Correlation and causality | Statistical studies | Probability and Statistics | Khan Academy
ฝัง
- เผยแพร่เมื่อ 14 ต.ค. 2024
- Courses on Khan Academy are always 100% free. Start practicing-and saving your progress-now: www.khanacadem...
Understanding why correlation does not imply causality (even though many in the press and some researchers often imply otherwise)
Practice this lesson yourself on KhanAcademy.org right now:
www.khanacadem...
Watch the next lesson: www.khanacadem...
Missed the previous lesson?
www.khanacadem...
Probability and statistics on Khan Academy: We dare you to go through a day in which you never consider or use probability. Did you check the weather forecast? Busted! Did you decide to go through the drive through lane vs walk in? Busted again! We are constantly creating hypotheses, making predictions, testing, and analyzing. Our lives are full of probabilities! Statistics is related to probability because much of the data we use when determining probable outcomes comes from our understanding of statistics. In these tutorials, we will cover a range of topics, some which include: independent events, dependent probability, combinatorics, hypothesis testing, descriptive statistics, random variables, probability distributions, regression, and inferential statistics. So buckle up and hop on for a wild ride. We bet you're going to be challenged AND love it!
About Khan Academy: Khan Academy offers practice exercises, instructional videos, and a personalized learning dashboard that empower learners to study at their own pace in and outside of the classroom. We tackle math, science, computer programming, history, art history, economics, and more. Our math missions guide learners from kindergarten to calculus using state-of-the-art, adaptive technology that identifies strengths and learning gaps. We've also partnered with institutions like NASA, The Museum of Modern Art, The California Academy of Sciences, and MIT to offer specialized content.
For free. For everyone. Forever. #YouCanLearnAnything
Subscribe to KhanAcademy’s Probability and Statistics channel:
/ @khanacademyprobabilit...
Subscribe to KhanAcademy: www.youtube.co...
Done with whole statistics playlist and I can say it was a great help. Now I believe I have the working knowledge of the topics covered in this series of lectures. Thank You!
Unexpected sudden ending of 68 video's. Thank you very much for clarifying all this material for me, your explanations are very clear and easy to watch!
I love this talk. This is one of the best treatments of Causality vs. correlation.
I come back to this talk every few months to re-inforce the arguments...it is really necessary to apply these lesson to so many things in life (Global warming, most of Modern cutting edge medicine, etc)...Remember they are measuring correlation and selling it as causality, to get FDA approvals or more research money.
@@datingprofile what's your profession?
Just finished the playlist. It is remarkable how Sal has the talent to build concepts from ground up making the student really got into what is going on. This is a great example of humility, something usually missing on education (specially higher).
I really like these videos from Khan Academy. I encourage my students to use them to do independent study. There is a definite correlation between using the videos and success in the examinations, but not proof of causality yet :)
I now finished the whole playlist to get some introductory ideas about statistics. I think I got some now. It really provides a good value for learners, great videos, he explains well to make understand intuitively. He's great.
"Breakfast is the most important meal of the day" says the breakfast-related industry
I owe you big time, Sal. Nicely and Perfectly articulated the concepts. Thank you very much.
So basically causality only works in one direction and correlation is 2 things that move together and are related.
My thoughts exactly
In all these examples, yes. But you could also have 2 things that are correlated completely coincidentally. See for example the website/book "Spurious Correlations".
@ZergAteu The researchers probably understand the difference and may have even been careful with their wording. The article, however, is clearly trying to have the reader walk away with the thought that eating breakfast makes obesity less likely.
This is to the point. As I watched I kept thinking about the many articles that "might be wrong" that come out on a daily basis. These are powerful concepts to understand and apply regularly.
Thanks Prof for these mind boggling videos. Studied stats in my MBA from one of the most revered Faculty from an IIM and he was nowhere half as good as you. I had no idea stats could be this much fun. Thanks a lot. A bit sad to see the abrupt ending of 67 videos.
Same for me
Wow! This video, apart from explaining the difference between correlation and causality, tells us why it is important to know the difference too. Causality is a correlation. However, not all correlations are causalities. What we need is the objectivity to see a causality as a causality and a correlation as a correlation. Above all, we must have the ability to picture the interactions between a multitude of correlations and causalities while evaluating an event.
After all these years of hating... I finally start liking statistics :)
I am really sad because of the fact that this is the final lesson.
Thank You Sal!
I Appreciate for all the statistics videos You are the one of the best teachers I ever meet！Your teaching is clear and kinda funny sometimes make me laugh lol anyway thank you again！！
Causality always creates correlation, except in a few pathological cases. We generally say that X causes Y when doing X changes the probability of Y occurring. As a result, X and Y will be correlated. The point of this video was to say that correlation does not imply causation. The converse, however, is generally true. Causation implies correlation.
Correlation is saying that two things can go on simultaniously and cause and effect is saying that one thing happens which affects the next thing in the sequence. Directing this to the video, correlation is saying that sleeping late could impact your ability to eat breakfast because maybe they dont have enough time and vice versa and cause and effect is saying that because they are obese, they could not eat breakfast. That is just an example.
