Ummm actually, yes it does.
The saying you're looking for is "correlation doesn't nesscary mean causation"
It most definitely implies it though.
Are you a statistician or having you every run an experiment where you had to prove the data you collected was statically significant?
Correlation does not imply causation is literally out of the text book.
-So you're talking statistical terms now? Yes, then that phrase appears, as a quip, when mainly discussing regression analysis. But it is to drive home the point that regressions are are at their core are mathmatical operations. Yes they try to shed some light on real world phenomena, but all regression should be take with a grain of salt and it must be remembered that they are showing correlations not A results in casuing B. The is no single regression that economist, statisticians or scientist present as proof that one thing causes the other
For example if say I were looking at the between ice cream sales per week and murders per week for an urban city. And I made the number of murder the dependent variable and ice sales the sole explanatory variable. I hit run in STATA, R, or whatever and I will probably get a regression with a high R, high R-squared, also the coefficient on murders would probably be statistically significant.
The regression is telling me the two variables are highly correlated, imply/suggesting that ice cream sales and murders are linked, and someone to falsely conclude that eating ice cream makes people want to commit murder. However, I know that they are probably (really definitely) not directly linked, or that one is dependent on the other. It is the weather that is indirectly linking them.
A similar thing actually happen with Polio and ice cream, because both ice cream sales and polio infections spike during warmer months, some scientist thought ice cream consumption was correlated with Polio .
I actually work as a Economic Analyst/Econometrician, so I know how dangerous it can be blindly trusting the result of a regression.
-Looking at the comment not through the lenses of statistics, really the key work here is "implies", which really means it suggest it. Yes, the suggestion might be wrong, but if two things are happening at the same time (which would look like it has a linear correlation on a scatterplot) that implies there must be a link. With America, you have countries with lots of guns not seeing the same things happen. If guns don't kill people, people kill people, then it is a fair question to ask what is wrong with our people. And here it seems that "mental health" is pushed to not have that conversation about America's relationship with guns, as a country.
In science, all kinds, that is why peer review, replication studies, and alternative studies are so important. To strengthen the argument towards causation or to provide more context that weakens the previously suggested correlation.
I can tear the black culture argument apart in seconds. The arguments dealing with America's gun violence are much more nuanced.
So maybe people should be open to more studies into gun violence (which the NRA is against, of course), and not dismiss people on the other side of the argument as sheep.
