DeSciNYC

DeSciNYC

← Back to videos

Talk & transcript

What is the plague in academic publishing?

Dr. Mu Yang

Transcript

Download text

Automatically generated transcript. May contain errors. Select a timestamp to watch that moment on YouTube.

0:04

editors at the same time saying look at look at this repetitions and then two people decided ah I'm going to swap it out with something else it's fine so that is really terrible because that is it's it's like money laundering it's a fraud laundering it's even worse than publishing fraud because the first time around you're negligent you didn't pay attention you didn't do your job second time around you still didn't look into it you know so I think I think the bad correction is terrifying and terrible um this is this is also so the

0:49

double double-sided error point to the traces that are identical kind of hard to see so this the single-sided point you that Tracer there identical within each figure and then these are connecting traces are being reused AC cross figures so you can see this within paper and across paper reads of of traces what you do and then decided next slide please they decided to publish a new one it's a correction it's absolutely unbelievable um right next

1:30

like so okay we um so where we're getting into the part where we're going to start blaming at capitalism so you get you got just so that I'm not spreading this information so you can you can take out your phone and Google lcv profit margin oh wait anybody want to Google LC coffee margin yeah so that's a very good idea yeah that's that's insanely high it's way higher

2:19

than any of the notorious company that that that um would typically call them out for like over 30% profit margin so and lsee is one of the one of the publisher that I battle with the most because they published so many journals and then there's so many journals are absolutely completely polluted with uh paper mill stuff um so what we can do I think what we can do is very critically important so it's very the thing I hear the most

3:00

is like yeah maybe there just some some something in the methods I didn't read maybe I just don't understand their their technique there's always fancy techniques no a lot of times if you have done an experiment a lot of times your God feeling is probably right do not ignore your own God feeling and say I did what do I know you know maybe you don't know but it's always reasonable to doubt and and um PP here as I mentioned before is very uh important to to um use and then you can always re uh report to AIC Department of

3:43

Publishers um all right next slide please so okay uh wait okay here here we go so I think I don't know how many people are still in graduate school how many people are are post off go go back a little bit go back sorry yeah um but it's important to identify because when we see frosters uh we don't want to we don't want to get into like a go oh my God look at that guy he's so terrible or like oh yeah the paper meals are only by crazy people I I think I think it's it's important to to understand

4:25

that if you're in science every single every single of us will be at some point under the pressure of generating pretty or prettier data you know you're like okay what do I present in my next uh lab meeting wh why is my data always so so ugly somebody else is reporting something you know foreigners uh like myself one used to be okay I need my Visa sponsored I need to get my one be you know um and a lot of times the pi because I work at Colombia I work with a

5:07

lot of research lives I hear a lot of Pi goes like this okay if I don't get this Grand I don't think I can support you next year that sentence is usually followed by here's what I really want for that Grant nobody is stupid you don't have to tell tell people hey can you go fix something or you have to put pressure first and then your desire result second and I'm probably put a lofty a to that will be really wonderful if we develop a drug treatment for Alzheimer's using this fantastic animal

5:50

model that has been replicated so many times you know uh so so this strategy really works on the V vulnerable folks so by vulnerable meaning like you need your degree you need your paper you need your tenure you need your Visa sponsored so what I'm trying to say is that when we look at the frost it's like oh that guy is so but is in is the vulnerability is in every single one of us right so it's it's really more important to identify the pressure that from external sources is like your piis

6:32

your bosses and then from internal I really need this paper I rating need whatever none of them are wrong goals but like which drugs are going to work in people which drugs the FDA is going to approve how much people are going to sleep each night like um what people want to bet on that yay okay good good okay so I want to explain how that works in five minutes or less next slide so um why I started it is because I think what like we're going to hear a really interesting I think this is like um the talk that we're about to hear is going to be really interesting because that's actually like part of the reason we

7:12

started D NYC was uh I thought Academia I was doing a PhD before and I thought Academia was pretty bad and it incentivized like the wrong things so um I was looking for ways to like uh create a world in which like Academia was like uh more incentivized with the correct with incentives and we're seeing here you know uh Ellie elazer masel or Stanford they all had people who were basically falsifying data and um they were making things up and it cost it created like billions and billions of dollars of waste all right next slide so um you know with poly Market

