Extra footage: ua-cam.com/video/9NIvKzAgDGg/v-deo.html Note from Phil: A crude and basic introduction to the idea of error bars is covered in the video. No mention is made of, for example, the effect of repeated measurements on reducing the standard error, adding in quadrature, or of confidence limits. For a slightly more detailed discussion of experimental uncertainty, see ua-cam.com/video/ThP51oPttS0/v-deo.html
An oft overlooked concept to be honest. Science is always trying to refine its theories, and the best way is to find the error in measurement over repeated trials. Sounds horrible, though, Shakes. I hope the next mission doesn't have that happen. >.>
When I was a kid, I always thought science was about exactness - like numbers to the 15th decimal place. But as I grew up, I began to realise that it was actually about measuring inexactness.
As a ham radio operator I've built dozens of antennas over the years so I'm very familiar with the formula for converting frequency to wavelength but I've never seen the equation set up to solve for c. It threw me back for a moment.
you can get them in the us at places like world market and some grocery stores in the euro/international sections, though they are not as widely sold here
Thats not so bad, at least everyone was dressed. My buddy in MOS school absentmindedly put a bowl of ramen in the microwave without water. Got a whole Marine barracks evacuated at 2 AM. Just shy of 500 Marines in their underwear in a parking lot in the middle of freezing desert winter with two hours before the start of a standard 12-14 hour work day. I didn't think he was going to make it through the night.
12:04 Thumps up! Thumps up! Thumps up! I HATE when my students say: "The real value is x, so we are pretty close to being right." Your result is ALWAYS right per definition! It might be uncertain, or you might have measured something different that you tried to measure. But it is never right or wrong.
QUESTION FOR MORAIRTY; How does one manufacture/machine a pinhead of width 1 atom for a Scanning Tunneling Microscope, and then on top of that, have a machine that is capable of making physical adjustments on the scale of individual atomic radii?
Maybe not a very exciting topic, but important nonetheless. But Brady, please use × (Alt+0215) in stead of x as a multiplication sign and ± (Alt+0177) for the plusminus sign. Keep it up!
mpandurov okay, then this vid is reformulated to measure the frequency of the microwave. Same idea, but I think less accessible, they made the right call.
I think it is... Because you can go full ham with error calcs... Like he said, there's an error even in the ruler, which would not only matter when finding the distance between the center of mass between valleys in the chocolate, but even when having to find the center of mass themselves - which you probably can't do without even more errors. And then we have the setup of the microwave itself - how do you know there's not some bouncy effects of waves going on in there :( and so on.
Moriarty is easily one of my Top 3 favorites and I'm happy to see him back! However, I originally came to ask if they updated their Sixty Symbols thumbnail photo, or did I just mis-remember what it looked like?
I did this in my electromagnetism class. It's very tricky when you start involving multiple measurement instruments. And you need to dig through a lot of PDF documents to find the correct error percentiles for the range of measurements you're doing. It's fun, if you can understand it.
that genius! i never thought that those are standing waves. i thought everything gets kinda heated equally! wow thats actually something i was kinda interested in. nice one!
Statistical siginificance only makes sense if you are comparing your observations to a null model/hypothesis. So e.g. you could ask whether the speed of light is the same in this microwave oven as everywhere else. :D (That the true speed of light was in the range of what was predicted by the observations is basically a crude indicator that it doesn't differ significantly from our expected result.)
first of all, i dont really get how is it calculated and im too lazy to make any effort searching, also i think explaining it for dummies would make a great video
I know how much Phil loves the nitpicking criticisms on these videos so I'm not going to mention that the way I remember Hank Green teaching me sig figs on scishow, you don't do any rounding, because of course you're not meant to know what that next figure is. Anyway, you guys are awesome, and I could listen to Dr Phil all day.
This was very interesting. Repeating the experiment was discussed, it would be interesting to discuss if increasing the number of experiments or increasing the precession of the measurement to improve the accuracy of the result. Is the Δ%=(measured - expected) /expected not considered a valid measurement of accuracy?
