Fair of you to throw in the disclaimer but that was really well explained. "If you can't explain it simply, you don't understand it well enough" - Einstein.
I agree but has it been dumbed down so much that it actually doesn't tell the truth anymore? This is my biggest worry with easy explanations for sophisticated phenomena.
@@garmarianne3533 it certainly requires an understanding of statistics. I don't know about you, but the first time I encountered statistics was during university.
@@garmarianne3533 But I don't understand p if I don't understand statistics. I have to accept p and whatever is told me about p, but I don't know why p is the value which tells me that a finding is most likely not due to randomness. It's the difference between understanding a thing and acousmatics.
The p-value is used to determine if the outcome of an experiment is statistically significant. A low p-value means that, assuming the null hypothesis is true, there is a very low likelihood that this outcome was a result of luck. A high p-value means that, assuming the null hypothesis is true, this outcome was very likely.
Love this :) Needed a quick refresher on p-values and got it without dying of boredom or wading through overly complicated definitions. Funny, simple, educational. Thanks!
Thank you for that. p-value is one of those things that make no sense until you understand them, but once you do, you wonder what the problem was. Glad I could help.
The alpha value can be decided ahead of time, and is really a remnant of the days when p-values were hard to calculate. You would decide on an alpha value and then find out the cutoff for the statistic you were measuring. These days we have p-values so can see from them how much evidence there is that the null hypothesis is false. An alpha value of 0.05 or 5% is the most usual, but it is arbitrary. The alpha value is also interpreted as the probability of making a type I error, which is rejecting the null hypothesis when you shouldn't.
This could not be more incorrect (and is now repeated in a number of terrible stats textbooks). In order to calculate the p-value you're deriving it from an assumption that the null hypothesis is true. But the p-value is uniformly distributed when the null hypothesis is true which means all of them, .05, .5, .0001, are equally likely to occur and are not, in and of themselves, evidence of the effect. Without stating beforehand some sort of cutoff you're just like the person who stumbles across something odd and calls it a surprising finding. I liked Feynman's little story that starts with him talking about this amazing license plate number he stumbled upon in the parking lot when walking into class. The video already has issues arguing that p-values are measures of evidence, an idea that even Fisher gave up on later in his career (thus likelihood ratios).
Thank Helen I am beyond the eleventh hour trying to understand Null Hypotheses and p values for college all the while battling Breast Cancer. Your video has made my life soooo less stressful. Hopefully I can submit my stuff as the research project is past due, and get to bed at a decent hour...keep em coming. I have another Data Analysis course this summer I will be looking out for more of your videos. Thank you for making it so simplistic...not every one is a math genuis:) You Rock
Thank you for making statistics easy to follow in an entertaining way! You definitely help students like me who have difficulty following the concepts...your method of teaching works perfect for me b/c it helps me get it! I appreciate how your videos simplify the explanations so that its easy to follow. Thank you!
I need to say thank you Dr.Nic Because currently I am undergraduate student from science school. My dream is to being a scientist ,however my statistics. is so bad,but.after watch the video ,I find my passion to pickup and overcome statistics !!!! So,Thank you so much for making this video.Hope we can learn more from you in the future day!!
That is great to hear that I can help you reach your dream. I have videos to help people know how to study for statistics as well. You can find them all laid out nicely here: creativemaths.net/videos/
After watching this video, I'm still clueless about p values and the "null hypothesis". The null hypothesis needs to be better explained and WHY a low p value implies rejection of the null hypothesis.
Null hypothesis means that you are testing your variables at 0.05 (%95 level) significance level your variables if they are correct to estimate your model or sample then youe variables is not equal to zero,they are affecting your sample which null hypothesis was Ho:B1=B2=0 this implies if you have right variables to solve your model then you reject null hypothesis because b1 and b2 has an effect on your model.p value determines your null hypothesis is really true or wrong.So if mean average of your variables more than 0.05 in p value than you have to accept that your variables are wrong and has no effect or your null is true.
