This video's a little bit personal for me -- because my own Masters thesis completely, unquestionably proved the null hypothesis. I still got my degree... but it hurt, just a little.
I'm with Alex on this one. Tom, you should do a video talking about your thesis, whether it's on this channel or the Park Bech. That would be very interesting.
My favourite negative-result experiment is the Michelson-Morley experiment. It was designed to measure the luminiferous ether, and completely failed to do so. Instead, it ended up leading to the completely rejection of luminiferous ether as a theory, and ultimately to the development of special relativity. Not bad for a "failure".
honestly, with how little we know about the fundamentals of particle physics and the universe, we'd take any new information. An even remotely popular theory being disproved obviously wouldn't be as good as it being shown to be true, but even knowing that one of our theories, for instance the "dark photon" idea, is definitely wrong would be amazing.
fahoudey Remember, this is the Internet, there will be many many people that would read Jan's comment and take it as fact. As much as it ruins the joke, this might be one of those times when it's probably better to ruin it at some point earlier rather than later.
One of my favorite things about your videos is that you don't waste my time with an intro. Straight to the point. Love it and your content. Keep it up!
The info:filler ratio is high. I've given up on some channels that cover interesting topics, because swimming through their sea of filler is too exhausting.
The under-publishing of negative results is a serious and worrying problem. Surely there ought to be better regulations in place to make sure that all valid, completed experiments are published?
I agree, I imagine that there are plenty of experiments where they get bizarre results and have conversations along the lines of "...well we can't possibly use this data." Science works well but only if you refuse to question some of the fundamental axioms that science uses.
Many top journals won't publish negative results or replication studies, unfortunately. I'm not sure how you could force them to accept papers they don't want to publish. Open access would be a good way to go, like with conditions on grants that mandate open access publishing of any results (positive or negative) that come from the grant.
There's a thing that people are pushing right now, called pre-registration. Basically, you "pre-register" your experiment and experiment design with a journal, then as long as you conduct your study right and rigorously, your work will be published no matter what is the result. And you upload all of your data: raw, processed, everything for sharing and double-checking. That's the general idea of how science "should" publish. Otherwise we have what we have right now: a glut of unreplicable results.
Wouldn't this also create the problem where different scientists perform similar tests because they don't know that the test had been performed in the past and had negative results?
RossH There's a lot of incredible content creators! At least 1 out of every 1000 content creators are actually good at what they do, and there's millions of them so statistically there are several thousands of good channels out there. I know I've found many!
I swear, your videos are some of the only ones I genuinely LOVE seeing pop up in my recent videos. They're fascinating, easily digested, and even if it's something I already knew, a ton of fun to watch. I don't think you've got a single 5-10 minute video I haven't watched to completion. Keep up the fantastic work, Tom.
Reminds me of the public distaste for NASA after the faulty hubble telescope. It annoyed me so much when I heard everybody just started making fun of NASA who worked so hard and almost completed perfectly it but the tiniest error caused people to ignore all the success in that mission.
I think that was a different situation. Its not like an idea was tested and failed/succeeded advancing science. An avoidable error was made in machining the lens which resulted in more money having to be spent on an already very expensive project just to correct the error (or let the whole project die).
@@stiimuli So what. The money was spent here on Earth and went to circulation creating jobs etc. It was not wasted. Wasted money is the one that only stays on a bank account earning more money to sit on the bank account. And the pure material cost of satellites are miniscule compared on the amount paid for the research and manufacture. And money on reaseach and manufacture stays on Earth and creates more jobs.
Tom Scott Thank you for responding at all. I just thought that the clip should have made the main cut and not the Behind the scenes. I will go and watch the Behind the scenes now, thank you for telling me about that. Edit: that’s the park bench though and I saw that one already. That’s why I know of its existence. I will rewatch it anyways, because I like that footage.
I belive this one of your greatest videos for three reasons: - It discusses Science - It is about Space or at least Zero-G - And it shows women in science. Thanks so much!
The idea that negative results are less 'flashy', is the reason why why a lecturer at my university helped to set up the 'Journal for Negative Results'.
I remember the pain of my high school graduation project (in biology, studying a bactericidal compound) dealing with the null hypothesis. Wasn’t easy when you have to balance a mini research project and 5 other subjects. My results were tainted by poor methodology so I could neither disprove my null hypothesis or confirm it.