The +/- 2 value quoted in this video is actually the standard deviation - a direct reflection of the scatter of the measurement data. Error bars are normally determined from the actual inherent uncertainty in the measurements. That is, each quantity or variable in the experiment has an inherent uncertainty associated with it which can depend on many things - including HOW the measurement was taken. For example if measure the diameter of a metal sphere using a micrometer or veneer caliper you will get a different uncertainty than if you measure it using a wooden ruler. The error range used in this video is purely a statistical measure of the standard deviation of the data set - and in this case there were only 3 data points used. It's crude, but I suppose the purpose of the demonstration was to highlight a practical way of measuring the speed of light using a household appliance and basic Physics equation (c=l*f).
I remember having to do this in undergrad lab. But it was a lot more convoluted. We had to take enough measurements that we could justify quoting a probability distribution to our final result. It could easily have happened that all three chocolate bar measurements were closer together. In this video your error bars would have been smaller. What we had to do in lab, is quote a probability that the true value was between two values. For example C = 3.4 x10^9 m/s with a standard deviation of 500 x 10^9 m/s. Meaning that there is a small probability that the true speed of light is X.XX m/s even though we never measured anything near that value. (by the way, we didn't measure the speed of light in lab, it was all thermal and acoustic measurements).
I would like to see a video looking at the mathematics/engineering involving ball bearings and their use. Stuff like calculating the strength of a bearing and other things like energy loss to to friction and whatnot.
If we're doing a full analysis, instead of error bars, we think of the underlying probability distributions and in this case, the distribution of a sum a variables is the convolution of the individual distributions. However, what about the product? Do you just take the log, convolve, and then pull back with an antilog (i.e. exponent)?
"We have 8cm, 6cm, 7cm. Whats the best thing to do here, lets take an average..." Looks at the camera and realizes he needs to do math. "Lets use 7cm" 5:57
The 2 cm are not the error of the mean wavelength but the error of every single measurement! Explanation: The way the error bars of delta_lambda=2 cm were obtained is misleading: it suggests that when you perform more measurements the error cannot become smaller but might even become larger as you will get bigger deviations from the mean with a certain probability. Just wanted to point this out, I see this is a UA-cam video and Phil said himself this was a "crude and basic introduction". And I still think the point of the video was made very clear. But maybe a future Sixty Symbols video could cover how estimated errors in your measurements translate to errors in a derived quantity, e.g. if c=f*lambda and you assume f=f+-delta f and you measure lambda several times each with an error of say 2 cm, how do you obtain the mean, the standard deviation and the error the of the mean of lambda and how do you get the error of c in the end.
Question: What's the purpose of the "relative error" thing? Can't I just convert the units of the error and plug it in the formula for c, replacing c with delta-c and lambda with delta-lambda? It also gives the value 0.49*10^8, which is unsurprising given that it cancels out to be what I'm talking about, delta-c=f*delta-lambda. T=2*pi*sqrt(L/g) Also, what if we were trying to estimate a pendulum's period by measuring L, given pi and g (for simplicity's sake)? deltaT=2*pi*sqrt(deltaL/g) or deltaT=T*deltaL/L?
i freaking love this video. great explanation of fundamentals. so many people get tripped up on the notion of certainty in science. hell yeah. top stuff. would watch again.
I watched the „age of the universe“ video before, and it featured an asymmetrical errorbar ( for the value of the Hubble-constant as measured by „standard-siren“ of 70 +12/-8 km/s/Mparsec). How do you end up with asymmetrical errorbars? Why is one side weighed more? Probably in the details of the measurement, but I‘d like an explanation...
The one thing that blew my mind was the size of "microwaves". If I can measure those with my hand, then they are normal human sized, not micro. We've gotta rename them.
Moriarty Day is a good day. :) One issue with teaching the concept of error bars might be simple terminology. The "error" in error bars make it seems like a bad mistake. Maybe "uncertainty bars" is less confusing?
I think you forgot to allow for the error in the rule in your description of the error calculation though I presume it is small compared to the wavelength. When I worked in a sheet metal shop you had to be careful with the measurements as it was quite common to accidentally chop the end off rules in the guillotine so often the first few centimeters or inches were missing ! The other thing is that the Frequency written on the back may have used the speed of light in it's measurement if so you can't really use the labeled frequency to determine the speed of light.