Someone correct me if I'm wrong. The null hypothesis is the assumption that every variation to expected results is due to chance. In this video, the assumption was that , on average, choconutties contained 70g of nuts. The difference between her sample mean and the expected mean is put through some complicated equations to get a t value. You use that t value to find the p value from a pre-made chart. The p value in this example was 0.18. That means that there is only an 18% probability that the difference between her sample nut-mean and the 70g nut-mean is due to chance. That means that there is a 72% chance that these results are NOT due to chance and she really doesn't have enough nuts in her chocolate bars. However, it is convention in statistics that we have to reach a 95% confidence level before we can reject the null hypothesis (our assumption that whatever observed events that differ from the norm are due to chance.) For example, if a scientist observes a series of really hot days in the middle of winter, he can take take a large sample of the temperature of these few days in the year across 50-75 years in the past. He then finds the average (mean) of the temperature. This is the expected. He subtracts the actual temperature from the expected, then finds the t and then p value. If the p value is 0.1, then there is a 10% chance that these few hot days are just happenstance. However, if the p value is 0.01, then there is only a 1% probability that these hot days are hotter than average due to chance. There is likely something causing these days to be hotter. By rejecting the null hypothesis, we're saying that these results are NOT due to random variation in the universe.
Hi. See my response to Youbillyblake above explaining why we go fast. (Just stop and go back if you miss something.) I have done captions for some of my videos for people who have trouble with the way I speak!
Watching this has made a huge difference to my understanding of p-values, I was struggling to grasp the concept but your content has explained it perfectly. Thank you!
In a doctorate program and in a stats research class...it's mandatory...the funny part...I have never had stats before. In summation...THANK YOU FOR MAKING THESE VIDEOS!!!! :-)
You are most welcome. You might find my blog post about stats courses resonates: learnandteachstatistics.wordpress.com/2017/01/11/why-people-hate-statistics/
Hello! Could you please explain how the P-value was calculated to be 0.18? I tried using a one sample t-test on the same values in SPSS with a test value of 70, and there the P-value is showing as 0.357. What am I doing wrong :/ ?
What an incredible video. Of course very simplified but I think this is a great introduction to the concept. I feel that people have a really stilted understanding of what is 'likely', however. If anyone is reading this for that reason (or whatever), say you have your .18 P-value on these choco bars. 18% may not seem like a high chance of error in your analysis, but when you consider that across however millions of such bars may be produced in the future, 18% is MASSIVE. In physics in general, hell, a 5% chance may as well be a certainty. But the amount of permissiveness depends on the test, of course.
Your videos are really saving my butt, they're so clear and concise and also I love the art, the facial expressions always get me haha, it really helps me to stay engaged!
The p-value comes from a table of probabilities for the (in this case) t distribution. In most cases it is given as the output from a computer statistical program, but you can calculate them for yourself. This video: ruclips.net/video/Y3mGoW5w28c/видео.html will give you a good idea of where the p-value comes from.
This is excellent! I have been trying to get this stuff down all year at uni and it's been going in one ear and out the other. Your video really helped! Thanks :)
I love how innovative your videos are! Please make more of them, I can't wait to finish some more playlists :D These are amazing and thank you so much for all your explanation!
Good point - I think it is just that Helen is really bossy and likes to push him around. Also he isn't that good at it. But I will see if I can get her to do her own analysis next time.
Hi. I'm not sure what you are asking. It depends on which way you set up the null and alternative hypotheses. It often seems backwards, as we are trying to disprove something in order rather than prove it.
thank you so much its been like 3 sessions of stadistics and i havent been able to understand this, this 4:42 video should get paid instead of my teacher
It's a very nice video! You could even say it like this "There is an 82% chance that the mean weight (in this sample size) is more than than 68.7 g using the null hypothesis."
Almost. You are best to stick to the usual interpretations. You could say that the probability of getting a sample mean of this size or greater if the population mean is 70g is 0.82. You said "the mean weight" of the sample, whereas we can only talk in probability terms about a theoretical sample mean.
Hi I understand your problem. What you need to do is stop the video when it gets too fast for you and think about it, then go back and listen to any bit you need to think about again. The videos are made fast on purpose. It is easy to slow it down and control the speed yourself if they are fast. If we make them slow then it is very difficult for the viewer to speed them up, and then they get boring. You control the pace!