This is an incredibly important video, and I'm glad you made it, Mr. Scott. We non scientists have to celebrate results, sure, but also think about everything else that didn't went anywhere, but with the fact that they didn't went anywhere being an important data point to think about. It could be compared to the Higgs Boson going faster than light and that result ending up being an error, or even people watching a juggler playing around with 10 plates with ease, those people not knowing the amount of broken porcelain that came out before what they are seeing.
I heard someone talking about a journal that exclusively publishes negative results. I'm not sure if this exists or someone was suggesting that it ought to, but I think if it doesn't then it should. I wonder how much research time is wasted as research teams explore one avenue that another team followed earlier, found it nonviable, but didn't publish their conclusions as a warning to others. Plus, there's every chance that a new technique you tried might not have helped your own work but could be useful to someone else.
Would be especially useful if the large drug companies data could be made public with every drug they synthesise then test. Ben Goldacre's work on the subject has been great showing up these companies hiding bad results.
Yes! Thank you for spreading the word on publication bias; it's genuinely a huge problem in any field which publishes, and not one which is talked about often enough for my liking. Maybe one day these papers which prove the null will get exposure, and we'll all learn that learning about a dead end is still learning. Who knows. Also gg to those scientists - glad they got the results they wanted in the end!
The ability to control something with your mind through a brain interface would improve so many lives, and indeed save a few too. Imagine an amputee with a prosthetic that moved exactly the way a person would normally move their limbs, or being able to use heavy machinery remotely via an interface that moves a robotic arm.
I'm studying in Arnhem, so I'm acutely aware that the short hop across the rivers to Nijmegen would make me live in a place with an astounding number of very beautiful people. However, because of the unique way the Dutch education system separates universities into lower and higher tier ones (only the latter officially being called universities), I'm not able to study there. Still, not unhappy; Nijmegen may have the better people, but Arnhem is the better city. :p
+Ror shach - Interesting. Our higher-tier universities are simply called University in English and Universiteit in Dutch, whereas the lower-tier ones are called University of Applied Sciences in English and Hogeschool in Dutch, which literally translates to "High School", so you can probably tell why the official English name isn't a direct translation of the Dutch one...
I assume you mean "straight, single guys". If so, guilty as charged. Thanks for not being crude about it though. Their experiment is what really fascinated me, though.
I dream of the day I will take a newspaper an see the headline : "Breaking news : Scientists have just replicated a 1984 experiment on the classification of simplified sequence protein folds. The results held up!"
This is very true, We are continually taught the ways that media perceives and the way that journals perceives findings. I'm in Archaeology, if you haven't found fascinating and groundbreaking finds then you aren't publicized. Also at the same time in Bio anthropology, there's constant in fighting about species and the political motivations behind publishing in science or in a free to view journal. It's a real issue, I think. People have a right to forward their own careers but not at the expense of actual scientific progress! Love the video dude!
Knowing how this falsifiability is central to all science makes established scientific knowledge even more amazing to me. Usually, the best you can say for any piece of evidence is "this agrees with our theory" but that doesn't mean it can't also agree with an alternate theory or competing explanation or just be a fluke. It's another thing entirely to say "we've all spent a century diligently trying our best over and over and over and over to disprove this relativity thing and despite all that literally no one has ever managed to do so." It makes my brain have a real *Keanu WHOA moment*
In the example you give: the null hypothesis was actually the one they were hoping for. Because it's defined in terms of difference between experiment and control, rather than whether or not it's the result you're trying to prove.
I think we, as people, often tend away from experiments that we expect to fail to disprove the null hypothesis. But, sometimes learning from how and why it fails, even when it's expected to, can bring us closer to understanding how to succeed.
In science disproving something is just as good as proving something. It is a step forward either way. If you have an idea and it turns out not to work or be true then that is good for science as you know that route does not work. It can help stop other scientist going down that blind alley, it overall improves science even if it does not forward your own career.
This video is really important. More opportunities for publication of null results must be found if science (especially Psychology) is to gain some of it's reputation back.
Oh my god, is this really finally video games that you control with your MIND instead of a controller!! Haven't we all been waiting our whole lives for this!@!#
Null hypothesis. That is a mathematical concept to ascertain the difference between result sample data sets as "the same" or "statistically different" (indicating that the discrete variable between the data sets was the cause of the variation). It has nothing to do with zero gravity or mind reading, but it does look fun (with a P value of 0.71).
I would like to see a proper Journal of Failure. It could serve as a repository of failed experiments. To be interesting, what the journal would publish a response to articles or topics that are in vogue. This would be done to either cast doubt on certain findings or to give a broader picture to other researchers in that field of what works.
