Plagiarism is bad. This is even worse. | Wrong Number
Вставка
- Опубліковано 7 вер 2024
- Academia values the appearance of truth over actual truth.
reason.com/vid...
---
After surviving a disastrous congressional hearing, Claudine Gay was forced to resign as the president of Harvard for repeatedly copying and pasting language used by other scholars and passing it off as her own. She's hardly alone among elite academics, and plagiarism has become a roiling scandal in academia.
There's another common practice among professional researchers that should be generating even more outrage: making up data. I'm not talking about explicit fraud, which also happens way too often, but about openly inserting fictional data into a supposedly objective analysis.
Instead of doing the hard work of gathering data to test hypotheses, researchers take the easy path of generating numbers to support their preconceptions or to claim statistical significance. They cloak this practice in fancy-sounding words like "imputation," "ecological inference," "contextualization," and "synthetic control."
They're actually just making stuff up.
Video Editor: Adani Samat
Audio Production: Ian Keyser
Photo Credits: Michael Brochstein/Sipa USA/Newscom, Walter G Arce Sr Grindstone Medi/ASP, Graeme Sloan/Sipa USA/Newscom
Music Credits: Strange Connection by Nobou, Digital Dreams by Jimmy Svensson, Nothing Can Stop Us by Nobou, Hero Is Born by idokay, Sneaky Shenanigans by Charlie Ryan
She resigned as president but was brought back as faculty with a six-figure sum
900k YEAR salary
@@moneyobsessed but she's a black woman! so our arguments are instantly negated
Harvard = Barely qualified to work at Arby's.
Those who hire for serious jobs absolutely ignore institutions as ruined as Harvard. Their graduates are a joke, but much more so a liability. Whether it's those that were admitted and allowed to pass because of their skin color or because of who their dad was...they're all serious liabilities that will ruin your company.
Harvard, can’t help themselves pop…
I'm sure she's struggling after the 500k pay cut...
i'm surprised to learn that my greatest misconception was in thinking that falsifying data was actually difficult.
Committees check selectively. They chose to give her a free pass.
What terrifies me is that they appear to be ignorant and/or lazy about it. My statistics education is limited, but I would never use Excel to interpolate imaginary data into anything. I would also not just toss out outliers just because like Gino or Hauser. What were their schools? We cannot reasonably assume their conclusions are worth using in our lives. Therefore, they are irrelevant to life.
@@chipcook5346 i guess that's why you're not a famous academic, grifter, or university administrator. ¯\_(ツ)_/¯
@@jaewok5G You know it, baby! It's all monkeys and dart boards for me.
Journals and universities don't require any controls or accountability of data.
Here's one scenario: A grad student (who is doing all the work) takes research notes on paper. Then that data is copied into Excel. The professor (who is the author of the paper) realizes it's not statistically significant, and edits the numbers in Excel. The paper notes are never published, and might even be destroyed. So there's no evidence of the fraud. The grad students might never know their work was tampered with.
Excel and other consumer software should be banned from research. Instead, scientists should be required to use software that maintains a log of all changes, including which user made the change.
There's a replicability crisis in science right now. I think it's in large part due to manipulated data and junk science.
I'm a lifelong academic, have served on the editorial boards of seven international journals, and in my expert opinion there is no question-not the slightest-that Gay committed plagiarism, and at a level far far greater than as specified in Harvard's codes on this subject. And no... the fact that half of Gay's publications bear plagiarized material cannot be ascribed to some graduate assistant or such. There are SO many examples to provide (more than I've ever heard of from a single person), but here's my favorite: Palmquist & Voss (1996) wrote: "… the average turnout rate seems to DECREASE linearly as African Americans become a larger proportion of the population. This is one sign that the data
contain little aggregation bias. If racial turnout rates changed depending upon a precinct’s racial mix..." which Gay plagiarizes as (Gay, 1997) "… the average turnout rate seems to INCREASE linearly as African Americans
become a larger proportion of the population. This is one sign that the data contain little aggregation bias. If racial turnout rates changed depending upon a precinct’s racial mix,..." Note the verbatim (and unacknowledged) copying... save for one word, the change of "DECREASE" to "INCREASE," which the careful reader might ascribe to an attempt to forward a political agenda. In my expert opinion, this is far WORSE than plagiarism: It is a) not doing the original data collection or analysis, b) copying verbatim the scholars who DID do the data collection and analysis (without attribution), and c) inverting the conclusions (likely to further a political agenda). No wonder Professor Carol Swain of Vanderbilt, whose work was plagiarized by Gay (and who happens to be black), said Gay stole Swain's ideas and hence doesn't deserve to be called "Dr." because she (Gay) didn't do "original" research for her PhD. And Gay is the best Harvard could find? Really?