Are microwave ovens dimensioned to an integer multiple of the wavelength to encourage standing waves? Why? An improbable accident? If uneven intensity distribution is undesirable it would seem trivially easy to avoid by having non-parallel sides and irregularities in the pressing of the steel. The fact that standing waves are present is taken for granted here and I've not heard an explanation.
If you don't choose the width so you get standing waves, the waves would interfere destructively everywhere and mostly cancel out, or you get even weirder uneven patterns. Getting a lot of power into the food is already hard (you need big transformers to get high voltages), so you just deal with the uneven heating by rotating the plate and letting the heat distribute in between heating.
Also, the equal distance between the melted spots is a giveaway (also presumably between cold spots & the walls), so assuming standing waves shouldn't be strictly necessary.
They are actually "dimensioned to an integer multiple", and this is why your microwave tray is programmed to rotate. The standing wave would be heating only certain sections of your food without the rotation!
Will To Win thanks, and to nibblr for the responses, but does this lead to the conclusion that if the microwaves were scattered they would not heat the food? I bet you can both guess the next question: where would the energy go?
my conclusion would have been: the speed of light in chocolate filled with bubbles - c(aerobar) - is (dare I write 'significantly') greater than in vacuum. what implication will this finding have to the future of space travel? oh my! fun aside . great explanation to a problem virtually everybody struggles with at the beginning!
Wait, if you get your error bars from [delta]C/C, then aren't the value of the error bars highly dependent on your measuring scale? If I'm measuring a temperature in Fahrenheit or Celsius, I would get radically different error bars, since the ratio and zero point are different. Doesn't this render the whole exercise vapid and arbitrary?
Some people are entrenched in their ideology to such a fundamentalist level that they cannot separate what somebody says about science from what they say about identity politics.
phil has been a naughty boy outside the context of this channel, and some people won't ever let him live it down. imagine watching a science video where hitler explains quantum superposition, and nothing else. people would be losing their minds about it even though it was simply a video about quantum physics.
Wow, someone actually Godwinned a 60 symbols video. The fact that you went with Hitler as part of your comparison is rather pathetic. Disagreeing with Dr Moriarty is fine. Being mad is fine. But comparing the guy to Hitler? Wow.
Instead of just putting the chocolate bar in as a big block you would probably get more precise results if you separated the bar into its individual small squares of chocolate. (It looked like the bar is a 7x5 grid in which case 35 individual squares.) Put each square on the plate in the same rectangular shape but separate them just slightly so they're not touching each other as some of them melt. That way the heat from one square won't be able to easily transfer to a neighboring square and the main source of energy to melt a square will come from the microwave itself, not from convection within the chocolate. Cool experiment though. :)
This makes me wonder if a microwave constructed with non-parallel sides would eliminate standing waves, especially since they don't have that rotating metal reflector in the ceiling any more.
Hershey's did (not sure if they still do) something like this with their chocolate bars. They called them "air delight" over here in America. Clever way to save the amount of chocolate per bar, but sell it for the same price, and trick the human brain into it tasting better. ;) Im suprised Aero doesnt have a distribution center here in the US. They would sell like hotcakes.
Extra footage: ua-cam.com/video/9NIvKzAgDGg/v-deo.html
Note from Phil: A crude and basic introduction to the idea of error bars is covered in the video. No mention is made of, for example, the effect of repeated measurements on reducing the standard error, adding in quadrature, or of confidence limits. For a slightly more detailed discussion of experimental uncertainty, see ua-cam.com/video/ThP51oPttS0/v-deo.html
a crude and basic introduction from a crude and basic man, then
GroovingPict What is the problem with Phil?
I hope they're joking, Phil is my favourite speaker on 60Symbols.
Note: In the first taping they used Mars Bars, but there was a mission failure due to a unit conversion snafu.
An oft overlooked concept to be honest. Science is always trying to refine its theories, and the best way is to find the error in measurement over repeated trials.
Sounds horrible, though, Shakes. I hope the next mission doesn't have that happen. >.>
Brady: "How do you know anything?"
prof Moriarty: "Exactly"
*vsauce music starts playing*
You had me at 'measuring the speed of light with chocolate'.
Yes, I love when Phil does a video related to experimental physics. Thanks Brady, keep it up!
When I was a kid, I always thought science was about exactness - like numbers to the 15th decimal place. But as I grew up, I began to realise that it was actually about measuring inexactness.