Wow. TY. I only need to know this 1/8th of test and I WILL NEVER need again. TY for explaining simply. It made the more complex explanation understandable,
Brilliant example of how statistics is generally taught, and why most research using stats is wrong. It's ok to simplify the explanation, providing it doesn't lead to terrible consequences. I consider almost all medical research being demonstrably incorrect as a terrible consequence.
Wonderful explanation. So in brief: A null hypothesis is a statistical hypothesis that assumes no difference or relationship exists between two variables or groups. It is commonly denoted as H0 and is used as a starting point for statistical testing. The null hypothesis is usually tested against an alternative hypothesis, which is a statement claiming that there is a difference or relationship between the variables or groups under investigation. H1 is commonly used to represent the alternative hypothesis. The purpose of statistical testing is to see if the data provide enough evidence to reject the null hypothesis in favor of the alternative hypothesis. This is accomplished by computing a p-value, which is the probability of obtaining a test statistic at least as extreme as the one observed if the null hypothesis is true. If the p-value is small (typically less than 0.05), the data are unlikely to have occurred by chance if the null hypothesis is true, and the null hypothesis is rejected in favor of the alternative hypothesis. If the p-value is not small, it means that the data are not sufficiently different from what would be expected if the null hypothesis were true, and the null hypothesis is not rejected. In summary, the null hypothesis is a statistical hypothesis that is tested against an alternative hypothesis, and the p-value represents the probability of receiving the observed data or something more extreme if the null hypothesis is true.
One thing that is not clear from this video: It says P-value = 0.18 means that the the probability of getting a mean of 68.7 or LESS from a sample of this size is 18%. But it also says that the smaller the p-value is, then we have more evidence that the null hypothesis is wrong. P-value being small (such as 0.05) indicates that there is 5% of getting a mean of 68.7 or less from the sample, which means there are 95% chance that the bar contains more than 68.7g of nuts. How could this reject the null hypothesis (= the mean is 70g)?
It does take a while to get your head around the p-value. I have corrected the first sentence in your second paragraph: P-value being small (such as 0.05) indicates that there is 5% of getting a mean of 68.7 or less from the sample, which means there is 95% chance of getting a sample mean of 68.7 or more if the population mean is 70g of nuts. The probability expressed in the p-value is about how likely you are to get that result, not what is happening in the population. Ask if you need more explanation.
THANK YOU, I obviously needed this really dumbed down, but as it's to help me proofread tables in my editing course (ie I'm better with words than numbers) I needed to know the simplest explanation.
I am gratified to hear that you are learning about this in an editting course. I wish more journalists, writers etc had a good understanding of statistics.
At my job now I'm poking through a lot of data with significant p values, and since I haven't taken a statistics class in over 10 years... this video was a great refresher, thank you!!! I'm still curious as to how one actually calculates the p value though.
Totally! It doesn't help that the word "significant" is often taken to mean important or notable - which it is in real life. In statistics all it means is that there is evidence that the effect shown in the sample exists in the population. I am currently in the throes of my next video, which addresses these ideas.
Sorry, that is not the correct interpretation. Your original idea is either correct or it is not. There is no probability involved. The p value is how likely you are to get this result if your original idea is correct.
You might notice that the p-value you got is exactly double mine. You have got the two-tailed value, and I've given the one-tailed value. For an explanation of tails, see our video about hypothesis testing, difference of two means.
level of significance... The null hypothesis is rejected if the p-value is less than a predetermined level, α. α is called the significance level, and is the probability of rejecting the null hypothesis given that it is true (a type I error). It is usually set at or below 5%.
The best thing I've ever heard "if P is low, Null must go"
This is golden!
I agree. I didn’t make it up and tried to credit whoever did, but couldn’t find out. So helpful!
Fair of you to throw in the disclaimer but that was really well explained. "If you can't explain it simply, you don't understand it well enough" - Einstein.
The simplest explanation I have found in the internet. Avoiding the jargon made this very easy to understand.
