Population and Estimated Parameters, Clearly Explained!!!
HTML-код
- Опубликовано: 29 июн 2024
- One of the most basic and most important thing we can do in statistics is estimate population parameters. This video explains what population parameters are and how they are used to gain insight into the world around us.
NOTE: This StatQuest assumes are already familiar with...
...Histograms: • StatQuest: Histograms,...
...Statistical Distributions: • The Main Ideas behind ...
...and the Normal Distribution: • The Normal Distributio...
...and if you want to learn more about Confidence Intervals, check out this 'Quest: • Confidence Intervals, ...
For a complete index of all the StatQuest videos, check out:
statquest.org/video-index/
If you'd like to support StatQuest, please consider...
Buying The StatQuest Illustrated Guide to Machine Learning!!!
PDF - statquest.gumroad.com/l/wvtmc
Paperback - www.amazon.com/dp/B09ZCKR4H6
Kindle eBook - www.amazon.com/dp/B09ZG79HXC
Patreon: / statquest
...or...
RUclips Membership: / @statquest
...a cool StatQuest t-shirt or sweatshirt:
shop.spreadshirt.com/statques...
...buying one or two of my songs (or go large and get a whole album!)
joshuastarmer.bandcamp.com/
...or just donating to StatQuest!
www.paypal.me/statquest
Lastly, if you want to keep up with me as I research and create new StatQuests, follow me on twitter:
/ joshuastarmer
Correction:
2:16 I meant to say 10 and 30. However, you should still get the point either way.
#statquest #statistics
Correction:
2:16 I meant to say 10 and 30. However, you should still get the point either way.
Support StatQuest by buying my book The StatQuest Illustrated Guide to Machine Learning or a Study Guide or Merch!!! statquest.org/statquest-store/
About 11:19 ... In other words, the results would be if the 2 distributions (i.e., that of the population & that of the samples collected from the population) are overlapping even with high enough statistical power. Would you agree with this?
@@Pl15604 I'm not sure I understand your question, but the more data you collect, the better the estimates tend to be and the more power you have to correctly reject the null hypothesis (if that is something you want do). Having more power to correctly reject the null hypothesis is a way of increasing reproducibility. If you're interested in learning more about the null hypothesis, see: ruclips.net/video/0oc49DyA3hU/видео.html and if you want to learn about power, see: ruclips.net/video/Rsc5znwR5FA/видео.html
I have a slight doubt in liver cell experiment ......if mean is 20 and sd is 10 then mean +/- 2(sd ) is the range for 95% population but in our case 100% population is between 0-40
@@statquest yes but what if the new measurements we took after considering only 2 cells are on left hand side .....then the mean will shift away from actual mean . when you considered 3rd cell or even 4th and 5 th cell you placed those on graph accordingly .....that cannot be the obvious case
The way I PANICKED for a second while my brain was decoding "mRNA transcripts in liver cells". Man, why am I so scared of math I'm weak with relief. Bless you, bless you for the apples.
bam! :)
this is pure dripping free gold
:)
My thoughts exactly.. I mean I am sure the universe is paying me back for all those hundreds of boring unproductive lectures and cruel profs I endured
With all due respect, I love you man. I have been binge watching your videos for sometime now. A big shout out to you and your team who make these videos possible and make them see the lime light. These video's are evergreeeeeennnnn. Your videos have been the learning blocks for some many around the world and many more people to come and start viewing who are not yet born.
The content is very easy to understand. The concepts that I had learnt many year's ago when I look at your video's. They just provide me a different perspective all together.
Thanks for your effort and time that you have put to get these content on this platform.
You have made life easy for so many people. Please keep of doing this great work. Thanks for you and your team. LOVE you man for these fabulous content.
Triple BAM!! initially I use to get irritated with the music bit. But now those are the parts which make them more interesting. Sometime I just watch the video for the music piece.
Thanks Sir. Please don't stop and keep creating more contents. It will help the future generations to come.
Thank you very much! :)
Hey Josh Starmer, you are the miracle I was hoping for, as a physician I do not have time for reading these strange wording and foreign stat books, I stumbled upon your videos randomly while I was looking for a video on ANOVA, now I am a fan, I have started watching your videos from the beginning though they are easy, just to rebuild a stronger base, cannot thank you enough, love and respect, DOUBLE BAM
Hooray!!! Thank you very much! :)
Thank you so much for these videos. They’re the most clearly explained I could find all over the internet. I speak Japanese and Chinese as well. I had looked through all the free sources in those languages. Never ever had I come across anything as clear as this. 🙌
Hooray! I'm glad the videos are helpful.