7:53

what happened is is like um people were looking into pre prior to poly Market everyone thought that the expert were like these pundit on CNN or Fox News but when py Market came about people realized that actually like there's a lot of really good people predicting who's going to be the president who weren't necessarily quote unquote experts and the reason that they were able to predict this is because they were like given the correct incentive so instead of um being incentivized with popularity or Prestige as a news show um they were incentivized with money so people would bet and if they were correct they would actually earn money and when when when they were betting with their own money it caused people to be very very careful with who they're betting for and actually do a lot more

8:34

research and I was thinking I was like oh that would be like a really cool idea to have like a similar type of style thing where instead of like a peer review system where people were just betting on which project would be which paper was going to be successful in the future and it was all Anonymous and they weren't paid if you actually tried to incentivize it with uh connecting the dots between uh who who is paid based on how good of our prediction they were prediction of what was a good paper or not um you'd have people looking over and peer reviewing papers much more carefully and this would create a much higher um a much better type of Science and it would also pay the people who deserve to be paid all right next

9:14

slide so we wanted to create a prediction market for science next slide and um I don't like being up here so I want to go through these quickly um next slide so we have like scientific data we have predi uh prediction markets and we have like a social network feature where people so the idea is to create like an ecosystem where people are basically paid to curate each other's data and paid to like and paid to make predictions on which type of data is going to be accurate uh next slide so uh it all starts with like um the data source so we have a variety of data sources we have like wearables from like the aura ring or like your scale or

9:57

um a clinical research organization or a lab test and these are the things that are fed into the system so we have like these real-time data points that are always coming into the system um next slide um and then once we have these real-time data points uh we can create these competitions so um currently we have uh we're doing studies on like intermittent fasting so we're seeing how intermittent fasting affects people's sleep and so people enroll in a study and then they go They're randomly assigned to either control or the um the the a not control uh study and we get to learn okay this is actually this is this this does or intermittent fasting does

10:37

is or isn't a good idea or we have one on hair racing where we're working with a hair D to like they're testing different types of drugs and seeing which which which drug uh causes hair to grow faster and people can like bet on that in real time we also have one Brian Johnson is very he spoke previously at D NYC and he's like doing a lot to like gamify and create metrics around every part of his body so like every so often he creates um uh these he releases this like uh uh form that has like maybe a hundred metrics of like how much he's able to bench press uh how much he's able to sleep what his uh resting heart rate is everything including like his um sperm count and so this gets a lot of publicity so we're creating like a a

11:19

market around like Brian Johnson sperm where you can bet if it's going to be better or worse than random than what he previously had all right next slide um we also have markets around like vitamin and sleep we have markets around the FDA so you can bet on which drugs are going to be approved or not approved by the FDA I think this is going to be like a very interesting one and you can bet on every part of Brian Johnson's body so that every time that he releases a thing you can see if uh it'll if you can guess where his Biometrics are going to go if they're going to go up or they're going to go down in like across like a hundred different categories all right next slide uh next slide uh next slide next slide sorry S I should have

11:59

taken some of these out next slide all right uh next slide it's fine uh next slide or we'll just end next slide next slide oh next slide this is a deck okay so I think we're really changing how research is done um we're changing like uh we're making research happen faster we're we're having more eyes on on the data sets themselves and um we're incentivizing it with like new new modes of incentives so hopefully in the future you know you're not dumped this like 10-page research project uh as a paper or PDF and you have to figure out like where's the error um but actually like uh research will happen in more real time and um people will be paid to like find bugs in it and uh or

12:42

accurately predict if it's if it's correct and I think this is sort of how how science is going to go in the very near future cool all right um I think was there anything else next one one more is that it oh D NYC U thank you all for coming we have a lot of people coming to this next one we' had a lot of speakers next slide um here's all the here's some of here's just a handful of the speakers we've had um next slide um oh that's all next previous slide all right so um is there any questions any questions yeah go ahead how do you know that this is can we place a bet on whether or not this is g to make science 100 times faster sure we