Two thumbs up!!
I agree but has it been dumbed down so much that it actually doesn't tell the truth anymore? This is my biggest worry with easy explanations for sophisticated phenomena.
@@KommentarSpaltenKrieger how come p-value is a sophisticated phenomena?
@@garmarianne3533 it certainly requires an understanding of statistics. I don't know about you, but the first time I encountered statistics was during university.
@@KommentarSpaltenKrieger It leads to understanding of satistics, it doesn't require it.
@@garmarianne3533 But I don't understand p if I don't understand statistics. I have to accept p and whatever is told me about p, but I don't know why p is the value which tells me that a finding is most likely not due to randomness. It's the difference between understanding a thing and acousmatics.
The p-value is used to determine if the outcome of an experiment is statistically significant. A low p-value means that, assuming the null hypothesis is true, there is a very low likelihood that this outcome was a result of luck. A high p-value means that, assuming the null hypothesis is true, this outcome was very likely.
That's a nice way to break it down, thanks Eliot!
but how many are low, is it should below 0.5?
A low p-value means that, assuming the null hypothesis is true (or you mean false) ?
what is null hypothesis
Your explanation is more easier than this video ❤️❤️
This video absolutely saved me on my exam today. Thank you. People like you are amazing
Love this :) Needed a quick refresher on p-values and got it without dying of boredom or wading through overly complicated definitions. Funny, simple, educational. Thanks!
0:40 Helen?? No. That's Mello from Death Note. That's not Helen.
Adryanne Parker Maybe... (Blame my editor)
+Adryanne Parker LOL xD
lol
YEAH MELLO ALL THE WAY!!!
Adryanne Parker I nearly screamed when my teacher showed this to our class, haha!
The perfect video! I don't understand why college makes it sound like a concept of god; it's simple: if p is low, the null must go!
Thank you!
Thank you for that. p-value is one of those things that make no sense until you understand them, but once you do, you wonder what the problem was. Glad I could help.
The alpha value can be decided ahead of time, and is really a remnant of the days when p-values were hard to calculate. You would decide on an alpha value and then find out the cutoff for the statistic you were measuring. These days we have p-values so can see from them how much evidence there is that the null hypothesis is false. An alpha value of 0.05 or 5% is the most usual, but it is arbitrary. The alpha value is also interpreted as the probability of making a type I error, which is rejecting the null hypothesis when you shouldn't.
This could not be more incorrect (and is now repeated in a number of terrible stats textbooks). In order to calculate the p-value you're deriving it from an assumption that the null hypothesis is true. But the p-value is uniformly distributed when the null hypothesis is true which means all of them, .05, .5, .0001, are equally likely to occur and are not, in and of themselves, evidence of the effect. Without stating beforehand some sort of cutoff you're just like the person who stumbles across something odd and calls it a surprising finding. I liked Feynman's little story that starts with him talking about this amazing license plate number he stumbled upon in the parking lot when walking into class.
The video already has issues arguing that p-values are measures of evidence, an idea that even Fisher gave up on later in his career (thus likelihood ratios).
plus you should be selecting your alpha level a-priori which means before you do anything
We are so glad you like it. Statistics is such an important subject that it needs to be fun as well.
If you're here to quickly and easily understand the concept of P-values, then you've found the right video.
Good on you, Dr. Nic.
That is good to hear. I think it works best if you already have some idea of what a p-value is, but it isn't clear.
Thank Helen I am beyond the eleventh hour trying to understand Null Hypotheses and p values for college all the while battling Breast Cancer. Your video has made my life soooo less stressful. Hopefully I can submit my stuff as the research project is past due, and get to bed at a decent hour...keep em coming. I have another Data Analysis course this summer I will be looking out for more of your videos. Thank you for making it so simplistic...not every one is a math genuis:) You Rock
I hope you are ok, health wise.
get well soon :)
How are you doing now Jennifer?
Zaχ I’m doing well. Thank you for asking
Pray for your recovery. God is good! Math is amazing.. once you understand it. Wishing you all the best for your studying and health journey x
"P is low, Null must go" love it!