So glad I found your channel, I really really wish I found it earlier. But here I am, so grateful for all of your videos, thank you so much!
Bam! :)
Glad to see you update this video list. Your videos help me a lot. Thanks Josh. Love you.
Thank you! :)
I'm a bio/stats major in college now and your videos are the best! Thank you soooo much for providing quality content.
Awesome! Good luck with your major.
Never seen such a cool way to make statistics so approachable! super work StatQuest team @North Carolina
Thank you! :)
I'm a big fan of you and your channel. I've only joined since last week though, I have fallen in love with statistics that I used to hate. And I love and play ukulele! You're amazing!
Thank you very much! :)
Just started binge watching StatQuest and this is my 6th video. :D
BAM!!!!
Awesome!!!! I'm so glad you like my videos!!! :)
This is my third try tackling statistics, my old nemesis - only this time, I'll actually succeed. Thank you!
Hooray!!! Thank you so much for supporting StatQuest and good luck! Hopefully the 3rd time will be the charm! :)
Finallyyyyyy.... It was longgggg breakkkkk...
MEGABAMMMMMM
Dude you're an amazing teacher. Thank you for this!
Thank you! :)
This is a really nice resource to quickly revise statistical concepts and that too visually, so this video is Bam!😁👍🏽
BAM! :)
Your videos are great. since im not a English speaker, it is very helpful you have subtitles like this. I'm studying Psychology and I'm afraid of Statistics. Thank you very much.
I'm glad my video is helpful. :)
Is statistics also used in Psychology and in what way? If u dont mind sharing your thoughts... Thank u
wonderful lectures. It was title that attracted me to your videos 'clearly explained ' as I was mostly looking videos with high numbers of viewers.
Thank you! :)
Hi statquest, i really enjoy your job it's thousand time more easy to understand what we use when we are not mathematician due to your videos. Your video about t-SNE helped me a lot. Can i abuse and ask if you can do the same on the UMAP ?
what a way of giving detailed and brilliant explanation!!!!
Superb....
BAMMMMM!!!!!.......
Thank you very much! :)
Congratulations to your content. I am trying to learn statistic not in order to pass in a test ( or stuff like that) but to apply in my job rotine. Your channel focus in statistics concept without a huge load of algebra. I am very happy with this content. Thank you by sharing this aproach of knowledge
Thanks! I'm glad my videos are helpful.
I ironically love this channel. Just became a state major after switching from cs.
Awesome! :)
Amazing!!! I finally understand the meaning of p-values and confidence intervals.
Hooray! :)
Best stats teacher ever!!!!!! Thanks Josh 😄😄
Thank you! :)
Double BAM!!! Thank you for these useful refreshers.
Thanks!
A very great explanation as usual. Could you please make a video explaining "Identification of spatially variable genes"? I don't understand at all about spatial variable genes. My background is mathematics, but I want to learn about this spatially variable genes. Thank you so much.
Be ready for seeing me in anyway at all research areas. I have started learning statistics from best teacher Statquest. I will do a lot of researches with this knowledge. Thanks for your clear English . I am from Türkiye and I have no problem with listening you with my poor English skills. I will focus doing statistics of my own researches. Thanks for all infos. I hope I can keep learning and reach Regression analyzies. I am medical doctor I have to learn using multivariate statistics. Thanks for everything
Good luck! :)
Dude your video saved my grades! Youre a life saver!! ❤️❤️❤️❤️
You're welcome!!
This was the best intro of the whole series
Bam!
THANKS FOR THESE SUPER HELPFUL VIDEOS!
Glad you like them!
Thank you for this I'm a med student and this helped me through a lot of confusion.
Awesome! :)
med students in stats, India is now changing!
Minor correction: "between 10 and 30" 2:14
Thanks! There's always one or two things like that. Oh well.
@@statquest Yeah, but still a great video!
omg i love ur channel!! glad i found it :D
Hooray! :)
What an awesome explaination , Thank you dear Professor
Thanks!
I came for the stats. I stayed for the jingles.
bam! :)
Amazing videos !!! Finally I got, thanks, thanks a lot Dr. Starmer. Some suggest reading s or videos ? Definitely the best !!!
For more videos, see: statquest.org/video-index/
Really Fantastic as usual! Thanks a lot Sir
Thank you!
The question is, how confident are you that you will deliver the p-value StatQuest? Love your videos!