@2:45 As a theoretical physics graduate, this is what I imagine experimental physics to be like.
Could be the start of a new series called "Ghetto methods to measure the physical constants"
Assuming this chocolate bars aren't too expensive...
Gentel Noober i know (although the reference i had in mind was more Chem player ) but it would be a nice follow up from math to physics.
Yes, I'd love to see a video series where people got important science results using cheap materials that could be done at home.
I really wanna see what kind of ghetto watt balance they come up with.
As a ham radio operator I've built dozens of antennas over the years so I'm very familiar with the formula for converting frequency to wavelength but I've never seen the equation set up to solve for c. It threw me back for a moment.
This is such a fundamental concept to understand in science. I find the importance of this so difficult to communicate to my non-physics colleagues.
Just tell them it's yummy!
Brady: "How do you know .. anything?"
Moriarty: "Exactly"
This is a good practical video about physics and science in general. Thank you much appreciated!
Melting chocolate bars for the measurement of the speed of light? Sounds Ig Nobel-worthy. :D :D
3:04 my old chemistry teacher would be furious if they saw someone licking their fingers in a science lab
We should probably call it an "uncertainty bar" as opposed to an "error bar". That seems a whole lot more grokable.
He is one of the best guests on this channel. I would like to see him more often.
you can get them in the us at places like world market and some grocery stores in the euro/international sections, though they are not as widely sold here
Thats not so bad, at least everyone was dressed. My buddy in MOS school absentmindedly put a bowl of ramen in the microwave without water. Got a whole Marine barracks evacuated at 2 AM. Just shy of 500 Marines in their underwear in a parking lot in the middle of freezing desert winter with two hours before the start of a standard 12-14 hour work day. I didn't think he was going to make it through the night.
7:20 "It's actually 3, and you got 3.4 - I think that's pretty good"
Dammit Brady, you're spending to much time with astronomers... :P
We have Aero bars in the U.S. Not everywhere, but they were in the international foods section of the grocery store
"Well how do you know anything?"
"Exactly."
I love this guy.
I missed you professor
We do have a product like that in the US, it's called Hershey's Air Delight.
We do get them in the US, usually in the "foreign foods aisle".
12:04 Thumps up! Thumps up! Thumps up!
I HATE when my students say: "The real value is x, so we are pretty close to being right." Your result is ALWAYS right per definition! It might be uncertain, or you might have measured something different that you tried to measure. But it is never right or wrong.
I used to hate error bars, then I learned about adaptive optics on telescopes and now love error bars!
QUESTION FOR MORAIRTY; How does one manufacture/machine a pinhead of width 1 atom for a Scanning Tunneling Microscope, and then on top of that, have a machine that is capable of making physical adjustments on the scale of individual atomic radii?
Maybe not a very exciting topic, but important nonetheless. But Brady, please use × (Alt+0215) in stead of x as a multiplication sign and ± (Alt+0177) for the plusminus sign. Keep it up!
You CAN actually say that exact value for c is 299,792,458 m/s because the unit of length, the metre, is defined from this constant.
I like error calculations, they're not hard at all and make a lot of sense. It's fun when you do measurements and get errors that are over 100%.
mpandurov okay, then this vid is reformulated to measure the frequency of the microwave. Same idea, but I think less accessible, they made the right call.
I think it is... Because you can go full ham with error calcs... Like he said, there's an error even in the ruler, which would not only matter when finding the distance between the center of mass between valleys in the chocolate, but even when having to find the center of mass themselves - which you probably can't do without even more errors. And then we have the setup of the microwave itself - how do you know there's not some bouncy effects of waves going on in there :( and so on.
This was a great way to revisit error bars and enjoy some chocolate :) loved the combo. Nice work from the professor and Brady!
Moriarty is easily one of my Top 3 favorites and I'm happy to see him back! However, I originally came to ask if they updated their Sixty Symbols thumbnail photo, or did I just mis-remember what it looked like?
I did this in my electromagnetism class. It's very tricky when you start involving multiple measurement instruments. And you need to dig through a lot of PDF documents to find the correct error percentiles for the range of measurements you're doing.
