22 ms·
Inaccurate. This is a overused misnomer. Correlation is a necessary precursor to causation, but it's existence does not guarantee causation. If correlation is
by deltasixeight 5y ago
Inaccurate.
This is a overused misnomer. Correlation is a necessary precursor to causation, but it's existence does not guarantee causation.
If correlation is established it doesn't automatically establish causation as well, but it functions as evidence for two possibilities:
The possibility of causation and the possibility of a shared causative source.
It is wrong to say they are not equal. They are equal in a really fuzzy way.
- AHappyCamper 5y agoActually, it's totally accurate. And I respectfully disagree with your opinion ;)
- deltasixeight 5y agoAnd respectfully. You're wrong. Utterly and completely. This is beyond opinion. This is logic. I urge you to think about it.
- lvass 5y agoCorrelation !== Causation // true Correlation != Causation // false So Javascript really was right about having both operators. /s
- bagacrap 5y agoI think you could say correlation and causation are correlated, but I don't think it's fair to repurpose the word "equal" to mean something it doesn't.
- deltasixeight 5y agoIf two things have a causative connection they ALWAYS have a correlation as well. If two things have a correlative connection they don't always have a causative connection as well. I never repurposed the word equal. But to say they are not equal is highly, highly inaccurate. Correlation does not imply causation but causation DOES imply correlation. Applying the word not equal to this is not accurate at ALL. Seems like pedant-ism but it's not. Trust me. Correlation is evidence for causation. But people always throw this term around as if finding a correlation between two things is utterly useless in determining causation. It's like me presenting a witness who saw a murder occur. Technically speaking, the witness is evidence for murder, not proof of murder. This fact is obvious to everyone. It's completely useless to say something obvious along the lines of: TESTIMONY != PROOF The testimony is strong evidence for murder, as correlation is strong evidence for causation. Seriously, someone just posts one line CORRELATION != CAUSATION and everyone just gathers around this misnomer as if it renders the correlation utterly useless and pointless. No. The correlation is a TESTIMONY to a murder. It is powerful evidence. Do not dismiss it based off some stupid statistical catch phrase.
- dragonwriter 5y ago> If two things have a causative connection they ALWAYS have a correlation as well. Only “ALWAYS” when you completely control for all other influences. It's possible for a causal relation to not have a correlation otherwise.
- deltasixeight 5y agoYeah you're right, but you do get my point. I should write, causation ALWAYS IMPLIES correlation.