Perhaps Gay WAS the best Harvard could find.
I choose not to guess what she might have been the best FOR.
Thank you! Great YT comment.
Yep. The best black woman with all the right opinions and connections that Harvard could find.
Oh, you thought they were looking for someone actually qualified for the job, didn't you?
Any commentary on the assertion that up to 40% of papers to some Journals are produced by Paper Mills: "A report by the Committee on Publication Ethics released in June 2022 confirms Day’s findings. The report, which looked at six publishers, found that two percent of papers submitted to journals may have come from paper mills and for some journals, the number may be as high as 40 percent." - SHARYL ATTKISSON. There are also problems recently emerging with many persons unable to replicate experiments as reported in scientific and medical journals, ... My biggest gripe however with relevance to Harvard is the much bigger problem that at least 2 high profile Professors at Harvard (in the African Studies Department) made their careers falsifying data and still retain their posts.
Do I understand correctly that she inverted the conclusion of a study without providing any reason for doing so? I am not sure how important the claim is for the article she wrote but if it is really relevant to the main point of the article it should lead to a retraction of the article and probably lead to an end of her professional career. I don't see this as plagiarism but as a far more problematic fraud as it means that she knew that the data said otherwise and decided to hide the evidence. This is incompatible with scientific pursuit.
Just when I think I'm cynical enough, somebody reminds me that I woefully underestimate the proper degree of cynicism I should have.
me too!
If a student did what she did, they would be given an instant F in the class (thus more or less permanently destroying their GPA) and expelled from the university, and possibly blacklisted from other universities.
Yes you are correct.
And consider how many student careers she has destroyed for exactly the same thing as a professor on a panel or instructor. For her to climb so high at such place as a fraud, others had to be in the know. I am sorry to say that I have seen such things even in the hard sciences.
Students actually do this all the time with no negative repercussions. Education faculty in this country, from K-12 to Ivy League universities, are immune from accountability and work in a culture where diligence is discouraged. Where do you think Claudine Gay learned this behavior?
When she did it, she was a student.
It would be nice if that were true. The administration of my university is very lax with academic dishonesty and therefore cheating is rampant. It puts students who refuse to cheat at a marked disadvantage. These are the people who support no cash bail, they love any behavior that Destabilizes institutions and damages society, why would they be harsh with cheating students?
This guy was a great hire for Reason.
Knowing nothing about the matter, I would say he's badly underpaid. What a f*ing rockstar.
Seconding that!
Gay KEPT her JOB as a Prof at Harvard
$1 million/year salary too (or so I've read). Unbelievable, considering what happens to students who are caught plagiarizing.
I hadn't heard before either that she won an award for that plagiarized dissertation.
Yet Harvard is highly regarded and superior to other institutions? Give me a break.
It may be Just A Job, but it's a Harvard Law job. How far the mighty fall.
With $900,000 🤑😂
Shocking! Political activist uses bad faith means to achieve ends!
Do not go where the path may lead, go instead where there is no path and leave a trail.
-Claudine Gay
(She was plagiarised by Ralph Waldo Emerson. Seriously. Claudine wrote this first. Ralph shamelessly copied her.)
@@darrellfuller8078 wow, that is the most insightful, original quote ive ever heard, i'm sure no one in history has ever said anything remotely similar 😂😂
@@aidananstey9848 Claudine Gay plagiarised the Acknowledgments section of her thesis. She couldn't even thank her mentors and family without committing plagiarism. I guess that's the standard for earning over $900K per year at "top" universities.
What would one expect from a black, lesbian, shaved head radical?
@@darrellfuller8078 he would go on to shamelessly copy Claudine 200 years prior
I entered a Dutch research university around the time of the Diederik Stapel scandal. A Dutch psychology “researcher” who made up entire data sets, like literally made up the data. We have waaaaay too many social scientists in the world, especially those who are subsidized by the state.
"Social science" is one of the worst oxymorons (emphasis on morons!) in existance. It is most often nigh on impossible to test a hypothesis in social (or political) 'science' -- takes too long, factors greatly affecting the study cannot be sufficiently controlled, and volunteers may be tough to obtain (esp. for the 'worst' side of the study!). True science requires hypothesis, then experimental design, then experiment, then data collection and analysis, then conclusions and follow-up (including, often, more experiments).
Should be called social or political arts...