13:25

can create a market for that absolutely um is there another side oh anyone else any else what is D side cool oh wait let me get a picture of everyone before we go okay we got a panorama smile so guard all right one more all right smile cheers okay cool all right so um without further Ado um I would like to welcome our main speaker um let me um she was the uh she

14:20

was the head sleuth for this amazing topic on trying to figure out um if there was there was fraud going on and she was actually recommended by her friend um to come to speak at dyc so I reached out on Twitter and she was very kind to uh accept the invitation so and I think it really fits in well with like the ethos of DC so without further Ado I would like to uh welcome Professor or Dr [Applause] moo thank you so much for having me based on the previous speakers would you

15:02

bet this one is going to be a female that's a good one like the movie they make you watch so many previews another one okay where's my slid is a wall

15:41

this is my SL anyway so um proper space my my social handle is actually Mo Mouse too so we're at Star Mouse it's easy to remember M Chinese we like one double thing you know like tendas is always so mum Mouse too is my social handle but you know actually assistant professor at Columbia whatever yeah I mean you can read I don't have to read it for you right so

16:29

next slide you guys are drinking so I'm might as well you know so okay so this is this is important because just to show you that um my personal history is important to understand why am I doing what I'm doing because I followed a very typical asan kid you know like uh trajectory you know you come to the state you go to graduate school you do your get your PhD you do your post and and then I was I was kind of smart so halfway I dropped off I was like this is this is not for me so 2016

17:13

I came to Columbia to run this course and this cor is basically what share facility where your your half service half Academia things um yeah and and then so in terms of publishing I publish like 60 plus citation number you know people C this kind of stuff it's important to to to to see to see the context where I'm coming from because I'm coming from a very typical um background of of you know like uh anyway next slide so you can just click a whole bunch of um yeah so go back a little bit

17:55

so so basically we're going to talk talk about my first first case how I started my biggest case and then what I'm working on now which is like paper meal stuff and I really want to talk about um we don't want to shame specific persons so these people are examples of the bad toxic cultures is what I'm trying to talk about not not the yeah everything is talk I know man um right yeah so but I think is a parish or uh publish or Parish cult is and then the publishing Industries um it's a lot of problem in that so next slid you can

18:35

just click next slide and then you can yeah so so more basically clay go back a little bit uh we're going to talk about personal stories B cases uh publish your par is the you can read and U capitalism sucks and um power to the little people things like that next slide uh so retractions I have my reports has caused have caused 150 retractions I know the intro says 130 but that's how quickly it went up this uh this year so from what you saw that to now it's like

19:18

20 more but my first retraction took me two years probably more than two years um but now people like they they know you then they yeah so anyway this is just a next slide please so and then you can yeah uh that's good so uh I want to blame the public or Parish culture I think this is probably the root cause a lot of problems that we have um because you say if you're always chasing Grand chasing papers then the Publishers and the froster will take advantage of what you

19:59

want here and we also want to address Publishers because publisher makes a lot of monies uh the Publishers profit margin is way bigger than Walmart and all that so um peer review system Michael Michael mentioned that um paper meal paper meal is a very huge problem but paper meals are taking advantage of the publish and Parish culture so a lot of these are op opportun component to that is subsidiary to the bigger problem that is a uh publish or Parish next slide

20:39

please so this is a really old New Yorker uh Cartoon that is basically said basically shows the kind of pressure that you know acade academics are facing I thought it was kind of uh appropriate still 20 years later next slide please so notable cases uh last year I got an entire book retracted I it it didn't really take a lot of time because it just it's um yeah but I wonder like the

21:19

book was selling on on Amazon at the time for like over $200 I was like how much money did you make I'm like you know um so I wonder if tell libraries uh who bought the the book and whether they you know when you we you have a contaminated chicken you you you know you tell customers to eat that where it's like do you tell people who recall right that's where I looking for do you recall your book I don't know do you refund people's money I don't know but next slide please and then this is in retraction was so this is another note I I kind of ha a whole bunch of papers from one