SIX-SH00T3R, it's the logical thing to do, and if you can graph their performance versus time for the entire flight and (after some slight smoothing to hide the slight distraction of transitioning between zero and 1.8 g) you can identify when the parabolas were, it would show that it's more than just statistically significant.
It would probably be a better experiment to make use of a centrifuge, already used by astronauts in their training, and have the same subjects try to play the game. It'd produce a higher G-force consistently, so would be a more accurate test.
Seriously, more people, and even more scientist and publishers need to understand this. Research done is always valuable, whether it has some amazing pop culture worthy results or not. It does not even have to be novel; to repeat experiments from others or to put their research into new context (like testing their findings on fruit flies for houseflies) can be just as valuable. Sadly, studying Computer Vision, I think there is not much to be done on a parabolic flight for me that can't be done in 1g.
Hey look Dutch people doing sciency things. They speak English remarkably well for being Dutch. I can't hear any "Dutch-English" as you usually hear us talk.
I mean in science almost everything is english so its important to speak it well. i can still hear quite clearly that they´re dutch though, their grammar is good enough but still doesn´t pass as native speaker. and the accent is qutie obvious :P
+Mister Hat - Absolutely, and the Netherlands leads the way in that; there is no country in the world where English is not an official or historical language with a greater percentage of English speakers than the Netherlands. Percentage-wise, more Dutch people speak English than Canadians. However, most Dutch people don't pay as much attention to their accents as these women clearly did. Their accents aren't perfect, but they're definitely way above average for Dutch people...
Pure research doesn't care about "failed experiments" only that you follow your curiosities and ask questions. Applied research will eventually be the death of science if we can't let go of "we need 'x' result" or "let's focus on making blank a reality" when those things may not be possible! We won't know if something is possible or not without pure research, and sometimes that leads to proving the null hypothesis, and that's ok. It still has expanded our base of information, and will, hopefully, lead to more questions. That is what science is all about after all., and why I love it so. So go follow curiosities, and report those "failures". They may lead to questions you wouldn't have known to ask otherwise.
The Inquiry had a recent episode on it. The short version is, yes there is a bias towards flashy findings and it is actually hurting scientific advancement since a lot of stuff that gets published is plain wrong because it was rushed by the scientists and not properly checked by the publishing journasl.
John Oliver did a great segment on it as well in Last Week Tonight, highlighting the financial incentives for only doing exploratory studies and not spending time on confirming hypotheses.
In this case both outcomes are publishable. 1: you can use brain controllers in 0g or 2: the brain releases different signals in 0g. Both are interesting and publishable
Henning Metzger he kept it there though. I don’t understand, how it didn’t make the main cut. The behind the scenes links to the park bench video as well. At least it made the cut there. I am still a little bit dissapointed that he cut it out for the main video.
My Masters thesis (many years ago) was about how badly ancient civi;lizations were affected by epidemics and pandemics. My conclusion was that they weren't (except maybe Justinian's Plague). I proved the null hypothesis... which is good news in 2021.
This is a deeply rooted issue. I'd like to think that academic journals and the media are to blame. It would be great if there was more funding for "boring" studies- stuff like proving a null hypothesis, or trying to reproduce the results of other studies.
When people say that science and religion don't go together, and in fact, are polar opposites, this is what they reference. You cannot assume a hypothesis to be true unless proven, for disproving a hypothesis devoid of evidence is impossible; You need evidence to disprove an evidence-less hypothesis. Religion assumes something to exist without any ability to prove it does. All forms of science, with the partial exception of those built around discovery (theoretical physics, for instance), are based on the concept of proof before belief. Nothing is true, everything can be proven.
Melody = O Tannenbaum : The null hypothesis, the null hypothesis; / It tells you nothing worth mentioning. / But you can get your name in lights / if you reject the null hypothesis. // A number from a sample is / compared to a criterion / and if the sample's too extreme / You can reject the null hypothesis!
I have three published papers that were all null-hypothesis provers, and two of the three took many different submissions before finding journals willing to publish them. Pretty unfortunate, because I think (biasedly, of course), that my conclusions were very valuable
I was pretty distracted for most of the video. Once I saw that guy use a computer without physical controls, all I could think about was its possible use for VR.
It's actually tremendously useful in creating solutions for people who have damage to their peripheral nervous system. The most famous example of this is Stephen Hawking, who uses similar brain wave-reading technology to be able to speak.