Ha! Good one. One day I want to do a straight up StatQuest on the t-test. However, if you want a sneak peak, check out my series of videos on linear models. Using linear models for t-tests is one of the coolest things.ruclips.net/p/PLblh5JKOoLUIzaEkCLIUxQFjPIlapw8nU
In the stats field, you beats Khan Academy dude!
Thank you! :)
You are really great bro. Helped a lot
Thank you! :)
Everybody love somebody, Thank you, Sir
Thanks!
Love the "I" friendly approach you have in every video 🤩
Thanks! :)
Hey Josh, could you do a new series of videos on probability just like statistics? Your channel will be a one-stop destination for interview prep. Can promise that :P
I'm planning on doing a few videos on probability soon.
I felt the triple bam after the p-value and confidence interval bit!
:)
Damn you're good! I gave cred in another video, subscribed and decided to do this thing from basic. I'm learning programming on my free time in order to attract women and started to doubt my presumtions when reaching Machine Learning. The note you did, what to think of training dataset and predictions in all of this, helped me more than you can imagine. Now I can run a proper test and measure again if my previous result was true, that women indeed tend to like lazy Swedish programmers more than Portuguese Football stars. Tripple Bam!
I'm glad we are both interested in statistics for the same reason. :)
@@statquest I started learning ML and came to stats for the same reason. Double Triple bam. I guess if we take the normal distribution of why people start to learn ML, I am sure its a very high probability that we came here to attract the ladieessss
@@bharathsf Totes!
this video just made me to learn more and more about statastice BAM!
bam! :)
I just want to say Thank You!!!
Thanks!
God bless you for this. This is real education :)
Thank you! :)
Hello Josh, your videos are awesome.
Can you make videos on q Value and Posterior Error Probability? It will be really helpful.
Thanks
I've made one on FDR, which is related to q-values: ruclips.net/video/K8LQSvtjcEo/видео.html
Wooooooo, that is supper great!!!!
Thank you so much! :)
You are awesome my dude
Thank you my dude!
Thank you!!❤️🔥
:)
Your video tutorials are BAAAAMMMMMMMM !!
Bam! :)
Really like your videos, Josh! I am building my statistics knowledge solely with these videos. I have a question at @3:40 that how is standard deviation 10? If we are looking at how wide the curve is, it widens from 10 to 30 which makes a difference of 20 but 20 is mean. I'll probably look at the next video about calculating s.d. but curious for now.
It's +/- the standard deviation. This is explained in the next video.
Thanks a lot sir!
Thanks!
"Imagine mRNA transcripts and genes"
😱😱😱😱
"Imagine apples and grocery stores"
🎉🎉🎉🎉
:)
You introduce R a lot in your pronunciation for examples you say rrrhhhey insteatd of hey and brraaam instead of baaaam..Hhhhhh. I'm from Algeria and it's a great pleasure to learn both american english and statistics with you
bam! :)
everyone whose learning statistics should go through this at least once!
Thanks!
Finding this channel for the first time feels like finding El-Dorado
bam!
Thank you for such god content. I hope such comments motivate you !!
Thank you! Yes! They motivate me! :)
Excelent!
Thanks!
First like and comment. Great video.
Awesome!
Thanks!
WOW! Thank you so much for your support!!! BAM! :)
Could you please share one video about variance covariance structure like unstructured covariance matrix or compound syemetri
I will tell to my teacher to come and learn with me here, rather than wasting our time at university.
You have to teach our teachers how to teach.
Thank you!
Hooray ):
I'm glad my videos are helpful! :)
There are really helpful !
One question : what did you use to make these beautiful slades?
@@nit235 Keynote
I know that this is the beginning of the course and maybe further it will be explained, but I will still leave a small remark. I think that in order to draw a normal distribution in this case one must normalize the data beforehand, because otherwise the fitted normal curve will suppose that negative cases exist. In this example we have 20 as a mean and 10 as a standard deviation. So we have 2.5% probability for negative values, which is impossible.
Great course anyway. Listening just for fun.
Sir please explain the same concept using business data like sales or brand ratings.
the “bam.” Caught me off guard 😹😹😹
:)
Hi,
Just stumbled upon your channel by chance while looking for something to get busy with during this pandemic.
I found your videos easy to understand and I got actually hooked into watching the videos. Great Job and thanks for simplifying the concepts!
By the way just a quick question though in this lesson. In the hypothetical data, you mentioned that most of the cells had between 20 to 30 mRNA transcripts, which represent the positive sigma level. May I know why you only chose the sigma level to the right?