22:00

Journal um so next slide please so my first C next SL so my first case was I still work here so let's be careful so what is it where happened so it was kind of like you know I I I I um so you know the host name that's not that's not show it here but this guy he came came to Columbia to give a talk and that guy is who who told me that this guy was giving a talk so which is when I went to the talk and asked them questions not well

22:41

received and I was like okay I will see you so next slide um so that's February 2020 you saw that this is June 2024 fast forward more than four years took me more than four years to put this guy in the Highlight as uh you know one of the guy who are problematic in Alzheimer research so so at the beginning phase of being a SL was extremely lonely and extremely hard nobody listen to you hit wall everywhere you're literally a little person like you know but it's took me really four

23:23

years to get to where um you he still has his job by the way to too so next slide please and then so okay we're going to talk about this guy so because the alimer research is all about spatial learning and spatial learning as some of some of you may know it's basic using uh the hippocampus structure in your brain so the hippocampus is very critical for spatial memories and how do we test that the mice you basically put them in a so this guy Richard Morris invented this thing called The Morris water is basically like a tub with opaque water and then you put a hidden platform in it

24:04

and then you train a m of R to find a hidden platform and over time you use like uh Landmark actually people do that in their Labs putting these kind of things to to give Myer like a special special cues to to navigate with anyway this is this is a this is a task that used by a lot of people who said Alzheimer's next slide please and so this is basically your the this is my setup this my data so is basically a wrong with a tub and then you train a out to swim and then the data so this is a typical learning data you know over time the M get faster and

24:47

faster but you see a lot of variabilities within group and across days next slide please next slide please um so here are some other people's data typically you can see that there's a lot of variability which is yeah my swimming a big top of P Water of course iability what you think um yeah and so next slide please uh one more one more sorry one more okay so this is what caught my eye right this is the same mice so this the Alzheimer my swimming a big top of op

25:30

water and then these are very old mice um so we're like where are the air bar sometimes people don't show air bar which is which can be fine but you can see some token air bar there represent like but and you say what what are the rest and then it's completely linear you know there learning is not like a completely linear process especially M that are naturally not natural swimmers so this is very bizarre to me next slide please uh one more so okay that's good

26:10

that's go what next slide so here's another here's the next study uh you know from the same guy from the Temple University you can this is not three groups by the way this is uh this is four because the two control so WT stand for well type which is s your control animals so these are control animals with and without uh Alzheimer drugs and these are the Alzheimer transgenic animals with and without drugs as you can see the well TI with without drug are completely overlapping meaning that their performance are completely identical and linear and perfect again you have barely any Arrow

26:54

bars so Richard moris remember the guy who invented the test I asked him to comment on it and his comment was you know this is like very strangely linear and he didn't believe it next slide please so again so this is uh personal communication this is what he told me um when I wrote him um he said this require suspension of this believe and then so so the the the scary part is that when I thought okay I got rich moris to back out my statement man you're probably dead no he's not he's

27:36

like nobody listen to me or Richard Morris that's how that's a Hu race and Cavalier um Stone Walling that we're facing next slide please yeah Stone Walling so as a as a first time SL or first time uh whistleblower nobody listen to me like I reported to office for research Integrity I reported to journals uh you guys laugh at like how how can how can like learning data be so perfect yeah I thought so too but absolutely nobody listen everybody was just like oh did you talk to that guy

28:16

yourself I'm like oh yeah if somebody run me over would you say that as well go run him over I'm like what so but that's exactly this the the kind of res respons you got from Journal editors and even um office for research Integrity which is run by the government um next slide please so and then I got a help from from other senior SCH these are still people that I I like very Dr Elizabeth be and then Kevin Patrick these are these are St that are uh they came way before me and then they came to um they

28:57

land ahe and which made a huge difference next slide please so what they found can everybody see okay so what they were able to find is that so the same color uh boxes indicate overlapping images so you can see that one and then this one so you can see the same color boxes the the images are identical but um but this is the Alzheimer Mouse this is the Alzheimer with uh drug treatment so they're completely different experimental conditions but the the images are reused so make your question