No please dont call it art. It needs a new term so we dont continue to besmirch hard science or any other established endeavor.
The soft sciences are beyond easy to fake. Because it's literally impossible to get the exact same results twice. No two humans are exactly the same so when you're comparing us you can't do it the same way that you would to bridges or two cells. Because those two things are actually supposed to be identical. If they're not you see the problem. Two people are not meant to be identical. We are meant to be different. So lying about it making it seem like you're right is super easy barely an inconvenience!
"They're actually just making stuff up." That quote makes my day. Thanks.
Amen. Plain, honest - even if brutal - communication is rare, underutilized, and under appreciated these days.
If you enjoy that style, you may enjoy clips of senator Josh Hawley as he grills various D.C. swamp creatures. Even if you disagree with him, it’s a refreshing and rare style. One of his best is the most recent example of him going “nuclear” on Deb Haaland. Entertaining.
Best Regards
There was a time I considered myself to be an academic. But that was so far in the past that it was "allowed" to publish a paper with a negative result. It was (almost as) valuable to show directions of research that didn't work as those that did. When that standard disappeared, so did the respectability of academia.
Being right is not nearly as important as being BELIEVED.
Is this your quote? I love it….
@@bjnowak It is. Thanks!
Being right is not nearly as important as being BLACK
@@cedricwilfordI'll cite you Cedric, great quote.
Wow, this is one of the best quotes I've seen in quite some time. Nailed it!
Figures don’t lie! But, Liars can figure.
And still she ended Richard Fry career on laughable grounds, just because he published a paper with well documented data stating that armed black men are actually shot LESS by the police; with Fry, the lab he created and that was actually helping black kids was also lost.
That’s because, in my opinion, exposing those types of facts make a lot of people, namely liberals, and BLM, a lot less money…
That’s because, in my opinion, exposing those types of facts make a lot of people, namely liberals, and BLM, a lot less money…
This is the best video Reason has ever made. Give this man an entire show.
WOW how far Harvard has fallen. Low standards and bad results follow. But sadly as a researcher this is very common and regrettable.
Gay is not an academician, she is a politician. Optics over objectivity. Between Gay's behavior and the rapping "Harvard Docs," I have started looking carefully at physicians' diplomas. Harvard? I'm out of there.
@@richardfabacher3705 bwhahahahahahahahahhahah I'm sure they're terrified
Was it like that when you was there ?
@@ZIGSVIDSI know you don't realize it but you're making black people look worse
One of my professors in my PhD was teaching us quantitative research messages. He flat out told us “look the point of all this is to make a bunch of worthless data mean something. If you can’t make that happen you cannot get published. If you don’t publish you don’t keep your job.” Still look at this class as the moment I realized I wanted to teach at a community college instead of a 4-year school and I have never regretted this decision.
...at the university, "Publishing matters more than truth."...
something something something, high priests of climatology, something something something.
Lies, damn lies, and statistics
Gay is a classic example of DEI.
you think women and minorities have a lock on academic fraud and plagiarism??? lmfao
Now I only have a bachelor's in Sociology and Political Science (and yes it was a bad financial decision), and since my graduation nearly 11 years ago I have forgotten a lot. But I will never forget how much of what academics know is utter baloney.
If you're going to read this comment further I will say this (indulge me): the first year of any social science program should include two courses in statistics and two in economics. Many social sciences students (like me) put off stats until then final year, and never took an econ course.
This^^ 💖I found myself interested in soc but after talking to a few faculty and reading some of the course material I felt quite repulsed by how biased they are despite sharing many of their views🤦🏻♀😕My thoughts were stats & econ would help a lot in filtering out the BS too. Perhaps even philosophy 101 or discreet math or courses to strengthen one's logic & critical thinking?🤔May be a bit of polsci, law & history can help build context too but I'm overthinking at this point 😅I'll probably take a soc 101 in my part time but it's certainly not something I'll be majoring in. 🙅🏻♀
But feel free to share what stat or econ courses would you recommend for a freshman, I could use any and all suggestions 😅😇
@@babyqueenxo As an actual economist, a basic microeconomics course (Supply and Demand) is fine. Understanding that one graph, and what happens when you start manipulating it, will inoculate you against so many bad political ideas, its just not funny.
OTOH just reading a few easily obtainable and digestible books will do the same job:
_How an Economy Grows and Why It Crashes_ (Peter Schiff)
_Rich Dad, Poor Dad_ (Robert Kiyosaki)
@@babyqueenxoBefore you take any social sciences course, read/listen to "Basic Economics" by Thomas Sowell.