+Robert Faber Really? Since when? I was under the impression it was something with his glasses and eye-tracking, and I have trouble believing it's a "brain wave-reading technology" (mostly because, if it is, it'd probably be frustrating due to noise and could only work if it was using his brain to move a cursor or somesuch)
Fun fact, the Dutch language doesnt have the word Teach, we only have Learn and use it for both cases. Our word for Teacher sounds more like Learner, and despite their best efforts we sometimes slip up, like the student at 01:03
More people need to approach things with the null hypothesis. This age where people can try to foist untested and unproven ideas on the world needs to end.
This video's a little bit personal for me -- because my own Masters thesis completely, unquestionably proved the null hypothesis. I still got my degree... but it hurt, just a little.
Tom Scott Do any of your videos talk about your thesis?
Failure is just success rounded down, my friend!
As you said, it's a step forward.
The only failed experiment is one that doesn't show anything one way or the other.
I think the classic Rutherford experiment is a VERY good example of this.
The experiment gave a very different result than they thought it would give.
I'm with Alex on this one. Tom, you should do a video talking about your thesis, whether it's on this channel or the Park Bech. That would be very interesting.
So the plane was full of scientists, researchers...and some dude screaming to a camera XD
he was even in his own "crib"
i think he said something abou there also being a swedish tv team arround
close, Danish.
Pro-science publicity arguably is even more important than research I think. Within reason, of course.
UrbanTarzan Duh It's not about the research; it's about people thinking we do the research.
My favourite negative-result experiment is the Michelson-Morley experiment. It was designed to measure the luminiferous ether, and completely failed to do so. Instead, it ended up leading to the completely rejection of luminiferous ether as a theory, and ultimately to the development of special relativity. Not bad for a "failure".
And according to flat earthers it proved the earth is flat.. or something, somehow
If earth is round, the Michelson-Morley experiment would be round too.
Just like your house
/s
honestly, with how little we know about the fundamentals of particle physics and the universe, we'd take any new information. An even remotely popular theory being disproved obviously wouldn't be as good as it being shown to be true, but even knowing that one of our theories, for instance the "dark photon" idea, is definitely wrong would be amazing.
and don't flat earthers try to twist it somehow as proof of their theory? wild
As someone who lives here, it is my duty to point out that this happened in Cleveland.
I can't believe people are still talking about the null hypothesis. I remember reading a study that disproved it years ago.
Jan Sten Adámek that's a joke comment right?
Of course. IIRC xkcd is the original author.
Luke96w come on, you ruined the joke
fahoudey Remember, this is the Internet, there will be many many people that would read Jan's comment and take it as fact. As much as it ruins the joke, this might be one of those times when it's probably better to ruin it at some point earlier rather than later.
I read that paper too. But it was all very abstract.
One of my favorite things about your videos is that you don't waste my time with an intro. Straight to the point. Love it and your content. Keep it up!
The info:filler ratio is high. I've given up on some channels that cover interesting topics, because swimming through their sea of filler is too exhausting.
Haha, I like how you can see the attempted personal flip record in the back
That's Melanie, who was camera operator for my sections! The full video of her flip's over on the behind-the-scenes video :)
#parkbench... :)
The under-publishing of negative results is a serious and worrying problem. Surely there ought to be better regulations in place to make sure that all valid, completed experiments are published?
*+*
I agree, I imagine that there are plenty of experiments where they get bizarre results and have conversations along the lines of "...well we can't possibly use this data." Science works well but only if you refuse to question some of the fundamental axioms that science uses.
Many top journals won't publish negative results or replication studies, unfortunately. I'm not sure how you could force them to accept papers they don't want to publish. Open access would be a good way to go, like with conditions on grants that mandate open access publishing of any results (positive or negative) that come from the grant.
There's a thing that people are pushing right now, called pre-registration. Basically, you "pre-register" your experiment and experiment design with a journal, then as long as you conduct your study right and rigorously, your work will be published no matter what is the result. And you upload all of your data: raw, processed, everything for sharing and double-checking.
That's the general idea of how science "should" publish. Otherwise we have what we have right now: a glut of unreplicable results.
Wouldn't this also create the problem where different scientists perform similar tests because they don't know that the test had been performed in the past and had negative results?
Actually one of the only good content creators left on TH-cam. Much love from Scotland.
RossH There's a lot of incredible content creators! At least 1 out of every 1000 content creators are actually good at what they do, and there's millions of them so statistically there are several thousands of good channels out there. I know I've found many!
Scottland
Always showing up on the weirdest places on weird time frames wearing the same shirt
are we sure Tom isn't a cartoon character?
Incendere 🤔🤔🤔
there’s multiple cartoon versions of him too
everyone in the plane dead serious doing experiments and Tom in the back having the best time of his life flying
I know I am a bit late, but...