29:37

whether like is any of this ever real so you know cuz there is a component of of you know CU we all clean up images and data to some point sometime it's like putting on makeup maybe even putting fuelers but sometimes it's identity Stu right but at some point you don't know where where this is what it is is right so okay next slide please so this is uh this is the electrophysiology data that I caught in four studies so I mean this is super tiny to see but the blue box is

30:19

basically indicate identical data in four studies May if you look at the um ring boxes they're all like super similar as well next slide please uh one more so and I did this for during the um one of our project I did this summary so it's kind of hard to see but each uh each circle is a paper so the overlapping part shows that uh the two paper or three paper share images so this is the extent of overlapping across as many papers just

31:01

look at how many circles there are and then look at how many paper overlap and then when they overlap the images are always used to indicate different experimental condition not the same because otherwise it will be um not a problem so you can see the the it's really what what I found was funny Mouse swimming a top data but this is what came out is the whole um is you know yeah so it's um is pretty overwhelming to look at next slide please so this guy had n retractions by

31:42

now oh here's a funny part he he fed a lawsuit and against his graduate student and then so but that guy only contributed to four of our retractions so what happened to the others uh um so next slide please and then he lost the lawsuit I mean and then they said considering such allegation practical at least should have known through the exercise or reasonable diligence about the Alle falsified of you know basically he was told there was some problem way

32:23

before I I reported so anyway it was embarrassing don't throw little people under the bus um next slide please uh this is my biggest case so this is uh this guy used to be the Director of National Institute of aing Nia so ni is I think the largest founding Agency for Alzheimer research globally I mean may be wrong but some somewhere in there but in his Department of neuros science department he was in

33:04

charge of 2.6 billion dollar of gr funding so he had huge influence I didn't get anything so um anyway this guy had 800 papers published andway flagged a over 130 so the collateral damage is that we have over 500 co-authors impacted and um and then something like 238 patent was was supported by his by his paper so people using his research paper to support pent application so there's a lot of drug development all that stuff and this guy had 22 papers um

33:49

this thing called cable license Ser license is this something from Russia and it came from p pig brain juice so ask me anyway it was problematic uh next slide please so okay so in then uh June of 2023 I was recruited by the science writer Charles pillar um to join the the sloes I work with Elizabeth big and P Kevin Patrick and Matthew shk who reported a lot of Alzheimer's scandals I was I was lucky to be recruited to join

34:31

them to do a lot of Investigation on the alimer related research scandals and then the book just came out and or will be coming out in February 4th I got my own copy yesterday um but so this is basically where I um how I found in the musia case next slide please uh just um just uh a disclosure I do use and then so do other other slues we we do use an AI is is pricey but for us cuz we start using using it during beta so we use it for free so nobody is

35:12

genius nobody has Magic Magical one we use Ai and is great uh next slide please so this is one again so whenever you see same color boxes look for basically images that are identical or near identical and I usually these are from different different experimental conditions next like um so you see the Yellow Boxes and the one of it the color got changed to be a little bit lighter so I don't know if somebody thought about it and decided to um change change things a little bit it's not good uh next slide please oh

35:54

this is very not good because you can see that uh this is non TG means this is a control control animals and then that is um that is a wiz drug with two type of drugs so you're mixing controls and and drug treatment drug treated animals next slide uh this is so so using AI is not just about it's a little more than feeding PDF into AI tools I had to do some I'm going to boring you with the detail but I had to do do some work to get AI to figure this out but you can see the cloned

36:35

sections um not sure why would somebody do that so this is this is this is why I think a lot of things are like makeups and fers but you don't know when they stop doing that that's a problem right because there's really nothing wrong with booty on makeup but at what point did you did you not even not use your own histology you know how do I even know that this is your experiment right so that's that's that's the the encroachment of trust that that this kind of thing uh generated next

37:16

slide uh same this is this is also like a lot of Clone sections as you can see uh again because when you look at the when you you look at it's it's in one of those who who even look at those and then what information do you get from those is like so which which Mak me think some of these some of these may may well be makeup of fers but but I just don't trust that this is your histology anymore you know it's like did you where did you get this thing right

