Correlação não implica causalidade: Perspectiva de um minuto sobre correlação vs. causalidade
O que veremos neste artigo:
Entendendo a complexa relação entre correlação e causalidade
Quando se trata de analisar dados estatísticos, é crucial compreender que correlação não implica causalidade. Mesmo que duas variáveis apresentem uma relação forte, não podemos presumir que uma cause a outra. Este breve artigo explora essa importante distinção, trazendo à tona a necessidade de cautela ao interpretar dados estatísticos.
Correlação vs. Causalidade: Compreendendo as Diferenças
Antes de mais nada, é essencial compreender a diferença entre correlação e causalidade. A correlação indica apenas a presença de uma relação entre duas variáveis, ou seja, quando uma aumenta, a outra também aumenta (ou diminui). Por outro lado, causalidade implica que uma variável é responsável por provocar mudanças na outra.
Um exemplo emblemático dessa distinção é a relação entre as vendas de sorvete e o número de homicídios em Nova York. Embora os dados mostrem uma correlação positiva entre essas variáveis, isso não implica que uma causa a outra. Na realidade, o clima ensolarado é o fator oculto que impacta tanto as vendas de sorvete quanto o aumento dos homicídios.
Fatores Ocultos e a Importância do Contexto
É crucial considerar os possíveis fatores ocultos que podem influenciar a relação entre variáveis aparentemente correlacionadas. Ao analisar os dados, é essencial dar um passo atrás e avaliar o contexto mais amplo. Muitas vezes, descobrimos que a correlação observada é meramente uma coincidência, sem qualquer relação causal genuína.
Assim, ao interpretar dados de correlação, é fundamental considerar o cenário mais amplo e investigar possíveis fatores que poderiam influenciar a relação entre as variáveis em questão.
[vid_player]
FAQ
Qual é a diferença entre correlação e causalidade?
A correlação indica a presença de uma relação entre duas variáveis, enquanto a causalidade implica que uma variável é responsável por provocar mudanças na outra.
Por que é importante considerar fatores ocultos ao analisar dados de correlação?
Fatores ocultos podem influenciar a relação entre variáveis, tornando crucial uma análise contextual mais profunda para evitar conclusões errôneas.
Conclusão
Em síntese, a compreensão da distinção entre correlação e causalidade é fundamental ao analisar dados estatísticos. A cautela ao interpretar a relação entre variáveis é essencial, garantindo uma análise precisa e livre de conclusões precipitadas.
Bro you change my view to the world
My happiness is positively correlated with seeing new videos from this channel in my sub. box!
Beautifully explained. Wow
These are so great. You're one of my favorite Youtubers out there.
Short and to the point!
When you say 'weather', how do you quantify that? I get that there would be temperature, but people seem to talk as though things that aren't variable quantities correlate, like facts and events.
not the best example as its not really clear that weather should / could affect homicides .
a more clear example could be age and arm length which of course correlate due to growth
this video should autoplay before and after nutrition science YT videos
One Minute Economics needs your help! Please give me a minute (heh) of your time by watching the following video if you find the channel useful, literally anyone can help (either financially or by spreading the word about my work): https://www.youtube.com/watch?v=io04ckq1X1M
tysm for clearing my concepts;)
reminds me of "poverty is the underlying cause of crime".. easy to debunk but yet people (especially academics) love to push that narrative.
im only 5 seconds in and- is this gru? so gru got a youtube channel and no one bothered to tell me? rude.
Your example just made me laugh
So hot weather causes homicides, noted.
Awesome! don't get why professors can't explain it so simple and clear
Death with Covid ≠ death from Covid.
Thank you!!
I just found your channel and am currently binging all your videos. I absolutely LOVE your content and can't thank you enough for taking the time to make them. You put broad concepts into simple terms.
Do you have a venmo account by chance? I don't have PayPal because I don't agree with their principles, but I'd love to donate!
I often find this is misused to make the argument that correlation implies no causation. People simply state correlation does not imply causation and think it wins the argument.
Had to literally text this video to one of my friends
The correlation is people fighting over ice cream to the death when it's hot out.
What if homicides cause ice cream sales?
Thousands of people have died suddenly and unexpectedly hours to days after warp speed injections from the same fatal conditions such as blood clots, heart attacks, and blood platelet failure. Doctors have confirmed many of these deaths to warp speed. Also, many warp speed recipients have died suddenly and unexpectedly but their deaths have not been ruled on by medical examiner. Can one infer that warp speed killed them? Or would it be better to say “Correlation does not imply causation.”?
100% of people who drink water die. This does not imply that drinking water causes death.
Important update! The situation in Ukraine is BAD and action must be taken. If this channel has been useful to you at least once, please give me a minute (!) of your time by watching the following video dedicated to what I believe must be done from an economic perspective (anyone can help): https://www.youtube.com/watch?v=nuZI_qBHyaY
The phrase should be "Correlation does not ALWAYS imply causation." There are many times that it does.