Thank you for your encouragement. I hope you don't mind, but I quoted you in my blog. Comments like these really make my day.
Honestly those videos are the only reason i still have hope on this subject. The comedy in the presentation is just the cherry on the top.
That is so good to hear. I love that my videos keep helping people.
Mello survived, changed his name to Helen and owns a chocolate business, lmao. Matt pretends to be his brother too.
I can neither confirm nor deny the presence of Mello in these videos. ;)
I mean regardless the association is there. **shrugs**
@@0nullnil lol imagine in the new one shot from this year with 27 year old Near it showed 29 yr old Mello and Matt owning a chocolate company
@@yeetusdeletus1229
I’m imagining it and honestly I feel like that could happen..
Omg I'm not the only one who thought it looked like Mello
I'm happy to hear that. Life-saving is a good thing.
"In general, ..." That was my favorite part LOL.
Thank you for making statistics easy to follow in an entertaining way! You definitely help students like me who have difficulty following the concepts...your method of teaching works perfect for me b/c it helps me get it! I appreciate how your videos simplify the explanations so that its easy to follow. Thank you!
I need to say thank you Dr.Nic Because currently I am undergraduate student from science school. My dream is to being a scientist ,however my statistics. is so bad,but.after watch the video ,I find my passion to pickup and overcome statistics !!!! So,Thank you so much for making this video.Hope we can learn more from you in the future day!!
That is great to hear that I can help you reach your dream. I have videos to help people know how to study for statistics as well. You can find them all laid out nicely here: creativemaths.net/videos/
I liked purely because I saw Mellow. I also clicked the video because I saw Mellow. You can denie it all you want but that is Mellow from death note.
After watching this video, I'm still clueless about p values and the "null hypothesis". The null hypothesis needs to be better explained and WHY a low p value implies rejection of the null hypothesis.
Paul Sandberg totally agree
Null hypothesis means that you are testing your variables at 0.05 (%95 level) significance level your variables if they are correct to estimate your model or sample then youe variables is not equal to zero,they are affecting your sample which null hypothesis was Ho:B1=B2=0 this implies if you have right variables to solve your model then you reject null hypothesis because b1 and b2 has an effect on your model.p value determines your null hypothesis is really true or wrong.So if mean average of your variables more than 0.05 in p value than you have to accept that your variables are wrong and has no effect or your null is true.
Were you able to find an explanation to this confusion?
Someone correct me if I'm wrong.
The null hypothesis is the assumption that every variation to expected results is due to chance. In this video, the assumption was that , on average, choconutties contained 70g of nuts.
The difference between her sample mean and the expected mean is put through some complicated equations to get a t value. You use that t value to find the p value from a pre-made chart. The p value in this example was 0.18. That means that there is only an 18% probability that the difference between her sample nut-mean and the 70g nut-mean is due to chance. That means that there is a 72% chance that these results are NOT due to chance and she really doesn't have enough nuts in her chocolate bars.
However, it is convention in statistics that we have to reach a 95% confidence level before we can reject the null hypothesis (our assumption that whatever observed events that differ from the norm are due to chance.)
For example, if a scientist observes a series of really hot days in the middle of winter, he can take take a large sample of the temperature of these few days in the year across 50-75 years in the past. He then finds the average (mean) of the temperature. This is the expected. He subtracts the actual temperature from the expected, then finds the t and then p value. If the p value is 0.1, then there is a 10% chance that these few hot days are just happenstance. However, if the p value is 0.01, then there is only a 1% probability that these hot days are hotter than average due to chance. There is likely something causing these days to be hotter. By rejecting the null hypothesis, we're saying that these results are NOT due to random variation in the universe.
Completely agree with you
This is by far the best tutorial out there in terms of p-value. Thank you so much you just saved my life.
Glad it helped!
Hi. See my response to Youbillyblake above explaining why we go fast. (Just stop and go back if you miss something.) I have done captions for some of my videos for people who have trouble with the way I speak!
Basically if the the significant level is 0.05, p0.05 means null hypothesis is true, the chocolate have 70g or more peanuts.