37:56

so next slide please next slide please uh say next slide we saw the same uh this is a particularly bad one because you can see these are these are actually two overlapping images um but then so these little dots are um plaques or path pathological indications that um that Alzheimer or Parkinson's might develop like Alan or plaques so these are neuropathologies these are missing in this one so using using this

38:40

dots as reference uh can you see this this little thing is missing in that one and then those two dots with a red circle they're not here so so that that one is not makeup that one is identi set because you're putting neur pathology on a slid that doesn't have neuropathology on it so that is a terrible one so this one makes the previous one just like okay if you can do this H how do I know where to

39:20

trust you right so that's that's a problem uh next slide please so this is also you can see that oh I see that um so again so vehicle animals and then that's anyway same overlapping next slide please oh this is a good one you can see this is this is the Alzheimer Mouse model with zero licing treatment this one is a control control of course we don't have pathology in your brain and but if you can see this

40:01

is two identical slides this one this one this one we got the pathology this one we don't get the pathology wow yes all right next slide uh this one is just a bit basically it's a mouse and then rats they have the same image so W work that way next slide please um this one is also bad because this is non-transgender meaning control Animal Control animal vehicle meaning

40:42

control animal uh with just uh placeable that one is Alzheimer with treatment so of course you want your Alzheimer with treatment to behave or look more like controls because that shows your drug worked but you cannot just use the same slides and show that that's how they work that's how work all right next slide uh same more of the same story so next slide so this is just to completely so that's a 2014 and 2016 two two papers from two different years uh having the

41:23

same images so making it makes you think that do you guys have like a library Archive of images that you just pick from and I don't know okay next one this one okay this is is not in NIH doca I found them after wait way reported to an you can see some of the corner clones next light uh this one is a little funny because you can probably see there's some kind of rectangular shape around this thing this is probably in the cells you don't see a lot of rectangular things uh next

42:05

slide uh I don't know how many people know how to do Western blog but this is you can see the copy and

42:21

paste uh yeah it's super little attack yeah cuz but if you see the original image is like why just okay next one so this I also found this one after then I and I report next one so so I enlarged that one just to show you how so I use this free software of forensic you can use all sorts of image tools but you can see these are identical uh look at not only the band but also the peripheral roll up and down so you can see there identicals and then so are

43:03

these um next slide so so you can check on pop here to see just typing the uh whoever's name eler to check all the record next slide so um I like AI tools but I I feel like it doesn't really prove me particular observant because everybody can just say you're just someone with the free AI tools so uh last year someone report like a whistleblower from Saudi Arabia got me into looking at chemistry papers even

43:46

though I don't know my chemistry God uh so I I mentioned I use free software called forensically next slide uses one called research rabbit it's pretty cool it generates the connections between so if you just typing papers or uh dii numbers you can see the connection between uh because this is particularly use for for paper paper meals because you can see how people they collab as well frosters you can see how they collab okay so I like these kind of papers because AI for some reason AI can

44:29

figure out the histology stuff very complicated stuff but AI cannot figure out squiggles like this it's it's I beat AI in this front like every single day and it just it's very satisfying like that that I can prove that I can use my eyes uh to do better than AI so which is why I decided to switch from the Alzheimer's stuff to to do this chemetry stuff which I call them squiggle cuz I don't know what Al they're called but you can see all these five traces are the same uh and next slide same this is all

45:10

um same same noises next uh this is fun I mean I I know that we have chemist people here and you can just appreciate the the going Westward Movement of our squiggles it's a very free hand yeah but they were published so um and then these are two uh identical edx are indicating different materials in two different papers uh next one oh this one is fun cuz you can

45:55

see uh if you look closely it was spliced in in the middle part see that splice here and then the other Green Arrow you can see the other end um but the the thing is that if you look at if you look at that edx in a regular paper in a from a regular distance unless you zoom it in you probably wouldn't see it because we're looking at this like you know microscopically so if you're just like looking at the paper on your phone or something you probably would not see it uh next one yeah okay so these are

46:37

two Ed you line them up you can see that the only difference the only different part is here every other parts so one thing I didn't know about this kind of stuff is that the the noises should not repeat um but you can see the even Peaks um are pretty much identical next one so these are also so these are supposed to indicate different materials um and the noises are are the same next

