Another busy week at Retraction Watch. Here’s what was happening elsewhere:
- Could Nature‘s 2-year torrent of paper retractions be a good thing? asks Paul Knoepfler.
- “An Ecclesiastical Approach To Peer Review” in The Scientist: “do not abuse the single-blind peer review system in order to exact revenge or waylay a competitor.”
- “Replication and reputation: Whose career matters?” Dorothy Bishop on another criticism of how replication is done.
- Scientists “should be held to a higher standard of honesty than the average person,” says Richard Smith.
- “[I]nstitutions need to be held morally accountable along with scientists who commit misconduct,” argues Zubin Master.
- In the social sciences, “There is a strong relationship between the results of a study and whether it was published, a pattern indicative of publication bias,” according to a new study.
- Some open access advocates are not happy with Science‘s new open access journal, Science Advances. Here’s a Twitter discussion going into some related issues.
- A definition of “timeliness of retraction by the amount of time that passed” from another sphere: Immigration law.
- “A trade group representing the fast-growing electronic cigarette industry is contesting a new government study that claims the largely unregulated products lead teens on a path to tobacco cigarette smoking,” The Hill reports.
- “With righteous vigour, like the George Costanza of statistics, Oransky demolished the failures in reporting.” Writeups of three sessions at which Ivan spoke at the UK Conference of Science Journalists in June: On statistics, reproducibility in science, and how to find your voice as a blogger.
- “[J]ournals and granting agencies should require that authors report failures to reproduce their published results,” argue Helene Hill and Joel Pitt in an analysis of two papers they’ve been criticizing for a number of years.
- Is humor risky in science communication? asks Hilda Bastian.
- A victim of its own success? A dermatology journal says “evidence-based” section is no longer necessary.
- Climate scientist Michael Mann has responded to filings in a defamation suit he brought against the Competitive Enterprise Institute and the National Review.
- A misplaced decimal convinced scientists and parents that spinach had much more iron than it actually does — and the tale of the paper in question’s citations resonates today, says Kent Anderson.
- “Competition among open-access publishers is getting stronger and stronger,” writes Jeffrey Beall. “To survive, these publishers must invent new ways to stand out and attract more author fees from researchers.”
- Germany needs an Office of Research Integrity, says Alexander Lerchl (in German).
- Elise Andrew, the creator of I Fucking Love Science and a “science popularizer with a spotty attribution record, gets a pass from” the Columbia Journalism Review, says Nadia Drake.
- “We have to be aware of, and beware, organisations that masquerade as lobbying groups, which we see a lot in climate change,” says Nobel Prize winner Paul Nurse. “We have to be aware of politicians that cherry pick scientific views, even ministers who listen to scientists when it’s about GM crops and then ignore them when it’s about climate change.”
- Science profiles a scientist who is organizing other researchers to call “for governments to curtail these controversial studies [of new lab-generated flu viruses] until risks and benefits can be thoroughly weighed.” (subscription required)
“When they are serial offenders they should be crushed and buried” – Nurse’s phraseology isn’t encouraging for a healthy future of science communication. He might have been thinking of Atari video games, Nikita Krushchev or Mitt Romney. In all circumstances such an approach indicates failure and an inability to discuss things openly.
Quite.
But a “healthy future of science communication” depends critically on influential figures NOT “distort(ing) scientific evidence to support their own political, religious or ideological agendas,” or on these figures “misus(ing) science to support their preconceived beliefs.” Distorting and misusing data/evidence is a big reason why a sizable percentage of the U.S. thinks the earth was “created” 6000 years ago and that evolution is “just a theory.” Crush and bury away, Sir Paul!
Freeheeler – I am puzzled by your comment. Not sure how wide-eyed one has to be before dreaming that facts won’t anymore be distorted by this or that political group.
As shown by this web site, there is nothing better to fight corruption, distortion and mendacity than by putting it out in the open and discussing it freely. And this included being free to discuss it without any Soviet-style threat of being “crushed and buried” if the other side wins.
Science communication ought be better than that.