I swear, your videos are some of the only ones I genuinely LOVE seeing pop up in my recent videos. They're fascinating, easily digested, and even if it's something I already knew, a ton of fun to watch. I don't think you've got a single 5-10 minute video I haven't watched to completion. Keep up the fantastic work, Tom.
Reminds me of the public distaste for NASA after the faulty hubble telescope.
It annoyed me so much when I heard everybody just started making fun of NASA who worked so hard and almost completed perfectly it but the tiniest error caused people to ignore all the success in that mission.
I think that was a different situation.
Its not like an idea was tested and failed/succeeded advancing science. An avoidable error was made in machining the lens which resulted in more money having to be spent on an already very expensive project just to correct the error (or let the whole project die).
@@stiimuli So what. The money was spent here on Earth and went to circulation creating jobs etc. It was not wasted. Wasted money is the one that only stays on a bank account earning more money to sit on the bank account.
And the pure material cost of satellites are miniscule compared on the amount paid for the research and manufacture. And money on reaseach and manufacture stays on Earth and creates more jobs.
2:08 I see they've begun tests on the human-fidget spinner hybrid.
Why is the epic slow mo walk not in this video?
It's in the behind-the-scenes video, linked in the end cards and the description!
Tom Scott Thank you for responding at all. I just thought that the clip should have made the main cut and not the Behind the scenes. I will go and watch the Behind the scenes now, thank you for telling me about that.
Edit: that’s the park bench though and I saw that one already. That’s why I know of its existence. I will rewatch it anyways, because I like that footage.
I belive this one of your greatest videos for three reasons:
- It discusses Science
- It is about Space or at least Zero-G
- And it shows women in science.
Thanks so much!
The idea that negative results are less 'flashy', is the reason why why a lecturer at my university helped to set up the 'Journal for Negative Results'.
I remember the pain of my high school graduation project (in biology, studying a bactericidal compound) dealing with the null hypothesis. Wasn’t easy when you have to balance a mini research project and 5 other subjects.
My results were tainted by poor methodology so I could neither disprove my null hypothesis or confirm it.
This is an incredibly important video, and I'm glad you made it, Mr. Scott. We non scientists have to celebrate results, sure, but also think about everything else that didn't went anywhere, but with the fact that they didn't went anywhere being an important data point to think about. It could be compared to the Higgs Boson going faster than light and that result ending up being an error, or even people watching a juggler playing around with 10 plates with ease, those people not knowing the amount of broken porcelain that came out before what they are seeing.
That Fly Your Thesis program sounds great! But I doubt I could ever justify any zero gravity research in Archaeology :p
"Does cleaning artifacts in zero-g produce fewer faults?"
Veritasium has a great video on a similar subject: "Is Most Published Research Wrong?" if you want to know more.
Of course it has a more attention-grabbing title
1:15 that name made me do a double take
On a scale of Donald Trump to Danielle Tump, how smart are you?
Same
it would have to be a logarithmic scale
+teslaTrooper - What logarithmic scale are we talking? 2log? 10log?
+Robert Faber Doesn't matter, they're proportional.
Words cannot describe how happy the ending made me.
I heard someone talking about a journal that exclusively publishes negative results. I'm not sure if this exists or someone was suggesting that it ought to, but I think if it doesn't then it should. I wonder how much research time is wasted as research teams explore one avenue that another team followed earlier, found it nonviable, but didn't publish their conclusions as a warning to others. Plus, there's every chance that a new technique you tried might not have helped your own work but could be useful to someone else.
There are a few, yeah. Though they arent as well known. Still, generally a good idea, and probably worth it if you want to publish something. :)
Would be especially useful if the large drug companies data could be made public with every drug they synthesise then test. Ben Goldacre's work on the subject has been great showing up these companies hiding bad results.
I'm glad that he told us the results in the end, I'm happy! And very excited for the future of science.
This video is fascinating! Really proud of all those hard working students working towards innovation.
It's funny to watch this as a neuroscientist working in a lab which also develops BCIs :-D
Great work, guys!
Yes! Thank you for spreading the word on publication bias; it's genuinely a huge problem in any field which publishes, and not one which is talked about often enough for my liking. Maybe one day these papers which prove the null will get exposure, and we'll all learn that learning about a dead end is still learning. Who knows.
Also gg to those scientists - glad they got the results they wanted in the end!
Major Tom to ground control
...who's floating in a most peculiar way.
tom scott, you never cease to amaze me!
Hey Tom! Bring back your Language/Linguistics videos! I love them!