47:18

one oh this one is funny I I do know that that pigs should not be stuffed from underneath anyway so uh QR code so uh this is this is a QR code for the for article I wrote about Environmental Research I reported 80 papers and then they already retracted 14 um so you can you can read about that um the next one please so oh usually I don't get into the science

47:59

part because I really don't understand anything about anything um but I did I did I did learn how to check this chart thing so I know that you can look at that stuff and look at where things are and then so this one this one was retractive because the Titan titanium pig doesn't should not be there so I mean that tells you how basic it is if a mouse is a mouse Behavior person can can learn how to check the chart I mean like that's how basic it is what why was the reviewer not checking the chart I didn't even understand the signs

48:41

behind it like some something about counting accer emission something but there's a chart that you can refer to that's how youy it is uh this is also the retracted because the squiggles are the identical uh next one this is a weird number I was remember we got a lucky number just very very strongly prefer 577 uh so okay Corrections so this is another article I wrote about bad Corrections because if you think the journals are

49:23

terrible for publishing papers with fraudulent data they're even worse when it comes to correcting fraudulent data because a lot of times the editor is like nah it's so annoying and then they will they they're obligated to contact the authors and authors say oh we're terribly sorry this is a mix up here is a correct image many many very often the journal editor will be like oh sure we'll publish a correction thank you for sending like did did you look into that like wait no but they do that all the time it's very

50:04

terrible next slide please next slide yeah so these are four paper that I reported you can see the pink box indicate a a segment that is that is that is yearly um repetitive over across those images but two papers I wrote All editors at the same time saying look at look at these repetitions and then two people decided ah I'm going to swap it out with something else it's fine so that is really terrible because that is it's it's like money laundering it's a fraud laundering it's even worse than

50:47

publishing fraud because the first time around you're negligent you didn't pay attention you didn't do your job second time around you you still didn't look into it you know so I think I think the bad correction is is terrifying and terrible um this is this is also so the double double sided eror point to the traces that are identical kind of hard to see so this the single sided point to the Trac that they identical within each figure and then these are connecting traces are being reused or across

51:29

figures so you can see this within paper and across paper reused of of traces what you do and then decided next slide please they decided to publish a new one it's a correction it's absolutely unbelievable um all right next slide so okay we um so where we're getting to the part where we're going to start blaming a capitalism so you can you can just said that I'm not spreading misinformation so you can you can take out your phone on

52:11

Google L's profit margin oh wait anybody want to Google I'll see your C margin yeah so that's a very good idea yeah that's that's insanely high it's way higher than any of the notorious company that that that um would typically call them out for like over 30% profit margin so and lcv is one of the one of the publisher that I battle with the most cuz they published so many journals

53:02

and then there's so many journals are absolutely completely polluted with uh paper meal stuff um so what we can do I think what we can do is very critically important so it's very the thing I hear the most is like yeah maybe there are just some some some thing the methods I didn't read maybe I just don't understand their their technique there's always fancy techniques no a lot of times if you have done an experiment a lot of times your God feeling is probably right do not ignore your own God feeling and say I

53:44

what do I know you know maybe you don't know but it's always reasonable to doubt and um P Pier as I mentioned before is very uh important to to um use and then you can always re report to the aics Department of Publishers um all right next slide please so okay uh wait okay here here we go so I I don't know how many people are still in graduate school how many people are are posted up he go go back a little bit go back sorry

54:26

yeah um but it's important to identify because when we see frosters uh we don't want to we don't want to get into like a go oh my God look at that guy he's so terrible or like oh yeah the paper meals are only by crazy people I I think I think it's it's important to to understand that if you're in science every single every single of us will be at some point under the pressure of generating pretty or prettier data you know you're like okay what do I present in my next uh lab

55:08

meeting what why is my data always so so ugly somebody else is reporting something you know foreigners uh like myself one used to be okay I need my Visa sponsored I need to get my H1B you know um and a lot of times the pi because I work at Colombia I work with a lot of research Labs I hear a lot of Pi goes like this okay if I don't get this grant I don't think I can support you next year that sentence is usually followed by here's what I really want for that