Omnologos, I guess the problem that Dr. Nurse encounters is that science misrepresentation seems repeatedly to gain an unfair advantage in the general media such that “open and free” discussion on scientific topics that have real or potential political consequences isn’t possible any any sort of balanced way.
Very occasionally we might be lucky enough that the science misrepresenters are “crushed and buried”, as happened for example as a result of the Intelligent Design attack on school teaching that was “crushed and buried” (at least for now!) as a result of the Dover trial and its aftermath.. another example is the belated judicial decisions against the then-S.-African-government and their appalling misrepresentation of AIDS science…
Unfortunately, we rarely get opportunities to present the science and address the misrepresentations in such a structured, open and coherent manner – science misrepresentation rather drips through the media and efforts to counter this seem to be rather futile in many cases – how nice it would be if there was a court of appeal in which a sanguine consideration of the evidence by an informed and disinterested “judiciary” could set the record straight on important matters of science! Until that happens it’s important that scientists and outer informed individuals speak out strongly against misrepresentation.
Of course this is very different from addressing scientific misrepresentation/fraud/error that occurs occasionally in the scientific literature which CAN be addressed in a rather structured manner, such that the scientific record is corrected absolutely and unequivocably (through corrections/clarification/retraction).
I don’t understand the reason for the implication that I am wide-eyed and dreaming. How are you reaching that conclusion from my comment?
I am not suggesting that anyone’s ideas should be suppressed. Actually, omnologos, I agree with you completely that any/all ideas should be put out there in the open and discussed freely, even if those ideas are a deliberate distortion of evidence. Perhaps where we disagree is with what should happen next: if the ideas put out there reflect a deliberate distortion/misrepresentation of evidence, then those ideas should be crushed and buried.
But you are perhaps taking Sir Paul’s crush-and-bury metaphor too literally. Based on the article, his crush-and-bury approach is, in responding to serial distorters, “to call offenders out in the media and challenge them in the strongest way possible.” Are you suggesting that this indicates a “failure and an inability to discuss things openly”? Because it seems very reasonable to me. And note that Sir Paul advocates this only if the forging of relationships fails to dissuade influential people from misrepresenting evidence.
Actually, I am still waiting for when Sir Paul will decide to bring the climate debate in the open and challenge the “misrepresenters” in a public way. So far nothing of the sort has materialized, and the one opportunity has been held behind closed doors as requested by Sir Paul and/or the Royal Society.
What is worrying is that he’s now moved instead to even stronger-sounding words. How can that “communicate” science?
Climate science is very much “in the open” omnologos. The problem (I agree with Dr Nurse) is that there are widespread misrepresentations of the science in the media (and especially on the internet much of which is a veritable industry of misrepresentation!); not just of climate science..
Not sure what you mean by “debate”! Science isn’t communicated by “debate” – it’s assessed and communicated by faithful recourse to evidence. In the scientific literature, flawed science/misrepresentation is addressed rather formally by processes of peer-review, post-publication peer review, corrections, retractions and so on. Unfortunately outside of the judicial reviews of instances of science misrepresentation that impact societal well-being (see examples in my post above) this doesn’t happen in the general case and science misrepresentation gets a considerable airing in the media sadly – “debates” don’t help, although they’re obviously central to issues of policymaking in response to scientific evidence
Paul Nurse isn’t a climate scientist and it’s not up to him to “debate” the issues of the subject. However part of his role as President of the Royal Society is to attempt that scientific matters are faithfully represented to policymakers – so his highlighting of the issues of science misrepresentation is part of his role and in my opinion one that he has the authority to pursue productively. One might note, btw, that in response to a request by the Royal Society to ExxonMobil some yeas ago to cease funding climate science misrepresentation, that company made undertakings to do so – so every effort at chipping away at misrepresentation is valuable – wouldn’t you agree?
You ask how Paul Nurse “can communicate science”. The answer surely is that he communicates science by being an excellent science communicator! Check out some of the YouTube vids!
Chris – if what you said were all there were about the latest from Sir Paul, what is the meaning then of his call for “misrepresenters” to be “crushed and buried”?