The animations crack me up. Fun video. Thanks.
This video moved my niddle of understanding P-value a notch closer to where I it to be.
That IS good news
my goodness... you explained it too damn clearly. I have been looking all over youtube.. for a good explanation.
p is low, null must go. thank you!
the video was dumb enough for me to understand. thanks
I think u misunderstood my comment fam. iim saying the video is made easy for dummies like you and me, which is a good thing.
haha all gud
I understand. I really tried to make a very difficult concept easy to understand, so I am happy you were able to get it. You are not dumb.
Hahaha
I agree. this video is much more helpful than my statistic book from my master course (... or maybe I'm just dumb lol). glad to stumble upon this!
One of the best channels I have bumped into so far. Way better than Khan Academy, I must say!
Thanks. Helen has taken on a life of her own over the years.
... Mello math?
Looks like Mello from Deathnote XD
Lol thats why I clicked. Mello from deaht note
Watching this has made a huge difference to my understanding of p-values, I was struggling to grasp the concept but your content has explained it perfectly. Thank you!
Glad it helped!
In a doctorate program and in a stats research class...it's mandatory...the funny part...I have never had stats before. In summation...THANK YOU FOR MAKING THESE VIDEOS!!!! :-)
You are most welcome. You might find my blog post about stats courses resonates: learnandteachstatistics.wordpress.com/2017/01/11/why-people-hate-statistics/
Who else came here from the Death Note subreddit because it has Mello in it
Same
Tee hee
@@DrNic 👀
Hello! Could you please explain how the P-value was calculated to be 0.18? I tried using a one sample t-test on the same values in SPSS with a test value of 70, and there the P-value is showing as 0.357. What am I doing wrong :/ ?
Hi. I obtained the p-value by assuming that I sampled from a normal distribution. So P(Z
Dr. Nic - thank you! We truly need more teachers like you.
You're very welcome!
The most simple and easy to understand explanation I have found. Thanks.
Great video - perfect for getting the concept in my head. Great jokes too :D
Glad you enjoyed it. Our editor is pretty funny.
That was an excellent explanation - thanks!
You just saved my life ,helping me wrap around this concept incidentally tend to be mathematically challenged
That is great to hear - I like to save lives!
This is by far the best video concerning p-value. Amazing job
Wow, thanks!
This is great! Thank you.
Extremely helpful n understandable.
What an incredible video. Of course very simplified but I think this is a great introduction to the concept. I feel that people have a really stilted understanding of what is 'likely', however.
If anyone is reading this for that reason (or whatever), say you have your .18 P-value on these choco bars. 18% may not seem like a high chance of error in your analysis, but when you consider that across however millions of such bars may be produced in the future, 18% is MASSIVE. In physics in general, hell, a 5% chance may as well be a certainty. But the amount of permissiveness depends on the test, of course.
Happy it helped
Omg after a dozens of video and lectures I finally get it form this vid... this is the perfect example guys...
Happy to hear it
OMG these videos are hilarious, i loved it, and btw im learning a lot! Thank you!
How do we calculate the p-value?
you still need the answer after 4 years or you found it?
Graphing calculator
@@ufatemav4509 lol
U Fatema V hey it’s been a year, are you going to answer it?
This is by far the best video on explaining p-value. Thank you.
Glad it was helpful!
Your videos are really saving my butt, they're so clear and concise and also I love the art, the facial expressions always get me haha, it really helps me to stay engaged!
Thanks Bethany - that is great to hear. Mission accomplished!
But what exactly is p-value? Where does it come from?
The p-value comes from a table of probabilities for the (in this case) t distribution. In most cases it is given as the output from a computer statistical program, but you can calculate them for yourself. This video: ruclips.net/video/Y3mGoW5w28c/видео.html will give you a good idea of where the p-value comes from.
why does Helen's brother look like Edward Elric from Fullmetal alchemist?
XD
Probably just a coincidence. Helen is a little like Mello from Deathnote, some people think. ;)
This is excellent! I have been trying to get this stuff down all year at uni and it's been going in one ear and out the other. Your video really helped! Thanks :)
I love how innovative your videos are! Please make more of them, I can't wait to finish some more playlists :D These are amazing and thank you so much for all your explanation!