The ability to control something with your mind through a brain interface would improve so many lives, and indeed save a few too. Imagine an amputee with a prosthetic that moved exactly the way a person would normally move their limbs, or being able to use heavy machinery remotely via an interface that moves a robotic arm.
Based on this rather tiny sample size, how many guys watching this are now thinking "I want to study neurology in The Netherlands now"?
Povl Besser I live in The Netherlands, but I know what you mean. I am thinking of studying in Nijmegen, so...
I'm studying in Arnhem, so I'm acutely aware that the short hop across the rivers to Nijmegen would make me live in a place with an astounding number of very beautiful people. However, because of the unique way the Dutch education system separates universities into lower and higher tier ones (only the latter officially being called universities), I'm not able to study there. Still, not unhappy; Nijmegen may have the better people, but Arnhem is the better city. :p
I am mostly just sad that Tom was in Nijmegen but didn't even make a video about the giant magnet that we have which can levitate a frog.
+Ror shach - Interesting. Our higher-tier universities are simply called University in English and Universiteit in Dutch, whereas the lower-tier ones are called University of Applied Sciences in English and Hogeschool in Dutch, which literally translates to "High School", so you can probably tell why the official English name isn't a direct translation of the Dutch one...
I assume you mean "straight, single guys". If so, guilty as charged. Thanks for not being crude about it though. Their experiment is what really fascinated me, though.
Absolutely brilliant way to end the video, Tom. You made the point perfectly!
Yes!!! Neuroscience, women in STEM, discussion of publication bias... this is everything I could possibly want in a video!
2:10 world record attempt in background
I dream of the day I will take a newspaper an see the headline : "Breaking news : Scientists have just replicated a 1984 experiment on the classification of simplified sequence protein folds. The results held up!"
Djorgal Haha, every day that goes by you're less likely to see something like that.. but maybe.. maybe one day.
Your videos give me back hope for critical thinking and humanity. No small feat. Thx!
Thanks, Tom! Currently doing an assignment on research and this is an area I need go cover (((:
Happy 1,000,000 subscribers Tom!
Two amazing videos shot at the same time. You just keep surprising.
This is very true, We are continually taught the ways that media perceives and the way that journals perceives findings. I'm in Archaeology, if you haven't found fascinating and groundbreaking finds then you aren't publicized. Also at the same time in Bio anthropology, there's constant in fighting about species and the political motivations behind publishing in science or in a free to view journal. It's a real issue, I think. People have a right to forward their own careers but not at the expense of actual scientific progress! Love the video dude!
Damn the ladies working on brainfly are pretty fly
True
I thought I'm the only one that noticed
Great video as always Tom
Knowing how this falsifiability is central to all science makes established scientific knowledge even more amazing to me. Usually, the best you can say for any piece of evidence is "this agrees with our theory" but that doesn't mean it can't also agree with an alternate theory or competing explanation or just be a fluke. It's another thing entirely to say "we've all spent a century diligently trying our best over and over and over and over to disprove this relativity thing and despite all that literally no one has ever managed to do so." It makes my brain have a real *Keanu WHOA moment*
In the example you give: the null hypothesis was actually the one they were hoping for. Because it's defined in terms of difference between experiment and control, rather than whether or not it's the result you're trying to prove.
Amazing.. UK's (industries and government) support for academia is enormously awesome and notable!
I think we, as people, often tend away from experiments that we expect to fail to disprove the null hypothesis. But, sometimes learning from how and why it fails, even when it's expected to, can bring us closer to understanding how to succeed.
In science disproving something is just as good as proving something. It is a step forward either way. If you have an idea and it turns out not to work or be true then that is good for science as you know that route does not work. It can help stop other scientist going down that blind alley, it overall improves science even if it does not forward your own career.
That moment when you realize the people watching TH-cam will forget about this as soon as the clip has finished. :)
You sound like veritasium.
This video is really important. More opportunities for publication of null results must be found if science (especially Psychology) is to gain some of it's reputation back.
Awesome video as usual! I was about to get upset if you didn't tell us the result at the end haha thank you!
WOW! I actually didn't want to see that very last scene :( Very nice video, Tom!
Best message ever for not giving up when you fail.
Oh my god, is this really finally video games that you control with your MIND instead of a controller!! Haven't we all been waiting our whole lives for this!@!#
Nah, I want true VR that you use your body to control. If you just use your brain, your body wastes away, but full body would be exercise.
Null hypothesis. That is a mathematical concept to ascertain the difference between result sample data sets as "the same" or "statistically different" (indicating that the discrete variable between the data sets was the cause of the variation).