Hi Khan. I'm a big fan of your work and like following you.
I hope you don't mind if I make a few suggestions. I think you did a great job at differentiating between causality and correlation.
A few things I would recommend mentioning in addition to this are: timing (which came first?), co-linearity (overcoming it to analyze causality), and randomization or multivariate models (how it takes care of the confounding issue you mention). Good luck and great job!
I wish more people thought like this, not taking things at face value, making up their own mind. would certainly help with the media and their inflexions towards news and current affairs.
Thank you Professor for this amazing videos. I watched the whole playlist of Statistics and I find it inspiring. Longing to follow with another playlist soon.
This video is simple and to the point. Great video!
Thank you so much for helping our minds to open up
THE END.
Thank you so much Sal for this playlist. :)
Ok i ve seen an article telling that most rock climbing accidents happen by men rather than women. But as far as I know there are more men that goes rock climbing than women. Can we say there is an correlation or how can we implement it in causality, can you explain i really didn’t get it. The more men rock climbing causes more accidents, am I wrong?
A fluctuation of A and B tend to be observed at the same time, Instead of A causing B; That's a very good example to illustrate Correlation and Causation, in my opinion.
Ideally, the authors would make the notes that Khan has made and provide some alternatives to causation as Khan did. This way, when the reader reads the article they don't come away thinking, "if I want to lose weight I just need to eat breakfast" but instead, "people are researching ways to help me lose weight, but they still don't know much besides "eat healthy and exercise".
Thank you sir for this playlist. You are excellent sir.💕💖💖💖
Great. I was also hoping you'd throw in an explanation of what a logical corollary is. Although it's implied throughout. Still it's easy to confuse that with a correlation unless you know why they can't be the same thing.
This series ending was more abrupt than The Sopranos cut to black. Thanks, Sal!
Really appreciate the info!
What is causality but correlation between two variables?......... mind blown.
Great playlist Sir. Thanks. This helped a lot.
Excellent contents - God bless you
Great videos. Loved all of it. I feel the regression part could be expanded to talk of linear and multiple regression analysis, and also explaining how to interpret excel outputs of regression. But overall, really good!
THANKS, THAT'S A BEAUTIFUL LECTURES.
Great piece here.
The funny thing is, even as a fairly fit guy, in the military, I started struggling with weight management in my thirties like a lot of people do, and one of the first things that really helped me start turning it around was compressing my eating window. Let's also be sure to acknowledge the fact that people dealing with poverty and food insecurity often eat the worst possible macronutrient composition (high fat + high processed carbs + low protein), which can be obesogenic via hormone signaling even if not massively overconsumed.
Thank u sir, you have expanded my way of thinking or looking at things to a different level, your videos are very helpful and I hope you make more of these useful videos. If you need our help through donations so you can make more videos, than add a link and I will gladly support.
Completed the entire series. Loved it!
Great examples! Well done.
This was an awesome journey Sal. Lots of respect and thanks.
I could not take the ending seriously! How he tackles that example near the end of the video made me laugh.
Informative ❤
great video! thank you
@ZergAteu
I didn't vote it down :-)
You might be underestimating the influence the title has over the way the rest of the article is read. If it starts out with "X may beat Y", then the reader will interpret the content as being in line with that summary. It's a tactic frequently used by tabloid journalists.
I MADE IT!!! Thanks for the videos, Sal, you helped me so much :)
Great Video
This playlist has been a great help sir, Thank you for the hard work.
Thank you.
So... since life is filled with variables, how does one conclude actual cause and effect other than in a controlled lab or controlled environment where even those are often misinterpreted based on unforeseen conditions?
This question is why "social sciences" are seen as weak, or not rigorous -- because it is simply impossible to answer this adequately.
thank you so much
best video on the subject
@ZergAteu
If what the study implied was a correlation rather than a causation, they shouldn't have said "eating breakfast may beat teen obesity" because it singles out one correlated factor. Why not name it "getting up early may been teen obesity" or "being more active may beat teen obesity"? Don't get me wrong, I agree with you that the article does not imply causality, but that's only because it's filled with weasel words.
that was awesome ,i think you only have to eat when you are hungry eating more doesn't make u skinny.
Thanks
Excellent video lesson, Sal.
This is a good teaching
Great post. This will save me so much tiresome re-explaining of bad health journalism to friends!
That how you end a series on statistics! With a 'Random' video :)
굉장한 설명이었다
Thanks a lot for the videos, really appreciate it. I was wondering if you've posted any videos on SPSS?
so does causation is more powerful instrument to proove our hypothesis than correlation?
As I was listening to the first couple of seconds I thought that maybe american titan General Mills is probably the one that funded that study cited in webmd. Anyway, great video, very khana - educational. People on jury duty should probably see this video before making any decision.
Very Helpful. 67/67.
My brains 🧠 , trying to process these information.
What are differences between correlation and magnitude? What is magnitude?