Crushed and buried by whom, exactly? You seem to imply, not by Sir Paul, since he is not a climate scientist.
Also…crushed and buried how, exactly? The ID types, like the astrologers and the homeopaths (and the ESP aficionados), enjoyed their moments in the public eye at various times in the past decades. Their ideas were taken down by and through very public debates. Just look at the reactions to Velikovsky’s book. Sagan himself IIRC came to realize you don’t counter pseudoscience by retreating into a tower and asking people to just believe what you say, and disbelieve what the other guy is saying.
If some people are misrepresenting climate science, why can’t the same be done? With the intellectual might of the Royal Society, what could stop science from winning?
Despite all the might of just about all major scientific organizations in the US and several court cases,
about half of all Americans still believe evolution is “just a theory” and prefer to believe in creationism. Never underestimate the ability of people to reject facts that do not fit their view of the world. Science will not win, but physical reality will.
By the way, what “very public debates” have we had with ID proponents and scientists? To the best of my knowledge ID has mainly lost in court cases, not because public opinion changed.
Paul Nurse is an outstanding science communicator (plenty of examples on YouTube for example) and one could hardly wish for a scientist more willing and able to “discuss things openly”. However he clearly is appalled by the dreary efforts to misrepresent science in support of dubious agendas; he should be applauded in my opinion for speaking out about science misrepresentation. After all precedent informs us that misrepresentation of science can have dreadful consequences.
Incidentally for an astonishing and rather amusing and touching account of Dr Nurse’s discovery of his own genetic background, this is a good view:
this is interesting! how about those pedigree analysis and family tree studies for gene discovery – i am not talking about Sir Nurse’s case…in general.
Thank you so much for the link to the YouTube document. Obviously the man who is speaking is not the Nobel prize winner, but rather Paul (Maxime) Nurse, and the result is fascinating, both deep and funny (“I know the homeland security has high standards…”).
As a complement I would recommend his BBC radio interview released 3 years ago. His account about his first research experiment dealing with the respiration of fish eggs is hilarious… and instructive.
http://www.bbc.co.uk/programmes/b015n3b7
The BBC arranged a nice transcript, which I found very useful:
http://downloads.bbc.co.uk/radio4/transcripts/lifescientific_paulnurse.rtf
What we didn’t say in our paper (Hill and Pitt Failure to Replicate: A Sign of Scientific Misconduct?) but I will say it now, this is an indictment of the referees of the two manuscripts and of the NIH Study Section reviewers (and of myself, before I woke up). We all should have realized that there was something wrong with the kinetics of killing of the Chinese hamster cells by tritiated thymidine under the conditions of the experiments in the reports. This required a knowledge of the very early experiments with tritiated thymidine and I have to admit that it took me a while to find those, the earliest of which dates back to 1959 when most of you who are reading this were not yet born. It’s important to know the literature in the field.
I don’t know how the cell numbers were measured, as your paper does not give details, but this is highly important, Typically 2-5% of cells in culture are dead/dying anyway; the number depends on the criteria and method used.
http://www.mdpi.com/2304-6775/2/3/71
Now it is impossible to have a good estimation of living cell numbers over three orders of magnitude. An assay that is tuned to measure live:dead/dying cells as 100 cells/area will have quite different properties at 1,000 cells/area and 10 cells/area. If you are good at 1,000-50 cells/area you will be lousy at 50-1 cells/area and vise versa.
Different people count cells, labeled with a vital dye, differently; deciding what is dead and dying can be a tough call when you have performed a toxic titration. Anyone can see the difference between a healthy and apoptotic cell, but if you have a population of poisoned cells, you have healthy, unhealthy, pre-apoptotic, pre-necrotic, apoptotic and necrotic.
You can account for all the difference between ‘A’ and the others if ‘A’ had a difference criteria for defining ‘living’ cells; if ‘A’ defines unhappy/unhealthy cells as non-viable then he will generate different curves than the others
You also have to acknowledge that although people follow the same protocol, they do not do the same things. When I trypsinize my cells, I am very gentle mechanically and incubate for a longish time. My technician adds trypsin and after a minute wacks the side of the flask with a karate chop. We both think we are doing the right thing and following the same protocol.