Glad you like them!
Mello? Is that you?
Maybe...
Very good clear videos. But I don't like that she's getting her brother to do all the Excel stuff... Ladies can use excel too :')
Good point - I think it is just that Helen is really bossy and likes to push him around. Also he isn't that good at it. But I will see if I can get her to do her own analysis next time.
Purely amazing explanation, simple, effective, clear and with a touch of humour. Great video :)
Glad you enjoyed it! That is exactly what we are aiming for.
Hi. I'm not sure what you are asking. It depends on which way you set up the null and alternative hypotheses. It often seems backwards, as we are trying to disprove something in order rather than prove it.
This Helen looks evil with that grin. For someone who sells “Choconutties”.
Helen does have a bad attitude
To be honest, Helen kinda reminds me of me. Cause I smile like that as well.
watched it again and i can safely say you may have scored me a neat 15% in my test this evening, Helen is awesome! :)
Good to see Mello is doing well.
maybe....
Thanks a lot! I'm currently studying for my CFA exam, I was having trouble understanding the Hypothesis Testing reading, this was very helpful!
Glad it helped!
thank you so much its been like 3 sessions of stadistics and i havent been able to understand this, this 4:42 video should get paid instead of my teacher
Very basic and essential concept of Statistical Sciences explained in a very very easy language, thank you, SLC. .
It's a very nice video! You could even say it like this "There is an 82% chance that the mean weight (in this sample size) is more than than 68.7 g using the null hypothesis."
Almost. You are best to stick to the usual interpretations. You could say that the probability of getting a sample mean of this size or greater if the population mean is 70g is 0.82. You said "the mean weight" of the sample, whereas we can only talk in probability terms about a theoretical sample mean.
@@DrNic thank you
Short and sweet. Thanks for the lucid explanation
Glad it was helpful!
Hi
I understand your problem. What you need to do is stop the video when it gets too fast for you and think about it, then go back and listen to any bit you need to think about again. The videos are made fast on purpose. It is easy to slow it down and control the speed yourself if they are fast. If we make them slow then it is very difficult for the viewer to speed them up, and then they get boring. You control the pace!
I liked the disclaimer at the end.
Yup! There are many levels at which to explain things and it is tricky to be both clear and rigorous
Very good explanation! Understood in under 10 mins with multiple replays.
That is great to hear. It is really sensible to watch it several times, and pause and think.
I like this explanation. Given a simple example and no bullshit
best video ever! i would love all my classes would be as simple as this for better understanding, great job congrats!
This is really helpful!! I finally get the idea of p-value after watching this video. Thanks for making this video :)
Really Great methodology of teaching. I would give more than one likes if i could. Thank you dear Miss!
Wow. TY. I only need to know this 1/8th of test and I WILL NEVER need again. TY for explaining simply. It made the more complex explanation understandable,
Pleased to hear it. Often you have most of the concepts, but you just need them arranged a bit better.
Pleased to hear it. We do our best.
Thanks for helping out with my college studies!
You are most welcome. I hope you find many of my videos useful.
Brilliant example of how statistics is generally taught, and why most research using stats is wrong. It's ok to simplify the explanation, providing it doesn't lead to terrible consequences. I consider almost all medical research being demonstrably incorrect as a terrible consequence.
Indeed.
Thank you so much for explaining this concept in such understandable fashion!
Wonderful explanation.
So in brief:
A null hypothesis is a statistical hypothesis that assumes no difference or relationship exists between two variables or groups. It is commonly denoted as H0 and is used as a starting point for statistical testing.
The null hypothesis is usually tested against an alternative hypothesis, which is a statement claiming that there is a difference or relationship between the variables or groups under investigation. H1 is commonly used to represent the alternative hypothesis.
The purpose of statistical testing is to see if the data provide enough evidence to reject the null hypothesis in favor of the alternative hypothesis. This is accomplished by computing a p-value, which is the probability of obtaining a test statistic at least as extreme as the one observed if the null hypothesis is true.