I am just quite lost with the positive and the negative sigma level. Hope you could give further clarifications on this. Thanks
Also, you mentioned that Standard Deviation corresponds to how wide the curve is from the mean or how spread the data is from the mean. From the example, you said it's 10, which made me think why 10? Shouldn't it be measured from the mean up to the last bin/sigma where one can see the data, which in this case is spread to 40. My understanding is from 20 to 30, it's 10 and from 30 to 40, another 10, so I assumed that it's 20..
Sorry about my question, I hope you get my point. Just got curious on this topic. Just want to have a better understanding of the concept. Thanks!
There are lots of ways to describe how data are spread about. One way is to use the maximum and minimum values (in this case, the maximum value is 40 and the minimum value is 0), but another method is the standard deviation. The standard deviation is useful because it corresponds to how wide a normal distribution is spread. To understand more about the normal distribution, see: ruclips.net/video/rzFX5NWojp0/видео.html
Amazing😍
Thanks 😄
The content is really amazing I need these slides to be a reference to me later when I need to remember something.
How can I get these slides?
This is (and a lot of other stuff) covered in my book: statquest.org/statquest-store/
please make some video of how to measure in SPSS or any statistical computer application. I LOVE YOUR VIDEO
I'm glad you like my video. Would it be OK if I gave the examples in R? In R, if we collect a sample of data, called 'data' and want to estimate the population mean, we use "mean(data)". If we want to estimate the population variance, we use "var(data)" and if we want to estimate the population standard deviation, we use "sd(data)".
@@statquest thank you for the explanation, i mean i really enjoy to learn stats from your channel of how to use SPSS and any statistical apps like JASP, as the technical tools for statistical problem solving ... Maybe you may add the following step of using the SPSS after the explanation of any formulation of all kind of topic you provided :)))...
, just some simple tutorial video of how to use spss and any computer stats software...
I really enjoy the way you explain... Hope the best for you, #StatQuestaddict. (besides the fact that my stats lecturer explain it with the most complex way that maybe only just 5% of the students understand it)
@@fairuzam1420 I'm glad you liked my videos. I actually contacted the people at SPSS and asked them to give me a copy of the software so I could add steps to my videos, but they said no... :(
I WAS CRACKED UP when you're using mRNA as example and suddenly change it to apple, thats hillarious thanks for making me laugh TuT
:)
Thanks
:)
Has un vídeo sobre como calcular las probabilidades de una funcion de densidad usando la formula e integrando.
I'll keep that in mind.
Lovely vedios
Thank you!
Tx sir
Glad you are enjoying my videos!
Could you also please make a video on Probability density function, probability mass function and CDF.
I'll keep that in mind.
@@statquest thanks so much.
can you provide me notes of total playlist? thanks for better explanation. you are the best.
I have study guides available here: statquest.org/studyguides/
Means no matter what the estimates of the sample are ....if P values of these samples r nearly equal then ....we have same confidence in their estimations
I'm just refreshing my stats knowledge and ur videos are so helpful! I'm having a question though: I didn't get the part about the Exponential distribution as well as the Gamma distribution and how it can be compared/used like the normal distribution (5:57 - 7:10). Could somebody help me, please 😵💫?
There are lots of distributions in statistics. The normal distribution is one, and there's another called the exponential distribution, and there are many more. However, all of them can be used in the same ways. If we collect a lot of data and draw a histogram, we can fit a distribution to it and use it to calculate probabilities (by calculating the area under the curve or with the histogram).
@@statquest Ah, got it. It appeared to me that they are kinda the same as the normal distribution which made no sense to me. Thank you so much for ur reply and keep up the great (and very helful) work! All the best from Germany :)!
@@claudiabeck1443 Happy to help! :)
Nice tutorials, how did you create a graphs? I need to create a similar to yours but didn't find any good tool.
I draw them by hand in keynote.
i'm here because i like the way you sing🤣
bam! :)
I have a question. I am having hard time understanding if Probability Distribution is the same thing as Statistical Population?
Or, is Population related to all events of interest that have had already happened and that we wish to analyze using statistics. But, since its to costly to gather information/mesurements regarding entire population, we gather a sample by which we estimate population parameters and approximate population curve.
When we actually approximate population curve, does that mean we have a approximation Probability Distribution, or is probability distribution concept related to sample space only? If so, how do we calc. probabilities for each element of the population?
