, on paper (even according to 'the laws") he didn't do anything wrong.
That's a very grey area and in some cases even the laws are old and not well defined.
I guess in the end it all comes down to moral conscience but more about pressure from higher up unfortunately. It's a difficult decision to keep your job or to do what's best for science.
Even some laws are heavily biased. (not speaking about the corruption)
-
How many of you were taught a model of transistor behavior, and then sent into the lab to do experiments to confirm this result? How often did a lab supervisor force you to a write up confirming the presented model, even if bad equipment, bad implementation of the tests, sloppy record keeping, noise or other problems made that confirmation problematic?
I didn´t experience that in my west-european university.
It is hard for me to conceive of undergraduate lab courses which don't enter with a strong bias on the models and results. Those courses are not generally intended to uncover new physics, but to provide some confidence that the coursework is based on reality and some exposure to measurement techniques and problems. If this is not how it is done in western Europe, well done to the educators.
The bias which seems endemic in US versions of lab courses is not surprising. The science behind the experiments is not controversial and is generally agreed upon. The experiments are chosen to avoid complexities beyond the theories which are supposed to be within the students grasp (for transistor experiments that means things like room temperature, modest current ranges, modest gains, modest frequencies, low slew rates and so on). But it often means torturing the data into submission when something like a non-opaque potting material shows up in the transistor, exposing photoelectric effects.
-
One of the major problems driving junk science is the institutional demand for large quantities of publications by faculty. Quality is rarely a consideration.
Well they were told to "publish or perish" so what do you expect? This is not so much about science but rather the human condition. This is more about the human need to succeed, to compete, and to try and be a winner.
That's reasonable to set a metric and evaluate people by that metric, but the trick is to define the metric such that maximizing it will indeed achieve the bottom line goal.
When you incentives people, they will find creative ways to increase their score.
I agree totaly. Hence my signature "measurement changes behavior".
Here is a recent article about metric that went wrong. A bank paid bonuses for new accounts so employees opened millions of accounts without customers knowing about it. https://www.bloomberg.com/view/articles/2016-09-09/wells-fargo-opened-a-couple-million-fake-accounts (https://www.bloomberg.com/view/articles/2016-09-09/wells-fargo-opened-a-couple-million-fake-accounts)
It also contain general wisdom regarding metrics:
Two basic principles of management, and regulation, and life, are:
1. You get what you measure.
2. The thing that you measure will get gamed.
Really that's just one principle: You get what you measure, but only exactly what you measure. There's no guarantee that you'll get the more general good thing that you thought you were approximately measuring. If you want hard workers and measure hours worked, you'll get a lot of workers surfing the internet until midnight.
-
"Here is a recent article about metric that went wrong"
True. But that's not the root cause. The issue here isn't just or even mostly due to employees gaming the new IC system. The issue is with the moronic managemnt that put that moronic IC system in place and failed to put a control framework around it.
Unfortunately, it is the shareholders and low level employees who ultimately paid the price, not senior management.
Zero accountability.
-
"Here is a recent article about metric that went wrong"
True. But that's not the root cause. The issue here isn't just or even mostly due to employees gaming the new IC system. The issue is with the moronic managemnt that put that moronic IC system in place and failed to put a control framework around it.
Unfortunately, it is the shareholders and low level employees who ultimately paid the price, not senior management.
Zero accountability.
I disagree that the employees that actually committed the fraud are not responsible, but will leave it there.
-
"Here is a recent article about metric that went wrong"
True. But that's not the root cause. The issue here isn't just or even mostly due to employees gaming the new IC system. The issue is with the moronic managemnt that put that moronic IC system in place and failed to put a control framework around it.
Unfortunately, it is the shareholders and low level employees who ultimately paid the price, not senior management.
Zero accountability.
I disagree that the employees that actually committed the fraud are not responsible, but will leave it there.