If you must, do your intro poli sci, sociology, and philosophy courses - but take intro stats and econ at the same time.
Economics is the superior social science, poli sci and sociology are downstream of econ.
Aaron Brown is smart, honest, clear, and factual. He is a mensch.
Incredible. Now THIS is journalism. Great work.
Statistics are difficult, but can be made a lot easier if you know the results you want in advance.
When I saw that testimony I realised:
"Oh, THAT'S why 10 Commandments were written. Society has been here before, when the educated lost their moral compass and we needed to create 10 rules (10 fingers) of 3 words each that self-reinforce."
"Academia values the appearance of truth over actual truth." - Beautifully summed up!!
She just played the game. Harvard was looking for a POC token to show that they weren't racist. They chose her, and all she had to do was throw some papers together as a formality. She didn't do the research because they weren't evaluating her on that.
Figures DO lie, when liars do figures.
If I were to start a university, the path to tenure would be simple: number of papers discredited. That's how academics worked in the good ol' days; they would argue things out in papers and truth would prevail.
If I were running a university, I would say no tenure. Everyone works year to year.
@@robertewalt7789 Yeah right. Twelve years of expensive study and debt to deal with, you gonna take a shitty job with no security. All the people that no-one else would hire would churn through your revolving door. You'd have the worst reputation and awful students.
Reminds me of the quite infamous-at-the-time "Bully Hunters" fiasco. They had a trailer video which stated:
_"Over 21 million female gamers have reported sexual harassment in-game"_
Note that it does not say "it's estimated that 21 million female gamers have experienced sexual harassment in-game", or even "21 million female gamers have experienced sexual harassment in-game". It explicitly says that 21 million HAVE REPORTED such harassment.
So, where are these 21 million reports? As you might have guessed, they don't exist, and never did. That number is based on an online survey where the number of responses was... drum roll... 874. Not 21 million, but 874. (From those reports 35% claimed having experienced harassment.)
And to top it off, it wasn't some kind of academic or governmental survey, or any kind of survey conducted by a company dedicated to such surveys. It was just a random blog post somewhere.
Where did they get the "21 million" number from? By extrapolating from that 35% value. But, as mentioned, that's not the extent of the distortion because, as mentioned, they say in the video that "over 21 million female gamers HAVE REPORTED". No, they didn't. Those reports don't exist anywhere.
The famous "97 percent of climate scientists" meme is similarly largely an invention. Are there really 10 thousand climate scientists? No. When I encounter that, I ask for a list. Got list? No?
@@thomasmaughan4798
What are you even talking about?
@@DjVortex-w "What are you even talking about?"
The topic of this particular comment thread is misleading statistics. Climate politics is a good example of misleading statistics. The 97 percent is accurate, but it requires careful examination "of what" exactly and it would easily have been 100 percent but that would be suspicious. 100 percent of people that claim AGW believe AGW. But 99 percent has been used; 98 percent has been used; 97 percent is a nice prime number that now is used only for climate politics.
Hypothetically, someone could easily decide on 97 percent FIRST as a meme and then *adjust the statistics* to arrive at the desired 97 percent. And it would not be wrong! Merely misleading.
@@DjVortex-w He's talking about the universal belief that 97% of scientists agree that climate change is real, caused by man and a crisis. BUT, the original survey of 10,000 scientists were asked 2 questions 1) Has the planet warmed in the last 160 years? and 2) Is human activity a significant factor in this warming? They got 3,146 responses back, but only 79 of those responses were from self-described climate scientists, and 77 of those agreed with the second question. There was nothing in the survey about a crisis, and the sample size was somewhat small. But this is the public consensus as it has been pushed by politicians.
@@DjVortex-w its very clear what hes talking about
This is a genuinely fascinating prospect. People will blow tens of thousands of dollars in a year to be taught by universities. How much of that value is literally just regurgitated information or falsified information? I’ve been saying for years that the education system is highly flawed, be that Public, Private or otherwise. If the fact that we invest millions of dollars into universities and can’t even trust them to pursue or teach the truth, I’d say that’s the most damning piece of evidence that our schooling system needs to be MASSIVELY reworked…
I learned how to fake data in high school.
If you understood the chemistry experiment, but your attempt failed, you just create data that shows success.
Even easier, copy someone else's successful data and shift the numbers a bit.