It has nothing to do with zero gravity or mind reading, but it does look fun (with a P value of 0.71).
I would like to see a proper Journal of Failure. It could serve as a repository of failed experiments. To be interesting, what the journal would publish a response to articles or topics that are in vogue. This would be done to either cast doubt on certain findings or to give a broader picture to other researchers in that field of what works.
failed as in... what? the experiment (equipment) failed, or they proved themselves wrong?
Proved themselves wrong.
Just out of curiosity, did they happen to look at the effects of increased gravity on brain function on the lower parabolas?
Now that you say it, I am quite curious about that as well...
SIX-SH00T3R, it's the logical thing to do, and if you can graph their performance versus time for the entire flight and (after some slight smoothing to hide the slight distraction of transitioning between zero and 1.8 g) you can identify when the parabolas were, it would show that it's more than just statistically significant.
It would probably be a better experiment to make use of a centrifuge, already used by astronauts in their training, and have the same subjects try to play the game. It'd produce a higher G-force consistently, so would be a more accurate test.
@@trojanhorse62 Coming into this 2 years later...ever hear of anybody doing this?
I had publication bias even in elementry school for the science fair.
Seriously, more people, and even more scientist and publishers need to understand this. Research done is always valuable, whether it has some amazing pop culture worthy results or not. It does not even have to be novel; to repeat experiments from others or to put their research into new context (like testing their findings on fruit flies for houseflies) can be just as valuable.
Sadly, studying Computer Vision, I think there is not much to be done on a parabolic flight for me that can't be done in 1g.
Hey look Dutch people doing sciency things.
They speak English remarkably well for being Dutch. I can't hear any "Dutch-English" as you usually hear us talk.
Vel0city she did say "learn a computer" instead of "teach a computer" around the 1:00 mark which gave it away to me immediately.
Norwegians say this too. They even look kind of similar.
I mean in science almost everything is english so its important to speak it well. i can still hear quite clearly that they´re dutch though, their grammar is good enough but still doesn´t pass as native speaker. and the accent is qutie obvious :P
I'm not too surprised. In most developed countries it's common to learn English as a second language at a young age.
+Mister Hat - Absolutely, and the Netherlands leads the way in that; there is no country in the world where English is not an official or historical language with a greater percentage of English speakers than the Netherlands. Percentage-wise, more Dutch people speak English than Canadians. However, most Dutch people don't pay as much attention to their accents as these women clearly did. Their accents aren't perfect, but they're definitely way above average for Dutch people...
Pure research doesn't care about "failed experiments" only that you follow your curiosities and ask questions. Applied research will eventually be the death of science if we can't let go of "we need 'x' result" or "let's focus on making blank a reality" when those things may not be possible! We won't know if something is possible or not without pure research, and sometimes that leads to proving the null hypothesis, and that's ok. It still has expanded our base of information, and will, hopefully, lead to more questions. That is what science is all about after all., and why I love it so. So go follow curiosities, and report those "failures". They may lead to questions you wouldn't have known to ask otherwise.
This is unusual, an intelligent video in the trending section...
Please do a followup with the thesis results from the plane, could be a lot of fun
Thanks for covering the probably biggest issue across science; the publication bias and it's consequences.
The Inquiry had a recent episode on it. The short version is, yes there is a bias towards flashy findings and it is actually hurting scientific advancement since a lot of stuff that gets published is plain wrong because it was rushed by the scientists and not properly checked by the publishing journasl.
John Oliver did a great segment on it as well in Last Week Tonight, highlighting the financial incentives for only doing exploratory studies and not spending time on confirming hypotheses.
Did not expect to see my Radboud University come up when watching this video 😊
Are you going to have a laugh at me for studying at HAN in Arnhem? Radboud students usually do... :p :p :p
Thank you so much for doing this video! There are more journals now in favour of publishing null and negative results, but it's slow going.
In this case both outcomes are publishable. 1: you can use brain controllers in 0g or 2: the brain releases different signals in 0g. Both are interesting and publishable
Where is the clip of you walking away from the plane in slow motion, you can't keep that to the parkbench 😂
Henning Metzger yes plain
Henning Metzger he kept it there though. I don’t understand, how it didn’t make the main cut. The behind the scenes links to the park bench video as well. At least it made the cut there. I am still a little bit dissapointed that he cut it out for the main video.
madkillah 423 that's what happens when your German auto correct tries to learn English
As great of a shot it is, there's not really anywhere you can slot it in this video that hasn't already been taken up by something else.
