🎯 Key Takeaways for quick navigation: 00:18 🎥 *The video aims to define causality and explore how to understand causal effects.* 01:13 🚑 *Causality involves changing a treatment variable (X) to observe a corresponding change in the outcome variable (Y).* 02:25 🍦 *Correlation between variables does not imply causation, as illustrated by the example of buying ice cream and wearing shorts.* 03:49 🧪 *Running experiments with randomized treatment assignment is a powerful way toestablish causality, but it's not always feasible.* 05:26 🔄 *Understanding the counterfactual, what would have happened without the treatment, is crucial in determining causality.* 06:49 🌐 *The fundamental problem of causal inference arises from the inability to observe the counterfactual, leading to the need for sophisticated tools for estimation.* 08:41 🛠️ *The series will explore various tools and methods to understand causality when experiments are not always possible.* Made with HARPA AI
If something changes the probability of an outcome doesnt that make it a correlation and not a causation? If it changes some outcomes and not other that means there must be a variable unaccounted for that results in different outcomes and that variable might be the actual causal property. Am I wrong?
If *manipulating* one variable changes the probability of another we'd call that a causal relationship by the standard causal-inference definition. On the other hand, if *observing* one variable changes the probability of an outcome, that's correlation. For example, if I observe you're wearing shorts, that makes it more likely you'll be eating ice cream (correlation). But forcing you to put on shorts doesn't make it more likely you'll eat ice cream (no causation). Changing some outcomes and not others doesn't require an additional variable as long as the probability is changed for all of them (if of ten people, two would have rolled a 1 on a six sided die, but me replacing all the dice with twenty-sided dice reduces that to 0, everyone's chances of rolling a 1 were casually reduced even if eight people saw no change in outcome). (note: the definition of causality used in the law, as opposed to in casual inference, would only count an effect as causal if the actual outcome, rather than the probability of the outcome, changed) Also, even if there is an additional variable, that's still causation. If a drug reduces chances of a disease by 10% for women and 20% for men, the drug is causally effective for everyone even if gender mediates that causal effect.
@@NickHuntingtonKlein Thanks that makes sense or it will after.I let it roll around my brain for a.while. another question if you dont mind. All neuro science is correlation because consciousness is always subjective. Can thoughts be causally related to prior thoughts given the lack of any mechanism that we understand?
@@jlrinc1420 the same causal pathways are there whether we understand them or not. I don't see any reason why there couldn't be causal relationships between thoughts. For example if I tell you a passphrase, like "toaster bunny", that will cause you to think of the phrase, and that current thought will cause you to think of it again later when you need it. Also, just because something is subjective doesn't mean it's correlational! If, say, learning that a wine is expensive causes people to give it subjectively higher ratings, knowledge of the price casually affects those ratings, and that's a valid causal mechanism.
I haven't even seen the full video yet, but i'm liking it cause you have Dixit on your shelf!
🎯 Key Takeaways for quick navigation:
00:18 🎥 *The video aims to define causality and explore how to understand causal effects.*
01:13 🚑 *Causality involves changing a treatment variable (X) to observe a corresponding change in the outcome variable (Y).*
02:25 🍦 *Correlation between variables does not imply causation, as illustrated by the example of buying ice cream and wearing shorts.*
03:49 🧪 *Running experiments with randomized treatment assignment is a powerful way toestablish causality, but it's not always feasible.*
05:26 🔄 *Understanding the counterfactual, what would have happened without the treatment, is crucial in determining causality.*
06:49 🌐 *The fundamental problem of causal inference arises from the inability to observe the counterfactual, leading to the need for sophisticated tools for estimation.*
08:41 🛠️ *The series will explore various tools and methods to understand causality when experiments are not always possible.*
Made with HARPA AI
Causality- when you sacrifice your best friend and your crew to monsters you'll get superpower
Berserk 😀
GRIFFITHHHHH!!!
I was looking for this comment
great video man , very informative and simply explained , My professor didn't explain half as well
Hey! Nick. Really great examples
great video, Nick, as always!
Hi Nick, really helpful! Would you consider doing a paid one on one tutoring session online?
Sure! My email is on my website at nickchk.com
Understandable😍
Hi Nick! A great video again.
Thanks a lot for the videos!
This topic is so interesting to me for some reason lol. Thanks
Trying to understand berserk be like
Please do you think that causality in english grammar is conditional sentences
Not by definition, but you would probably want to use conditional sentences to describe a causal effect
If something changes the probability of an outcome doesnt that make it a correlation and not a causation? If it changes some outcomes and not other that means there must be a variable unaccounted for that results in different outcomes and that variable might be the actual causal property. Am I wrong?
If *manipulating* one variable changes the probability of another we'd call that a causal relationship by the standard causal-inference definition. On the other hand, if *observing* one variable changes the probability of an outcome, that's correlation. For example, if I observe you're wearing shorts, that makes it more likely you'll be eating ice cream (correlation). But forcing you to put on shorts doesn't make it more likely you'll eat ice cream (no causation).
Changing some outcomes and not others doesn't require an additional variable as long as the probability is changed for all of them (if of ten people, two would have rolled a 1 on a six sided die, but me replacing all the dice with twenty-sided dice reduces that to 0, everyone's chances of rolling a 1 were casually reduced even if eight people saw no change in outcome). (note: the definition of causality used in the law, as opposed to in casual inference, would only count an effect as causal if the actual outcome, rather than the probability of the outcome, changed)
Also, even if there is an additional variable, that's still causation. If a drug reduces chances of a disease by 10% for women and 20% for men, the drug is causally effective for everyone even if gender mediates that causal effect.
@@NickHuntingtonKlein Thanks that makes sense or it will after.I let it roll around my brain for a.while. another question if you dont mind.
All neuro science is correlation because consciousness is always subjective. Can thoughts be causally related to prior thoughts given the lack of any mechanism that we understand?
@@jlrinc1420 the same causal pathways are there whether we understand them or not. I don't see any reason why there couldn't be causal relationships between thoughts. For example if I tell you a passphrase, like "toaster bunny", that will cause you to think of the phrase, and that current thought will cause you to think of it again later when you need it.
Also, just because something is subjective doesn't mean it's correlational! If, say, learning that a wine is expensive causes people to give it subjectively higher ratings, knowledge of the price casually affects those ratings, and that's a valid causal mechanism.
@@NickHuntingtonKlein Thanks will let this ferment a while. Appreciate your time
Thnks for reply but I am teacher of english grammar our principal gave me topic 'causality' for teaching so I am confuse how to teach it.
Oh. I'm not sure what causality means in the context of grammar
@@NickHuntingtonKlein thnks for reply
Appreciate the video but i was getting distracted by the board games in the background lol