It's fun, if you can understand it.
that genius! i never thought that those are standing waves. i thought everything gets kinda heated equally! wow thats actually something i was kinda interested in. nice one!
wtf else would they be if not standing waves?? theyre in a confined space ffs.
i thought the amplitudes move from one side to the other, idk im no expert?? i thought thats possible even in a confined space? im no physicists
can you show us in the next vid how to calculate statistical significance in that exact example?
Statistical siginificance only makes sense if you are comparing your observations to a null model/hypothesis.
So e.g. you could ask whether the speed of light is the same in this microwave oven as everywhere else. :D
(That the true speed of light was in the range of what was predicted by the observations is basically a crude indicator that it doesn't differ significantly from our expected result.)
first of all, i dont really get how is it calculated and im too lazy to make any effort searching, also i think explaining it for dummies would make a great video
Phil is back :)
yeah, unfortunately
it looks like it's filmed long time ago
I know how much Phil loves the nitpicking criticisms on these videos so I'm not going to mention that the way I remember Hank Green teaching me sig figs on scishow, you don't do any rounding, because of course you're not meant to know what that next figure is.
Anyway, you guys are awesome, and I could listen to Dr Phil all day.
- How do we know anything? - Exactly!
Perfect place to end the video! :)
This was very interesting. Repeating the experiment was discussed, it would be interesting to discuss if increasing the number of experiments or increasing the precession of the measurement to improve the accuracy of the result. Is the Δ%=(measured - expected) /expected not considered a valid measurement of accuracy?
The +/- 2 value quoted in this video is actually the standard deviation - a direct reflection of the scatter of the measurement data. Error bars are normally determined from the actual inherent uncertainty in the measurements. That is, each quantity or variable in the experiment has an inherent uncertainty associated with it which can depend on many things - including HOW the measurement was taken. For example if measure the diameter of a metal sphere using a micrometer or veneer caliper you will get a different uncertainty than if you measure it using a wooden ruler. The error range used in this video is purely a statistical measure of the standard deviation of the data set - and in this case there were only 3 data points used.
It's crude, but I suppose the purpose of the demonstration was to highlight a practical way of measuring the speed of light using a household appliance and basic Physics equation (c=l*f).
more videos like this, please
I remember having to do this in undergrad lab. But it was a lot more convoluted. We had to take enough measurements that we could justify quoting a probability distribution to our final result. It could easily have happened that all three chocolate bar measurements were closer together. In this video your error bars would have been smaller. What we had to do in lab, is quote a probability that the true value was between two values. For example C = 3.4 x10^9 m/s with a standard deviation of 500 x 10^9 m/s. Meaning that there is a small probability that the true speed of light is X.XX m/s even though we never measured anything near that value. (by the way, we didn't measure the speed of light in lab, it was all thermal and acoustic measurements).
I would like to see a video looking at the mathematics/engineering involving ball bearings and their use. Stuff like calculating the strength of a bearing and other things like energy loss to to friction and whatnot.
I'm going to buy a pallet of aero bars, and expense it to my work. When I file the expense report, I'll say "it was for science".
You know you're dealing with uncertainty when the experimenter mutters "bollocks!"
Great communicator. BTW did you guys et the chocolate?
except for the black part, I would love to after cooling it down
Didn't understand it. What were those little green dots in the melted chocolate bar?
One type of Aero bar has a minty green filling inside.
Sounds delightful.
Pure cancer.
*Measure the speed of light.... with chocolate*
has science gone too far?........
delicious delicious science has gone too far x3
If we're doing a full analysis, instead of error bars, we think of the underlying probability distributions and in this case, the distribution of a sum a variables is the convolution of the individual distributions. However, what about the product? Do you just take the log, convolve, and then pull back with an antilog (i.e. exponent)?
"We have 8cm, 6cm, 7cm. Whats the best thing to do here, lets take an average..."
Looks at the camera and realizes he needs to do math.
"Lets use 7cm"
5:57
The 2 cm are not the error of the mean wavelength but the error of every single measurement!