(Dry Labbing)
I had a stroke of great good luck in elementary school. We performed a chemistry experiment, and my partner and I didn't get the result we expected. I don't remember exactly what it was, but it was something distinct, like something burning with a tiny green flame. My partner talked me into reporting that we _had_ gotten it, to avoid the embarrassment of having botched the experiment, and that's what we did.
Then the teacher explained to the class in detail what the results ought to have been. My partner and I had misunderstood; the part of the experiment that we had just turned in was the control portion. There was _no way_ it it could produce the special result. I will never forget that feeling. The teacher never said anything to my partner and me about it, but I knew that he knew that we had lied, and rendered the whole experiment pointless.
I never faked data again, and when I worked as a teaching assistant, and one of my students reported that in a physics experiment he had found that momentum is NOT conserved, and he couldn't find any other explanation for the anomalous result, I gave him top marks, and presented his lab report to the class as the ideal of how science should be done.
That would work about four out of five times in my chemistry class. The fifth time, the procedure and ingredients are wrong on purpose, so if you follow it right, you won't get the right answer, and that's where the "explain your discrepancy" question earns full points. If you suspect the sugar was sand, or the reagent was diluted too much... congratulations, you get the grade. If you just do math with faked numbers to get your expected result... you fail the class.
@@williambarnes5023 That's why understanding the experiment is important.
Dry labbing is used to show you succeeded in doing it right, despite failing to do so or failing to even do the lab.
I was accused of plagiarism four score and seven years ago. But I didn't and you can hold these truths to be self evident.
Falsification of data has always been present in academia. It has gotten worse lately. The worsening of the issue is just one more symptom of the focus on ideology over academic integrity and merit.
People who argues that lies, their destruction of the meritorious system, etc are justified for their cause are often idealogs and/or narcissists. I think given that they scarcely ever succeed at helping their own causes they claim to champion suggests that it's all self-interest.
Thank you Aaron Brown! Your analysis is always interesting to watch. As a researcher working in a different domain, I am always amazed at how it works out in social/political sciences (his is not to say that there are no problems in my domain, but rather that they are quite different in nature and have no such implications for the society)
What's *_really_* "outrageous" here IMO, is the (relatively) indisputable fact that basically no-one - and especially in the Jewish community and academia + Harvard donors & alumni - cared one iota about Gay and her politics - which targeted Whites, until it became publicly known she might not passionately love Jews and Zionism. Had it not been for that, she would've remained a president at Harvard and could've continued her racial grievance politics and plagiarism disguised as "science and studies".
That she kept her job as a professor there speaks volumes.
Students were chanting to annihilate a specific population of fellow students. That put her in the spotlight and the plagraism was used etc….In a perfect world yes this would have happened regardless of israel Palestine…You disregard that as a cause for action and etc :)
Ugh. Medical and social science researchers often learn just enough statistics to fool each other.
Integrating over possible values for unknowns, rather than picking point estimates ("imputation"), helps with both avoiding unwarranted inferences and focusing future data collection to maximally reduce uncertainty.
Great piece! It requires a stunning amount of homework to get to the bottom of these bogus stats. By the time you do, 10 more studies have been published.
I would also suggest a vide on the topic of "idea laundering" which is paying for university data, then publishing it in the media, which makes the paid-for data a new fact.
The appearance of truth is more important than actual truth. Brilliant!
Her shamelessness is beyond measure. As someone commented.. she would have no hesitation in destroying a student if their ideologies conflicted. Actually a wicked person IMO. Imagine carrying on like she has if she was white... She'd be locked up. Black privilege strikes again.
I critically appraise research studies for a living, and it never ceases to amaze me how much "imputation" is tolerated when journals accept a paper for publication. So many worthless articles out there, and so many important questions that have yet to be studied fairly. 😢
Are they really worthless or do they serve an agenda apart from getting a paycheck and career growth? In the area of soft sciences I've seen they all have the same partisan narratives. It's more like activism for an agenda rather than scientific research to find the truth.
I work at an institution where there is a lot of social science input, so I get exposed to this stuff. I've seen things where it's clear no research has been done, data is misrepresented, and novel definitions for words have been used in order to "support" the conclusions.
The entirety of american critical theory, social justice theory, post mid century feminism, queer theory and everything after that is all based on that model- marx would be turning over in his grave except they are achieving the results he wanted.
Plagiarism is okay for our country's president, so it should be okay for university presidents too.
If everything presidents do is OK for everyone, the country would fall apart... Even more than it already has.
It's not okay for presidents, either.
Clearly rampant plagiarism was no bar to him sitting in the Senate for decades, why should that change just because he's moved his office a couple of miles across town?
Gay kept her $900,000 per year teaching job.