Yeah me and my crush…. Thinking i m getting it but no she asked me to stop hoping lol
Perfect!
Tell this to the Trumpies who blame Biden entirely for the high gas prices
I feel marriage counselors should use this
IT EXPLAINED MY IGNORANCE
A few years ago Redditors got super-exited to correct people with 'Correlation is NOT causation' and like this video misrepresenting it with the vs. or 'Versus'.
Correlation *could imply causation (we may not know yet) and so let's take the ice-cream sales.
Now, it could be true, as some experts say, the hotter the weather the more impatient and aggravated people get. Let's keep that in mind but put it aside here.
It can also be that hot weather often happens during summer which coincides with schools being out and sometimes it's where more people go outside.
Then upon some investigation the police and paramedics have started noting that on Martin Luther King Drive and Malcolm X near a large sprawl of urban welfare housing there is a tremendous problem with:
Rival gangs and criminal enemies looking for each other to murder each other,
AND
A problem with severe domestic violence that often gets worse in school break, summertime, often over custody, alimony payments and child support problems.
HOWEVER,
These are often mitigated by everyone staying inside their homes and rarely coming across each other face-to-face in the street itself. Especially in winter but worse in summer as again, more people go outside and less are wearing hats and hoodies etc.
There is a lovable older lady who, every time she hears the 'Ding Ding' and music of the incoming Ice-Cream Truck she comments "Here come the Rachet!". Apparently 'rachet' means something like traumatic fighting, arguing, a crime and chaos scene is coming.
Why does she say that and her son nods like he totally gets the meaning?
Because the demand for cold refreshing ice-cream will come at the peak of the hot blazing sun and guess what?
everyone and their kid forgets all risks and goes for that ice cream sale and every day at least one 'baby daddy' blows up on a 'baby momma' who claim she aint got no money but how she buying ice-cream out in here?!"
and a 'baby momma' whos got a restraining order against her shows up to find baby daddy and the kids and tries to stab him in front of the ice-cream truck.
and then that 'Rachet Ice-cream sales truck' gets worse because guess who's been waiting to shoot Lil Cornflake (in revenge for his gang shooting their Big Baby Jones) and guess what, they figured out Lil Cornflake got no AC in his crib so now they know he comes out to the Ice Cream Sales truck every day for Blue Freezies
That sound of the ice-cream truck is the sounds of 'get the guns loaded and homicide!'.
By the way, there is NO BETTER time to homicide a 'Lil Cornflake' or any enemy than when they are coming out of the story or ice-cream purchase, carefully holding that ice-cream cone in their right hand and trying to stuff the change or wallet back with their left hand while being drawn into that delicious fragile cone mmm. BANG BANG he never even let go of the ice-cream to reach for his gun.
It's actually on the grainy YT video where, no joke, AS bullet hit him and he turns to run, he actually holds the ice-cream cone like he's trying to save it from dropping or falling. Habit I suppose.
Of course, we aren't not supposing that the ice-cream sale itself suddenly made a receipt for the sale turn into a bullet or knife and eject itself, hypnotize the soon-to-be-murderer into throwing it into a victim.
What we MAY well find is there most definitely IS a 'Causation' from Ice-Cream Sales and Homicides.
Keep in mind, there absolutely was a direct 'causation' between snowfall in Toronto City and fatal heart attacks.
It's actually the exact same causation between the first week of NO snow (snow melt).
Now what?
GiganticWebsites.com is a project through which I make it possible for people to build truly gigantic websites (thousands of articles each!) at ridiculously low prices. If you have a great domain you want to turn into an amazing website or an existing site you'd like to upgrade/scale, visit our website or check out the One Minute Economics presentation video below:
https://www.youtube.com/watch?v=gE8yEOQFMvo
Please note that this comment is not an ad for a third-party service provider. GiganticWebsites.com is my baby 100% and I will personally be involved in each and every project so as to ensure the website turns out great 🙂
I have a question but I do not know how to ask it, how can you prove a cause in a debate, or I suppose how can experiments ever be proven as fact, if I stab someone and they die within three seconds how can I prove that the knife killed them and that they didn’t suddenly drop dead from a heart attack at that exact moment, what if you injected someone and they have a heart attack five minutes later, what if you injected another person and they also have a heart attack, then you did it again and again and now 100,000 people are dead from heart attacks within five minutes after the injection, out of 150,000 how can you prove that yes what ever you just did is the reason when someone simply says correlation does not equal causation, and they do not propose any other possible explanation, they simply believe they have won the debate, also can’t everything fall apart under further study, you could say no the knife didn’t kill this man, his artery was severed and he bled out, no it’s not because it was cut it’s because he lost this amount of blood, no it’s not because he lost blood it’s because these places the blood is meant to go did not receive blood, and on and on until you get down to individual cells and atoms, everything can be deconstructed but at the end of the day you stabbed him and now he is dead, how can this be proven if one can just say, “Correlation does not equal causation” what is the correct response to that to prove two things are connected.
What about my climate change?