Explanation:
The way the error bars of delta_lambda=2 cm were obtained is misleading: it suggests that when you perform more measurements the error cannot become smaller but might even become larger as you will get bigger deviations from the mean with a certain probability. Just wanted to point this out, I see this is a UA-cam video and Phil said himself this was a "crude and basic introduction". And I still think the point of the video was made very clear. But maybe a future Sixty Symbols video could cover how estimated errors in your measurements translate to errors in a derived quantity, e.g. if c=f*lambda and you assume f=f+-delta f and you measure lambda several times each with an error of say 2 cm, how do you obtain the mean, the standard deviation and the error the of the mean of lambda and how do you get the error of c in the end.
Question: What's the purpose of the "relative error" thing? Can't I just convert the units of the error and plug it in the formula for c, replacing c with delta-c and lambda with delta-lambda? It also gives the value 0.49*10^8, which is unsurprising given that it cancels out to be what I'm talking about, delta-c=f*delta-lambda.
T=2*pi*sqrt(L/g)
Also, what if we were trying to estimate a pendulum's period by measuring L, given pi and g (for simplicity's sake)?
deltaT=2*pi*sqrt(deltaL/g) or deltaT=T*deltaL/L?
>2.4 GHz microwave
>/me looks at router
>2.4 GHz
i freaking love this video. great explanation of fundamentals. so many people get tripped up on the notion of certainty in science. hell yeah. top stuff. would watch again.
Oh man Phil's videos are the absolute best !
We do have Aero bars in the US
I watched the „age of the universe“ video before, and it featured an asymmetrical errorbar ( for the value of the Hubble-constant as measured by „standard-siren“ of 70 +12/-8 km/s/Mparsec). How do you end up with asymmetrical errorbars? Why is one side weighed more? Probably in the details of the measurement, but I‘d like an explanation...
My man causes a Campus wide panic with his experiment and decides to change a single variable and try it again :D
At 6:17 above the formula c = f(14cm) it says STD on the plane ticket. Std is the abbreviation for standard deviation: how you describe error bars ;)
Am 1st year. Can confirm, error propagation is the bane of my existence.
Importance of error bars can not be overstated.
“Well, how do you know anything?”
“Exactly” 😂
The one thing that blew my mind was the size of "microwaves". If I can measure those with my hand, then they are normal human sized, not micro. We've gotta rename them.
The Sapien named as such in radio terms. Long wave, medium wave, short wave.....
milliwaves!
remember that "long wave" in radio is over 1km, 3kHz signal has a wave almost 100km long
"How do we know?" I'm getting ToK flashbacks.
we get aero in canada they're great. Especially the mint ones.
Anything on Einstein-Rosen bridges?
hey u did not answer my question on the video about why is glass transparent.
It's possible to get Aero bars here in the 'States, but you'd have to hunt for them a bit. They're not particularly common.
Moriarty Day is a good day. :) One issue with teaching the concept of error bars might be simple terminology. The "error" in error bars make it seems like a bad mistake. Maybe "uncertainty bars" is less confusing?
Video on using first principles method in learning science
Can I get the error bar of the speed of light by putting 2 cm as lambda into the formula?
you should do a video on Propagation of uncertainty
Isn't frequency v?
v is velocity I believe, they actually used v=fλ to calculate the speed of light.
I think you forgot to allow for the error in the rule in your description of the error calculation though I presume it is small compared to the wavelength. When I worked in a sheet metal shop you had to be careful with the measurements as it was quite common to accidentally chop the end off rules in the guillotine so often the first few centimeters or inches were missing !
The other thing is that the Frequency written on the back may have used the speed of light in it's measurement if so you can't really use the labeled frequency to determine the speed of light.
The IPCC needs to watch this video.
Thank you for the video, I always wanted to know this. 👍
Very interesting!
Are microwave ovens dimensioned to an integer multiple of the wavelength to encourage standing waves? Why? An improbable accident? If uneven intensity distribution is undesirable it would seem trivially easy to avoid by having non-parallel sides and irregularities in the pressing of the steel. The fact that standing waves are present is taken for granted here and I've not heard an explanation.
If you don't choose the width so you get standing waves, the waves would interfere destructively everywhere and mostly cancel out, or you get even weirder uneven patterns. Getting a lot of power into the food is already hard (you need big transformers to get high voltages), so you just deal with the uneven heating by rotating the plate and letting the heat distribute in between heating.
Also, the equal distance between the melted spots is a giveaway (also presumably between cold spots & the walls), so assuming standing waves shouldn't be strictly necessary.