And she doesn't even need to have a class to teach. She just gets the title and doesn't need to do anything.
@@SimonASNGthis is a good thing. Anybody who is taught by her is doing themselves a disservice. If foolish people want to keep giving her money then that's fine as long as it's not my tax money
D.E.I = Didn't Earn It
Fraud is a crime. Prosecute it as such and stop all the hemming and hawing.
Please Reason more of this guy.
Excellent synopsis. The issue is that 'those in charge' are also 'those who are culpable'.
Having published hard science research articles in engineering, i spent a TON of time gathering data. Sometimes, we had to revise experiments to get better data (more consistent). We always had to calibrate our models with real data, and you can only trust any model within the range of the experimental data. Outside that range, you are guessing. Sometimes, that guess is good enough, but you always mention that the inference was interpolated, and you do your best to make sure there is a good reason for it. For us, it was usually that data was not possible to measure something that small. We would then use secondary validation to firm up the model to test if the model was useful in describing a given phenomenon. We had to scrap very expensive experiments that did not pan out due to a lack of data. Would have never dreamed of falsifying it.
The bar hasn’t been lowered. It’s been removed.
This seems like the worst period.
Even the market are now very unpredictable. Started investing recently when the market prices were a bit high,today I am more than 60% down!
Please educate me, i'm willing to make consultations to improve my situatio
Please how do I find this financial counselor?
I'd love to know this mentor of yours
beep beep boop beep thank you sonia campbell
Woooow look at all these bots. So cool. Hey Reason, I don’t know how UA-cam works from a publishers point of view, but if there is an ability to remove comments, can you remove this one? I know freedom of speech if important but this isn’t speech, this is a bot thread completely filled with intentionally bad faith replies to form a mock discussion with the intent of misleading readers to think it is genuine.
Medically - Don’t give me a drug unless it has been thoroughly tested
This one will go into the DEI's greatest hits mixtape
As a Data Analyst, I really appreciate this piece and will be sharing it.
Hi, I'm curious if the field of data analysis has these issues to the same extent too? I was recently learning about the ecological fallacy in regards to a certain third rail topic and I couldn't help but notice how the presenter's bias was blinding him into committing the same fallacy & drawing non-sequitur conclusions. 🤦🏻♀
These episodes are tremendous. Always enjoy and value them so much.
she should have been fired
If we don't hold the President of Harvard accountable? I guess it's ok for students to plagiarise too?
Oh no! Harvard students will be kicked out instantly. But Harvard do not screen their professors for plagiarism. Ex-President Claudine Gay is still a Harvard social science professor with a hefty six-digits salary. And she is not the only one. Academic rules are only selectively applied to Harvard faculties.
My 3rd grade teacher taught us about this stuff with fancy sounding words to lie.
Two other 3rd grade teachers told us we would all die in 10 years from the next ice age.
Thank God for the teacher that prepare us.
And we supposed to believe she's not in place because of an agency
Claudine Gay needs to be stripped of her Doctoral degree, it is not worth the paper written on.
Great video. Honest and straight to the point. The basic test that could be applying to all of these examples is to watch how sensitive your finding is to the number you invent. You can't just guess and then wipe your brow that you didn't null your hypothesis.
Gay was probably not fit for Harvard if admissions didn't consider race, so in in some way it's not her fault.
I know this is satire, but we shouldn't be entertaining the slightest chance that rids them of their accountability.
She already KNEW the “truth,” any real data would just get in the way.
Data imputation isn't necessarily just making stuff up. For example, if you recruited 50 people into a trial testing a new drug, but by the end 15 had dropped out for some reason, you might quite reasonably say your sample size was now only 35 and conduct analysis only on that (a complete case analysis). However those who went missing might not look exactly like those who stayed in - that is, their absence has now **biased** your sample. Data imputation allows you to explore the implications of this. You can use the scores you did collect in your partial samples to generate a few statistical guesses as to what sort of scores you might reasonably expect those now missing people to have returned had they stayed in. You can create a few different scenarios here (SPSS defaults to generating 5 imputed datasets with different values in the missing cells) and compare them to your complete case analysis. If there's a big difference between your real and "made up" datasets, you know your sample is potentially very subject to bias caused by your missing values.
All that said, treating imputed data as if it's real data is definitely very suss and I have seen that done...
Yep. I'm shocked at how many medical research studies I've seen where the completer analysis *supported* the null hypothesis (not just a lack of statistical significance due to being underpowered), but the ITT analysis shows a big effect and very low p value. Like, the imputed numbers were what made it a positive result.
