LogFAQs > #896578463

@Selenara posted...

No matter how many paragraphs you write, you STILL do not understand the phrase "correlation does not equal causation."

Given what I do for a living, I highly doubt that.

Selenara posted...

Correlation means that we know two factors have occurred at the same time. You seem to think this term means causation, but it does not. Causation refers to the factor that made the two factors occur.

Close but not quite.

Correlation means that two properties - A and B - share a statistically significant relationship. Where one is found to be elevated, you can expect a proportional increase or decrease in the other. Note that it is entirely possible - through coincidence - to have two independent (and therefore uncorrelated) variables occur at the same time, usually through small sample size errors (for instance, if I go outside and see a bird land in a tree, that does not mean that me leaving my front door is correlated to a bird landing in a tree).

Causation is identifying the dependent/independent variables in the A/B relationship and determining if one causes the other. It is decidedly not "the factor that made the other two factors occur". That's the entire point of the "Correlation does not equal causation" phrase - pointing out that a correlation between two (or more) properties does not prove that one causes the other.

In brief:
Correlation: Where A, then B
Causation: A causes B
Correlation != Causation: A may cause B or B may cause A or C may cause A and B.

Or, to apply it to the topic at hand:
Correlation: Where gun laws have been tightened, gun crime has dropped (which... I can't honestly tell if you agree with or not at this point).
Causation: Gun laws cause decreases in gun crime
Correlation != Causation: Gun laws may decrease gun crime, or decreases in gun crime may prompt stiffer gun laws, or (unknown third factor - you still haven't said what you think this is) causes both tighter gun laws and decreased gun crime.

Selenara posted...

You are saying that decreasing and increasing is "stagnant"? That is preposterous. The numbers fluctuated before settling in a downward trend, and you are trying to spin it to fit your assumptions. Perhaps you should set aside your assumptions and look at the data before you form a conclusion.

I've done that, you're just not listening to what I have to say, no matter how many ways I try and explain it.

Since work is slow at the moment, I went ahead and plugged the numbers into excel and got the actual equations for the trend lines.

1981 - 1991: Trendline is defined as Y = -0.0103X + 21.216
1992 - 2002: Trendline is defined as Y = -0.272X + 54.906

In other words, the average rate of decline nearly tripled after the gun law than before it.

If you still disagree with my assertions, consider that the trend from 1986 to 1991 was actually positive - the trendline for those years is Y = 0.434X - 85.601. In other words, for the six years immediately prior to the gun legislation, gun killings were trending upwards, which immediately reversed upon the law's passage.

And lest you accuse me of cherry-picking my data, there's no five-plus year period after the gun law implementation that had a positive trend. The closest you get is if you clip out 1997 to 2001, which still has a negative trend but one that's very nearly flat (Y = -0.008X + 16.558).
---

Kill 1 man: You are a murderer. Kill 10 men: You are a monster.
Kill 100 men: You are a hero. Kill 10,000 men, you are a conqueror!