They are actually "dimensioned to an integer multiple", and this is why your microwave tray is programmed to rotate. The standing wave would be heating only certain sections of your food without the rotation!
Will To Win thanks, and to nibblr for the responses, but does this lead to the conclusion that if the microwaves were scattered they would not heat the food? I bet you can both guess the next question: where would the energy go?
my conclusion would have been: the speed of light in chocolate filled with bubbles - c(aerobar) - is (dare I write 'significantly') greater than in vacuum. what implication will this finding have to the future of space travel? oh my!
fun aside . great explanation to a problem virtually everybody struggles with at the beginning!
Non-physics students probably won't realize how genius this is...
We do get the Aero bar in the US, but not widely.
I would read the frequency as +or- least significant digit i.e 1Mhz
That zero at the end is suspicious too, though. :)
Wait, if you get your error bars from [delta]C/C, then aren't the value of the error bars highly dependent on your measuring scale? If I'm measuring a temperature in Fahrenheit or Celsius, I would get radically different error bars, since the ratio and zero point are different. Doesn't this render the whole exercise vapid and arbitrary?
I have seen Aero Bars in the US. Not common, though.
Wow wtf is going on in the comments, I thought this was a really interesting science video.
Some people are entrenched in their ideology to such a fundamentalist level that they cannot separate what somebody says about science from what they say about identity politics.
lots of irony in that statement
phil has been a naughty boy outside the context of this channel, and some people won't ever let him live it down.
imagine watching a science video where hitler explains quantum superposition, and nothing else. people would be losing their minds about it even though it was simply a video about quantum physics.
Wow, someone actually Godwinned a 60 symbols video.
The fact that you went with Hitler as part of your comparison is rather pathetic. Disagreeing with Dr Moriarty is fine. Being mad is fine. But comparing the guy to Hitler? Wow.
Gunhaver. So feminists are Hitler, but you're rational for not wanting to see this guy explain physics? Got it.
Instead of just putting the chocolate bar in as a big block you would probably get more precise results if you separated the bar into its individual small squares of chocolate. (It looked like the bar is a 7x5 grid in which case 35 individual squares.) Put each square on the plate in the same rectangular shape but separate them just slightly so they're not touching each other as some of them melt. That way the heat from one square won't be able to easily transfer to a neighboring square and the main source of energy to melt a square will come from the microwave itself, not from convection within the chocolate.
Cool experiment though. :)
I could smell burning chocolate as I watched this. Is that a new feature of UA-cam?
I was at UoN doing a Master's during 2017, and a fire alarm did go off during a modern cosmology lecture ... Now I know why!
this guy is amazing
Wouldn't you get a better melt pattern in the chocolate if you placed it at an elevated level inside the microwave?
you can get aero bars in Canada
Now do it again, with more variables that we get to do multivariable calculus on to determine the statistical error bars!
Mint Aero are the BEST!
I love Moriarty's videos, they always spring up 2 topics.
This makes me wonder if a microwave constructed with non-parallel sides would eliminate standing waves, especially since they don't have that rotating metal reflector in the ceiling any more.
nice and informative video :)
I actually fucking love Phil
He interviewed me in July for my Masters in Theoretical Astrophysics, I was pretty chuffed :P
We don' care if you like him or dislike him. Only science comments allowed.
Tha Pume at this point I can't even tell if this is ironic
Wow, we got the owner of UA-cam right here.
Why 2 cm is the error in lambda?
Because we have the range between the maximum and minimum value of our experiments within 2 centimiters from 14
Hershey's did (not sure if they still do) something like this with their chocolate bars.
They called them "air delight" over here in America.
Clever way to save the amount of chocolate per bar, but sell it for the same price, and trick the human brain into it tasting better. ;)
Im suprised Aero doesnt have a distribution center here in the US. They would sell like hotcakes.
Measuring speed of light with chocolate and microwave oven..amazing!!!
that's why Physics is beautiful.!!
“Well, how do we know anything?”
“Exactly.”
Yep what a great idea, sell air wrapped in chocolate for more than stuff without air
My high school took half a semester teaching us about units, accuracy and precision in Year 1 Physics
Cyan Sea wow very interesting