Sorry for long question, but these terminologies between Probability and Statistics REALLY confuse me. Please help me understand :(
Also, really appreciate your work! I am learning A LOT!
I have a StatQuest that might answer your question. It is called: What is a Statistical Distribution: ruclips.net/video/oI3hZJqXJuc/видео.html
Waiting for Fourth BAM!
:)
top class
bam!
Thanks for the videos for giving us more insight on statistics. I have a doubt. How can we calculate the mean and SD of an image when the image need classification with multiple classes( say the image need to classify into 6 different color patterns)? ie, the behavior of each data might be different from another. Can you please address to my query?
For the training data set the population mean and population SD can be found. But it can be treated with an image(testing data) where it is grouped with different patterns(or classes )
To be honest, I'm not sure I understand your question. The mean and standard deviation are metrics that are applied to quantitative numeric data, not discrete data like images.
Does it only work for features that can be ordered? (e.g. by heights, or by number of mRNA transcripts)
What if we count things having different colors, for example? (like, 3 green apples, 7 red apples, 5 yellow apples, 1 rainbow apple)
The concept of population and estimated parameters applies to all kinds of data. However, when we have discrete data, like your example, we often have different parameters other than just the mean and variance.
Are you a friends fan?
Because your introduction resembles me about Phoebe song 😂
See: ruclips.net/video/D0efHEJsfHo/видео.html
@@statquest oh my god 😂😂
I was right
Need these slides for revision
Hi Josh, Could you explain me how did you calculated "The area under the curve for all values equal to or greater than 30 at 4:20 and also what are populate rate and population shape
We use calculous to integrate the equation for a normal curve from 30 to infinity. Also, the normal curve does not have a rate parameter, so there is no population rate. However, other distributions, like the gamma distribution, have a population rate. These are just like the mean and standard deviation we have for the normal distribution. They define the shape of the curve.
@@statquest Thanks for your reply Josh. i got my answer
Sir, where I can find the details about the "Central Limit Theorem"? You have mentioned in your video about it.
Here's the link to the video:
ruclips.net/video/YAlJCEDH2uY/видео.html
...and here's a link to an index of all of my videos:
statquest.org/video-index/
This is a helpful video, but still hope to find an example of discrete probability distribution's parameters.... Such as binomial, poisson, geometric... Parameters... Because parameters distributions are not ebough clear for me
I have a video about the binomial distribution here: ruclips.net/video/J8jNoF-K8E8/видео.html
Yes I saw it and was very helpful , but actually I talk about parameters.... I wanna find a video about thr PARAMETERS of probability distribution
BAM!
Bam! :)
Do I understand it correctly that the more and more data we the more estimated parameters get closer to the mean and STD is an application of central limit theorem?
The more data we have, the better we can estimate parameters and less there will be less variation in those estimates if we do them a bunch of times.
8:04 says "the new measurements will come from the same population" sounds like "come from the same distribution?"
1. How do experimenters (both the 1st and subsequent people) know what distribution the samples come from and thus what population/distribution parameters to estimate?
2. What happens when someone later proves that the samples come from some other distribution that's not normal, does that invalidate all previous experiment results?
3. Even within the same family of curves, experimenters may use different tests which have different degrees of freedom, does this cause a reproducibility problem?
4. What does it mean for an experiment to be reproducible, or to have reproduced a previous experiments? Are the estimated population parameters and their confidence intervals part of defining reproducibility?
Han, the good news is that all of your questions can be answered by the Central Limit Theorem. In a nutshell, it says that regardless of the underlying distributions, the means (averages) are normally distributed and this means we can use basic tests (like the t-test) to compare means regardless of the underlying distribution. For more details on the Central Limit Theorem, see: ruclips.net/video/YAlJCEDH2uY/видео.html
BAM!!!
Thanks!
Is there any explanation for the rate and shape parameters?
Different distributions have different population parameters. The normal distribution has the mean and standard deviation. Other distributions have other population parameters. For example, the Gamma distribution has rate and shape parameters. However, the concepts are the same - these population parameters determine the the properties of the population distribution.
Bam!!!
:)
3:28 is this really a normal distribution ? The number of transcripts can only be positive integers. Therefore the distribution is not continuous and rather corresponds to a binomial distribution, which effectively look really similar but are not exactly the same. Or am I missing something here ?
If the number of observations is large enough, then a binomial distribution is accurately approximated with a normal distribution. So that's what's going on here. See: en.wikipedia.org/wiki/Binomial_distribution