Like they didn’t know about her plagiarism before they put her in that position. She is a typical DEI hire.
Thank you for a well thought out and logical video. A treat these days!
How much longer western men? TAKE. YOUR. COUNTRIES. BACK.
I really value and admire this vital public service of auditing agenda-driven flawed statistical analyses. Keep it going!
And now she teaches a class on academic ethics?
When I die, don't let me vote for Dumocrats.
I genuinely hope Wrong Number becomes a series. Thank you for making this video.
Can we replace the electoral college with Excel Auto fill?
agreed!! william henry harrison -would've been- will be a transformational president!
Maybe just switch the President with Excel Auto fill, and the congress.
That would make the goverment 109% more efficient
But we don't want to be dependent on a proprietary application that runs on a proprietary OS. Autofill should only be done on LibreOffice (on whatever OS you want) because you will be able to audit how the function works.
too late.
@@jaewok5GMAWA!
The next time someone brags about graduating from Harvard or another Ivy League school, remember exactly what that means. And also remember that many of them will end up, or have ended up, in positions of power and control. Gay was doing this 20-25 years ago, so graduates making excuses like "I graduated before all of this" or "that's just recently" are blowing smoke. On that note, I wonder if anyone has analyzed "Dr" Jill Biden's alleged work?
I read Michelle Obama's thesis; that was a day I will never forget...
Lack of principles
Will turn any society, any community, any institution, any individual
into a cesspool.
There are liars, damned liars, and academics...
Aaron must drive modern academia nuts. Good!
Thank you for this excellent video.
I read a paper that was defending wind power from criticism for killing eagles and hawks. It had a ridiculous argument about how cats hunt more doves and crows so wind power killed fewer birds overall. But the worst part was the basic math errors. The paper was cited by an article defending wind power
It's much easier and common to produce "research" that's garbage in the soft sciences than it is in the hard sciences. You will see that type of thing far more often in psychology, sociology, political science, etc. It gets far worse in topics like gender studies and such. On the hard sciences, that's why one waits for studies to be reproduced before getting too excited about any single study.
Yes. The 'studies' fields are largely derivative. Political science, economics and psychology are more serious than sociology, which is ground zero for ideology masking as science and sloppy empirical research.
@@InfrequentObserver Well said.
As an intellectual historian I've always had a professional bias against purely quantitative studies absent of good qualitative argument ;-). I remember the "Time on the Cross" controversy, where Fogl and Engermann presented such calculations as how often the average slave was whipped 😞, a truly questionable contribution to slavery studies. But seriously, while there has always been fraud and ideological bias in academia, the brutal neoliberalization of higher education since the 90s, which has turned academics into closely monitored "publishing automatons" has generated a whole copy & paste and "invent your own data" culture. It's pretty disheartening.
I love that after this video, I get an ad claiming a study in some overseas country shows that coconut oil balances hormones. I'm sure this "study" is even a bigger joke.
Thank you for this and for people in the comments showing examples of her plagiarism. Congress could have done better in doing so, it was a bit hard to entirely believe the narrative they pushed because Congress is not immune to pushing an agenda. But it’s clear now that she did indeed commit plagiarism.
No other country on earth would allow this woman to be in charge of anything.
This is the kind of content for which I subscribe and support Reason!
The problem is "publish or perish," and citation networks. Academia is rotten to the core, with the focus on novel publications, hypotheses, and findings has a horrendous inflationary effect. Just like your food and the dollar in your pocket, government meddling has destroyed the value of a core aspect of the human experience in our country.
If you are a very small group of the richest, most powerful power you BUY everything that will help you get your way - this is how science became 'the science'.
Yes, THANK YOU!
For this clear, simple & concise explanation of one of factors eroding the value, usefulness & accuracy of scientific research publications, as well as the reputation and public trust in science in general.
As a statistician I always tell people to be skeptical of the studies they cite in conversations because they don't know how those studies were conducted. To be fair, often times those studies are misquoted by the media and the general public.
But this type of BS is just rampant in Peer Reviewed studies as mentioned in the video and really gives statistics a bad name and causes people to distrust data analytics and thus doing nobody any favors.
Having just learned about imputation as a data science tool in my studies, this was super interesting and gave me something to think about. Thanks!
This just confirms what I was already thinking, that I should assume every published study is trash.
Not every study, you should look for replication of studies from independent researchers.
Look up the Grievance Studies Affair (James Lindsey, Peter Boghossian, Helen Pluckrose got fake studies with fashionable jargon published) to see just how right you are.