Well soldiers who needed to fight for their home country are also not responsible for the ideas of a lunatic.
Even if they didn't agree with it, they had to follow orders otherwise they would commit another type of crime.
Unless you're in a position high enough to have some kind of power to do otherwise.
-
Ehen i said that "The issue here isn't just or even mostly due to employees gaming the new IC system." I meant the following:
1. The employees are responsible.
2. The employees are not "not responsible".
So where did you disagree?
-
How many of you were taught a model of transistor behavior, and then sent into the lab to do experiments to confirm this result? How often did a lab supervisor force you to a write up confirming the presented model, even if bad equipment, bad implementation of the tests, sloppy record keeping, noise or other problems made that confirmation problematic?
I didn´t experience that in my west-european university.
It is hard for me to conceive of undergraduate lab courses which don't enter with a strong bias on the models and results. Those courses are not generally intended to uncover new physics, but to provide some confidence that the coursework is based on reality and some exposure to measurement techniques and problems. If this is not how it is done in western Europe, well done to the educators.
The bias which seems endemic in US versions of lab courses is not surprising. The science behind the experiments is not controversial and is generally agreed upon. The experiments are chosen to avoid complexities beyond the theories which are supposed to be within the students grasp (for transistor experiments that means things like room temperature, modest current ranges, modest gains, modest frequencies, low slew rates and so on). But it often means torturing the data into submission when something like a non-opaque potting material shows up in the transistor, exposing photoelectric effects.
Maybe on Belgian Universities don't make, i don't know other spanish universities if they do this, but on my university made this.
This made when at a practical lab had many problems*(almost all students) for obtained the result, so the great majority time ,the supervisor signed the lab part that you had could to do .After when you made the report you had to put of the lab part that hadn't done the theorical valor.
*An example: At practical lab, we had to test a switching power supply of tipology buck , so our group gave us a crushed diy board(Made in University) with the tracks and pads without copper, with tin gobs,etc. I fixed the tracks and the gobs but the board didn't work(too change the IC).
Until i began to suspect the transistor(the reference was hidden among components). When i discovered that the reference transistor didn't match with the documention.
I searched the datasheet and i found that it wasn't a transistor. IT WAS A SCR .
We told to surpervisor lab the problem .And he saw that we couldn't perform the practice by the transistor and remained few time lab. He signed us the lab after we had to put on the report all the theorical data
-
Ehen i said that "The issue here isn't just or even mostly due to employees gaming the new IC system." I meant the following:
1. The employees are responsible.
2. The employees are not "not responsible".
So where did you disagree?
I don't disagree with you two points here.
-
"Here is a recent article about metric that went wrong"
True. But that's not the root cause. The issue here isn't just or even mostly due to employees gaming the new IC system. The issue is with the moronic managemnt that put that moronic IC system in place and failed to put a control framework around it.
Unfortunately, it is the shareholders and low level employees who ultimately paid the price, not senior management.
Zero accountability.
I disagree that the employees that actually committed the fraud are not responsible, but will leave it there.
Well soldiers who needed to fight for their home country are also not responsible for the ideas of a lunatic.
Even if they didn't agree with it, they had to follow orders otherwise they would commit another type of crime.
Unless you're in a position high enough to have some kind of power to do otherwise.
Do you ask to german during WWII?
If german soldier negated to follow orders as shot the people or burning village ,they had all the numbers of be shot or be destined to Punishment Battalion as hunter-tank,
otherwise they will be accused the war crimes.
On resume,whatever they do are screwed
-
The same thing, blaming the little guys, is happening in the vw case. Given how pervasive the involvement is both in vw and in the industry, the issue there isn't simply a bad engineer. It is an issue of currupt culture and currupt management.
Yet, none of the big guys will go to jail. Instead, they will get paid handsomely for their services.
The same thing happened with the financial crisis, with our politicians as well.
As long as we keep the big gun pointed downward, we will continue to see this type of stories.