If the p-value is small (typically less than 0.05), the data are unlikely to have occurred by chance if the null hypothesis is true, and the null hypothesis is rejected in favor of the alternative hypothesis. If the p-value is not small, it means that the data are not sufficiently different from what would be expected if the null hypothesis were true, and the null hypothesis is not rejected.
In summary, the null hypothesis is a statistical hypothesis that is tested against an alternative hypothesis, and the p-value represents the probability of receiving the observed data or something more extreme if the null hypothesis is true.
Very nice summary
One thing that is not clear from this video:
It says P-value = 0.18 means that the the probability of getting a mean of 68.7 or LESS from a sample of this size is 18%.
But it also says that the smaller the p-value is, then we have more evidence that the null hypothesis is wrong.
P-value being small (such as 0.05) indicates that there is 5% of getting a mean of 68.7 or less from the sample, which means there are 95% chance that the bar contains more than 68.7g of nuts. How could this reject the null hypothesis (= the mean is 70g)?
It does take a while to get your head around the p-value.
I have corrected the first sentence in your second paragraph:
P-value being small (such as 0.05) indicates that there is 5% of getting a mean of 68.7 or less from the sample, which means there is 95% chance of getting a sample mean of 68.7 or more if the population mean is 70g of nuts.
The probability expressed in the p-value is about how likely you are to get that result, not what is happening in the population.
Ask if you need more explanation.
THANK YOU, I obviously needed this really dumbed down, but as it's to help me proofread tables in my editing course (ie I'm better with words than numbers) I needed to know the simplest explanation.
I am gratified to hear that you are learning about this in an editting course. I wish more journalists, writers etc had a good understanding of statistics.
Best video ever. Please please keep on making videos for us. Your videos are Khan academy level.
Wonderful explained!! Those wonderful people who take from their time in order to help others..
God bless you!
Thank you Calina. Do tell your friends, as we do get a little bit of money from advertising to help us keep going.
At my job now I'm poking through a lot of data with significant p values, and since I haven't taken a statistics class in over 10 years... this video was a great refresher, thank you!!! I'm still curious as to how one actually calculates the p value though.
Try Understanding where the p-value comes from - Statistics Help ruclips.net/video/0-fEKHSeRR0/видео.html
Totally! It doesn't help that the word "significant" is often taken to mean important or notable - which it is in real life. In statistics all it means is that there is evidence that the effect shown in the sample exists in the population. I am currently in the throes of my next video, which addresses these ideas.
This video explains p-values and significance levels really well. Thanks!
Thank you so much for this great, funny and simple explanation. You just saved me hours of trying to understand this concept :)
I’ve come back to this about a 1000 times
Hopefully it is helping!
Dr Nic's Maths and Stats it is! Every time I have to give a test or understand it again, I come back to it.
thankyou , so easily understood
p value = probability value = probability of our original idea ( = null hypothesis ) being correct
Sorry, that is not the correct interpretation. Your original idea is either correct or it is not. There is no probability involved. The p value is how likely you are to get this result if your original idea is correct.
I hope that is an extra 15%, not 15% total for the test. We are glad we can help.
You might notice that the p-value you got is exactly double mine. You have got the two-tailed value, and I've given the one-tailed value. For an explanation of tails, see our video about hypothesis testing, difference of two means.
Best video on p-value out there !! Thank you :)
Glad you think so!
greatest video ever! finally got the concept of p-value clear
Wow that is high praise
With your videos I am learning Statistics AND New Zealand's English :)
That's great! If you find the accent tricky you can turn on the subtitles.
Thank you. I finally understand what p-value means.
I am watching this 2020 how useful is this channel great work thank you doctor Nic's
Thanks for watching! The older videos still get a heap of views. It is most gratifying.
Thank you :) this helped me lots with completing my final paper. I have a hard time understanding stats
level of significance...
The null hypothesis is rejected if the p-value is less than a predetermined level, α. α is called the significance level, and is the probability of rejecting the null hypothesis given that it is true (a type I error). It is usually set at or below 5%.