B-Bud why not do an epic outro and roll some credits just for fun 😂
Two videos out of one trip? That's efficiency, baby
Hopefully he makes 4, the Park bench already had footage from this trip as well.
He had to go to Nijmegen as well, not just Bordeaux, so it's still two videos out of two trips.
+Robert Faber
True, very true.
2:06 They are doing the "most backflip" thing Tom talked about on the park bench. Just saying.
My Masters thesis (many years ago) was about how badly ancient civi;lizations were affected by epidemics and pandemics. My conclusion was that they weren't (except maybe Justinian's Plague). I proved the null hypothesis... which is good news in 2021.
This is a deeply rooted issue. I'd like to think that academic journals and the media are to blame. It would be great if there was more funding for "boring" studies- stuff like proving a null hypothesis, or trying to reproduce the results of other studies.
Can’t believe they’re still teaching the null hypothesis. I read a study debunking it years ago.
Oooooh, if you were in Nijmegen we might get a video on some big and powerful magnets?
You're on trending! In the US at least :^)
Very well explained Tom.
Brain controlled interfaces are so freaking cool oh my god
I wish I saw this yesterday, BEFORE my AP Bio test (which was today)
What's the best way to find their papers once they release them? I've always found this to be a fascinating field, and need to learn more about it.
1:22 That brain massage looks smooth. I want it too!
Love your videos Tom.
When people say that science and religion don't go together, and in fact, are polar opposites, this is what they reference.
You cannot assume a hypothesis to be true unless proven, for disproving a hypothesis devoid of evidence is impossible; You need evidence to disprove an evidence-less hypothesis.
Religion assumes something to exist without any ability to prove it does.
All forms of science, with the partial exception of those built around discovery (theoretical physics, for instance), are based on the concept of proof before belief.
Nothing is true, everything can be proven.
hum
I take this as a measure of Toms audience that this comment hasn’t blown up into a massive flame war...
I can't believe they keep teaching kids about the null hypothesis -- I saw a big study disproving it a few years ago.
Great video as usual.
Thank you
im also intersted about the human sign spinner in 2:08
That's really cool, didn't know brain interfaces were anything near that in outcomes.
You need to do a separate video on these research papers. Probably one on each
Here's how you spot a Dutch person: they say "learn" when they should say "teach".
Do you use the same word for those two in Dutch?
That's exactly what French Canadians do too.
Dude.. You were at my university! When were you there?
Approximately 3 years ago :)
The point of science is to settle arguments, prove one guy right and the other guy wrong in a final, inarguable way. Scientific facts.
Melody = O Tannenbaum : The null hypothesis, the null hypothesis; / It tells you nothing worth mentioning. / But you can get your name in lights / if you reject the null hypothesis. // A number from a sample is / compared to a criterion / and if the sample's too extreme / You can reject the null hypothesis!
The null hypothesis is the best thing to ever exist in science, as it is a nicer way of saying, "You dun goofed up!"
I have three published papers that were all null-hypothesis provers, and two of the three took many different submissions before finding journals willing to publish them. Pretty unfortunate, because I think (biasedly, of course), that my conclusions were very valuable
It's strange because negative results are just as important. People need to know so we can cut out a lot of the "they should"s and "why aren't they"s.
I was pretty distracted for most of the video. Once I saw that guy use a computer without physical controls, all I could think about was its possible use for VR.
Get ready for real SAO
It's actually tremendously useful in creating solutions for people who have damage to their peripheral nervous system. The most famous example of this is Stephen Hawking, who uses similar brain wave-reading technology to be able to speak.
+Robert Faber Really? Since when? I was under the impression it was something with his glasses and eye-tracking, and I have trouble believing it's a "brain wave-reading technology" (mostly because, if it is, it'd probably be frustrating due to noise and could only work if it was using his brain to move a cursor or somesuch)
I was distracted by something else.
These videos are more addictive than Pringles
Fun fact, the Dutch language doesnt have the word Teach, we only have Learn and use it for both cases. Our word for Teacher sounds more like Learner, and despite their best efforts we sometimes slip up, like the student at 01:03
'Leren' is the Dutch word for learn and teach, 'Leraar' is a teacher. I make the same mistake every so often.
It's the same with many other germanic languages as well. In norway, sweden and denmark it's exactly like that too.
You were like 100 meters away and I didn't meet you :( Do a meetup next time. You got a lot of dutch viewers, especially in Nijmegen!
More people need to approach things with the null hypothesis. This age where people can try to foist untested and unproven ideas on the world needs to end.