55:48

Grant nobody is stupid you don't have to tell people hey can you go fix something or you have to put pressure first and then your desire result second and I'm probably put a lofty to that would be really wonderful if we develop a drug treatment for Alzheimer's using this fantastic animal model that has been replicated so many times you know uh so so this strategy really works on the vulnerable folks so by vulnerable meaning like you need your degree you need your paper you need your tenure you

56:29

need your Visa sponsored so what I'm trying to say is that when we look at the fra was like oh that guy is so but is in it's the vulnerability is in every single one of us right so it's it's really more important to identify the pressure that from external sources like your piis your bosses and and from internal I really need this paper I really need whatever none of them are wrong goals but the pressure can become toxic and it

57:09

becomes a sleeper oh I I I'm pretty sure these are outliers if I just get rid of these two outliers it'll be perfect right or the other thing I see a lot if you do animal research is keep adding animals or just add two more M I add four more M I'll add six here and and then eventually and then take out outlier here and then and then so my analogy is that it's like it's like it's it's sculpting and signs are different but they're kind of similar in science because a lot of the senior people

57:49

commun of fraud I feel like part of is it's like they have spend their lifetime having this their their musees they're they're they're sculpting you know their theory of a Amoy or whatever they spend their their lifetime trying to develop this is their the gem of their life okay um um that they it's it's the the time and the money that you invested in it making them vulnerable to be more select to the to the to the evidence or the fraud that is you know um that that is making them to be more blind to evidence so um take-home

58:33

message is to start from every single one of us identify pressures knowing the knowing the you know how much money the LCB is making from our vulnerability and and then and then yeah

59:04

yeah I think that's a lot of um there is a movement to be like you know to bend and Pier reviewing T right yeah

59:32

so you could request right so you can request on PO here and saying could the author please share your raw data or raw code or whatever um recent paper they should have uh or they they should be obliged to provide that to you but you know some of some of the older paper you just like you know do you have any advice how to Scientific fra and get away with it um so so anything I caught are probably the dumbest ones um so Mo most fraud do not get caught so has the

1:00:13

quality of fraud improved I don't want to encourage anyone I probably cuz I I think every time I post on on blue sky or Twitter I think the watch learning oh she doesn't like this one she doesn't like when there's a break or you know yeah is very one question and this is very important for those of us who are pursuing PhD or a master's degree or any sort of study in public health can you please share the experience like the fall out the challenge facing what did these people sacrifice as a result of

1:00:55

attraction so the F my first case a temple Universe guy still has his job uh nothing changed the the the ni guy because his position was so high he stepped down from the director of science but um yeah not not much no most people have their job and fundings and just retraction just whoops yeah

1:01:34

[Music] I think I think researcher has to want something other than papers Hing Dax and Grand money right because what you want is what the market will feed you with so yeah had to change what exact what we want do we want real science or do we want the things that I grew up with like you want a peers and you want you want ten years you want you have to check all this which is I think we to rethink why we want this or what yeah I I mean you you guys can catch me

1:02:15

somewhere else three questions three questions okay go go go always be auor or somebody else at home um most question most time is the authors yeah about kind

1:02:39

pressures oh I think it's started with like like yes anytime yeah if you have you have a tendency to please yourself please your every single one of us has being there so that's that's just uh that's why that's I yeah uh I just question the percent how much of the research would like to estimate as a whole I not I'm goingon to lose this B so this is gonna be his next favor bat all right last question okay last

1:03:19

question um have you had any responses positive or negative from your colleagues inia [Applause] [Music] M yeah so I run the core right there have people were telling other people to not use my uh saying that I don't know what I'm talking about and stuff so um no force hading that yet so all right all right well let's give we have we have a little present for hold

1:04:00

have a mug t-shirt that

1:04:13

says all right if you have questions for her after and sneck around we have Jazz by the New York um New York BL New York Blu Rockets So stick around and they're going to set up and play so there there's going to be lots of uh hangout and if you want to join their band you can also just jump on stage and jam and jam along is that correct yes all right thank ask questions and stick around for the the Blues