Now perhaps it would be good if you did a little review of the literature and looked at the number of papers that actually state the algorithm they used to define ‘living’ and ‘dead’ cells when using
We did not count any cells or colonies ourselves, but only analyzed patterns that appeared in the numbers that were recorded in the notebooks. As for the survivals, differences that are as much as 70- to 100-fold apart seem hardly likely to be due to slight unrecorded changes in protocol such as you describe. Our point is that the kinetics of cell death in the hands of B and C are substantially different from A’s kinetics, B’s and C’s results are predicted based on radiobiological considerations, A’s are not. B’s and C’s numbers are consistent with random distributions, A’s are not.
It would be helpful if you could give an idea of what the numbers, across three orders of magnitude, were
[X]=0; 10,346+/-1,121 SD total 659+/-59 SD dead
[X]=10; 4,367+/-485 SD 1,897+/-46 SD dead
if we don’t know how it was done or how the dead/dying changed, ‘A’ or “B,C’ could be right.
If you are counting by microscopy then you need cells that are not overlapping and at maximum density the nuclei occupy 30% of the field; say you have an average of 1,050 cells in total and 1,000 are healthy and 50 dead or dying.
Add ‘X’ and wait.
Individual A counts 500 total cells and 200 dead and dying, so ‘B’ states there are 30% cells surviving.
Individual B counts 500 total cells and 0 dead and dying, so ‘B’ states there are 50% cells surviving.
You are assuming that ‘A’ is lying and and ‘B’ and ‘C’ are truthful. However, as we don’t know how the people interpreted numbers we cannot know, and an appeal to authority ‘based on radiobiological considerations’ is unhelpful.
Richard Smith somehow (unfortunately) seems to consider scientists to be above other “average” human beings. Most scientists I know slog away day and night in the lab and are exactly the same as other human beings that surround them. The key question is, should politicians, journalists, corporations and bankers be held more accountable than scientists and hold a higher standard of honesty given their much higher level of assigned responsibilities in society and the economy?
It is not a question of counting live, dead or dying cells but of counting colonies that can only arise from surviving (live) cells. Perhaps that makes things clearer. The protocols and the numbers that we analyzed are all available in the supplementary files on the Publications website. If you want to know more, you will find my email address on my website; http://www.helenezhill.com.
“t is not a question of counting live, dead or dying cells but of counting colonies that can only arise from surviving (live) cells.”
I have just looked at the supplement. They used the dilution method they used with >200 steps required for each determination.
Look at B’s Dec 18 2000 counts.
The fifth read gave 1511, but ‘B’ thought this low and so did a recount, got 2520, liked this number and kept it.
The seventeenth read gave 3534, but ‘B’ thought this low and so did a recount, got 4822, liked this number and kept it.
The thirtieth read gave 3877, but ‘B’ thought this low and so did a recount, got 4013, liked this number and kept it.
So ‘B’ changed 1 datapoint in 10, but all the changes were asymmetric, with ‘B’ discarding a low numbers and cherrypicking a number 67%, 37% and 3.6% larger than the discarded value.
‘C’s 10/1/01 appears to be a ‘neat’ copy of a raw data sheet.
‘B’ and ‘C’ are also using backgrounds that they are subtracting, whereas ‘A’ does not.
1. I don’t understand your reference to >200 steps for each determination. The protocols each have about 40 steps to them.
2. B seemed to be having a problem with the Coulter counter on Dec 18 but the rest of his/her counts for the 4 other 100% experiments seem to be OK and his terminal digits are random
3. C’s 10/1/01 probably is a ‘neat’ rendering of the numbers. The Coulter counter was not attached to a printer so the numbers were copied out from the screen by hand.
4. The backgrounds are very small relative to the magnitude of the counts and, although they were recorded, there is no indication that they were subtracted
And let me say further that there are no assumptions made about anyone lying or telling the truth. It is only a question of interpretation of the data and analysis of the numbers.