Amazing how writers in the medical field would write these kinds of articles. Either a lack of education or good amount of corruption
Do a tutorial on Correlation and Regression please!
I understand the message behind the video, but I disagree with Sal in that WebMD was indicating causality. MD used the words "may" and "seem," which are simply indicators of possibility. If MD would have used "will" or "does," then I would have called them on bad write-ups.
Great video Sal!
More about reading research papers please!
Even the same topic with a different example would be good.
Cheers!
How to know if correlation is causal through historical data, no controlled experiments?
I usually have coffee for breakfast. Now I'm wondering if I should at least consider brunch.
beautiful
Causality and correlation are synonymous?
So the following is true?
As the number of people who breath air increases, the number of people that die increases. Therefore, breathing air causes death.
The more people there are, the more likelihood of friction and conflict, which breeds a ton of other reasons which lead to death.
Yes!
Is this video about math or obesity? I can't tell.
Is obesity about math or breakfast? I can't tell. lol
Clearly about fuirt loops.
irrelevant comment. This topic forms the basics of the Data science analysis and very much related to statistics branch of mathematics
hmm one serious student who doesn't have a sense of humor
@@tkhasimbashaa This comment was irrelevant too. This topic ACTUALLY critically analyses the importance of ethical interpretation in statistics. n= 2n. bye.
In this case, the readers do not likely know the difference between causation and correlation so the weasel words result in the reader believing something is true and backed up by science when really it isn't true (not that it's false, just that it's not proven to be true or false yet).
So, how do you know if there is causality?
[INTERPRETATION
] what makes you so confident in case you encounter a "realistic" causality? is there any robust academic foundation? I couldnt find any clue in your
greatest lecture in this clip. Am I an idiot? I would politely ask you to answer or explain bits on this stupid kindergargetn level question please.
carbs are essential, I hate that there is this stigma behind them. Every cell in your body runs of glucose, aka simple/complex sugars... sorry just had to get that out there. great video!
+indigochick Yes, all three macro nutrients are important, all three of them have been demonized at some point in history when it comes to weight management.
It is still hard to believe for people that weight management is just as simple as: total kcal intake and total kcal expenditure. But you know, that is not fancy enough for articles etc.
For correlation you can’t say it will make things better
@ZergAteu
@qtutoringhelps
I believe a big part of the problem here lies in the target audience. If a paper was published in a scientific journal that used a weasel word here or there the readers, other scientists, would understand the difference between causation and correlation (hopefully). However, the referenced WebMD article is meant to be an easy to digest summary of the study fit for consumption by the masses (non scientists).
I think giving a more obvious example (e.g. the correlation between the number of known stars and the population of earth) would've been a better way to start, like c0nc0rdance did in his video on the subject.
So smoking correlates with cancer and does not cause it? Smokers tend to engage in other risky activity which promotes cancer?
i dont still get it. where's the correlation in there?
I hope you reply me. I really need the answers for our reporting
From what I understood, the text and the research that was talked about were more about comparing the patterns of people who ate breakfast and who didn't and that was that. Rather than saying breakfast causes you to lose weight, which may or may not be true in each case.
causality and correlation are synonymous. It doesn't matter if you eat cinnamon buns or bran muffins in the morning you will become overweight at some point in your life if you keep this type of diet. This study is typical research news the public is fed via the media. If the public is generally ignorant(or skeptical) of research methods how are they supposed to make informed decisions about diet.
When can we get to the really cool stuff(like logistic regression)?
Well explained case study
Lawyered.
@alique087
We cannot judge the intentions of the article but we can attempt to draw conclusions about the effects the article has on it's target demographic (non-scientists). A non-scientist reading this article is likely to come away with the impression that eating breakfast will help them lose weight. The article does not technically show causation, but it does imply causation to the average reader who is unfamiliar with causation vs correlation.
Then how to identify causality?
Yeah thats what I was thinking during this too!
u do it by randomized controlled experiments....divide into two groups randomly, and then give them your breakfast, etc..but have to be really randomized.
You try to induce B by applying A to a test group, rather than simply waiting for A to happen and looking if B is there too.
What is the conclusion of the video?
causation does not imply correlation.
Maybe breakfast eaters are more active because they are actually awake all day instead of sleeping half the day then waking up and not doing shit.
Hey, can anybody helps me. I am literally confused thatn is that topics is in term for python as well or not?
it makes me more carefull to conclude the result, but i become more confuse now
How about argue the difference between common sense and nonsense?
@blonkm You're right.
Salman Khan! Man Of Science! All Hail!
This video caused me to want bacon and Fruitloops.
Sorry, I think I wrote this in hurry and forgot to correct myself afterwards. Correlation is not necessarily causation but that isn't holy writ either. My point is that if the public doesn't know how research methods are conducted they can't make informed decisions.
I usually eat relatively high protein breakfasts.
I think eating breakfast will not help, if your breakfast is made of big mac, three cheesburgers, fried potatoe and a big coke portion...