Only usually in the soft sciences, hard sciences (biology, chemistry) are usually right
If there is only 1 study about something → the results are worth anything till it would get confirmed by other people doesnt the same tests.
If multible studies by diffrent people have the same results → you should be able to trust it.
BUT:
Dont blindly trust the media when they talk about "shocking results from a new study", from personal experiance, if you go and read the study for yourself you will probably find that the article is mostly wrong (by mistake, or intentionaly for clicks).
For example, every time you see an article with "X can cause things like ADHD" it is wrong, the media tend to insert this line for no reason. (ADHD is hereditary, the only way to have it is to be born with it)
@@gaijinbroken6571 Exactly. As a former scientist (biophysics) I'm a little insulted by the fact, that the video often generally says "academia" when all examples are from social "sciences".
Those first 10 seconds really tell you everything you need to know about the US education system.
So much hate in these schools
Getting published is the ONLY thing that matters...the truth is so damned annoying
To be fair, William Henry Harrison would receive my vote over either the Democrat or Republican candidate...
Maybe we should just skip the Presidential election entirely and install Emperor Norton.
But we have to lower the standards to give the marginalized people the same opportunities that are available to those who operate on merit.
People with actual PhDs do statistical research in Excel? So there’s R, Python, SPSS, Stata, Matlab and Eviews and people with actual PhDs do statistics in Excel? That’s just so messed up…
Excel is fine for basic analysis. Using the autofill feature on any program is completely unacceptable.
You can do a lot in excel, and none of those other options are part of standard curriculum in my primary domain of expertise. R has become widely known only recently, while something like MATLAB is way much more expensive than excel, which you can easily pirate. Knowing programming and knowing stats are quite different things, even more different is to know something about the domain where you hold a PhD (think taxonomy, relativity theory, being able to make and read maps).
@@iampdv I’ve got a masters degree in economics and I’ve been using Excel, R, Stata or EViews, depending on which one of those a given professor preferred. Later I learned Python on my own cause it’s very flexible. And I mean yeah, you can do a lot in Excel. You can technically even run Doom in Excel, it is Turing-complete after all. That doesn’t make it a good idea though if you’re doing advanced statistics imo. In purpose built data analysis programs you get pre-made models and tests for them that would be crazy hard to do in Excel, not to mention slow, error prone and limited by the number of rows that Excel supports.
my masters in a different (stem) field was obtained about a decade ago in a leading university in my country of origin. We had excel, excel and excel depending on prof's preference... I did learn MATLAB for some physics-oriented tasks during my PhD work, but that was not supposed to be part of my PhD work. In many if not most stem field there is a huge gap between programming, stats and the current problems, and you would need stats only occasionally which doesn't justify learning R or python or even spending time on choosing which one to learn (I have no clue what the other two things are). I am writing this as someone having a paper with MC simulations done in Excel in a q1 journal... Very cumbersome, I agree, but had I done this, say, in R, I would've forgotten how it works by now anyways... Even with MATLAB, which I use more often, I need to relearn every time I come back to it...
I mean ideally excel and all ms products should go, so maybe one day I'll finally decide between R and python when migrating to Linux. But I am not doing this right now, and I have lots of other problems to deal with trying to produce outputs with experimental data in two quite distant scientific fields...
@@iampdv well, first of all congrats on publishing in a q1 journal. Secondly, you’re probably a genius if you can pull off complex Monte-Carlo simulations in Excel, congrats on that as well. I certainly couldn’t do that. That said, I still find it very weird that PhD level researches in social sciences use Excel, given that specialized statistical software exists and is often free and is widely taught in universities. Maybe that’s not that weird in your field? I have no clue. Doing regressions in Excel instead of R is kinda like drawing maps in Paint instead of in ArcGIS to me. Like sure, you can do that but it’s such a pain and would take so long that you might as well just learn R. And hey, in the age of LLMs learning to code is far easier than it used to be just a couple of years ago. I bet you could get those MC simulations to work in Python with zero prior knowledge of Python in less than a day if you had ChatGPT to help you.
This report was amazing. More people should know about this problem in academia. There are a lot of problems with commercial data sources as well.
Next time a university student thinks they know more than you, just remember this video.
Just proves she was hired primarily for the color of her skin. Not near enough prior ascending Job history that would certainly have revealed this long before the public embarassment !
This is the first time I've heard this guy, I'm very impressed with his speaking ability on these topics. This guy needs to go big time!
“There’s lies, damned lies and statistics” - Mark Twain