correlation v. causation
- andy: "During the 1990s, an average college graduate earned in excess of a 70% premium above the average high school graduate, e.g., Pierce and Welch (1996)."
- Zia: I'm listening to the song through my bose earbuds
- andy: why do you think the statemtn above is true
- Zia: sexual favors
- Zia: we all know high schoolers suck at giving head
- andy: hmm
- andy: but we all were high schoolers
- Zia: I mean, many high paying jobs require a college degree
- andy: that assumes maybe that college kids are more ambitious than just high school grads
- Zia: no, it assumes that those are requirements for high paying jobs
- andy: yeah, but im going back to the original statement
- Zia: you're saying that from that statement
- Zia: you can determine that high school grads are less ambitious?
- andy: maybe high school grads didnt care to make as much money as college grads
- andy: its like, in america, we assume everyone wants to make a lot of money
- andy: and the statements like the one i showed you kind of try and isolate one factor
- Zia: the statement above is a statement above...
- andy: how so?
- andy: i mean, i see what you mean, but i dont really see that
- Zia: i mean, you gave me a statement
- Zia: which is a correlation
- Zia: and you're finding causation
- Zia: you do this a lot
- andy: i was arguing that the statement above was basing its impact on lots of factors, many that are assumed and not controlled for
- Zia: the statement above was basing its impact
- Zia: what does that even mean ^
- Zia: all i see is a statement
- andy: impact = legitimacy
- Zia: still, it would have to be an author who was basing the statement's legitimacy on factors, and it would have to be in the context of those factors
- Zia: an author or an analyst of some kind
- andy: all im saying is that maybe there's no meaning in the statement above
- andy: what he's trying to do is relate college education to salary as related to those who just have a hs degree
- Zia: there is no meaning in the statement above, except what's written
- Zia: if you've clipped this from an article, that would change things a bit
- Zia: if it's just a quote
- andy: maybe those who go to college are just more ambitious and they just care about making more money?
- Zia: it means exactly what it says
- andy: im not saying its not true
- Zia: I know that
- andy: but the context of the statement assumes lots of things
- Zia: you haven't given me any context whatsoever
- Zia: so I can't agree/disagree there
- andy: hmm
- andy: i dont think this is coming through
- Zia: all you gave me was a statement - if you want to discuss the reasons why we believe it might be true/false, then we can do that... what we can't do is talk about the author's assumptions or prejudices since all it is an allegedely factual statement
- andy: yeah, all im talking about is how that statement rings with our understanding of america
- andy: how other people understand that statement
- Zia: how do other people understand it
- Zia: still confused
- Zia: I guess I'm trying to figure out what your point is
- andy: ok
- andy: i feel like they're trying to isolate the causation based on the single notion of whether or not they went to college
- andy: but im asking, what if it starts before
- Zia: understand that there's no claim of causation in that statement at all, but we'll move to your question
- Zia: what if what starts before?
- andy: the causation
- andy: but i found some context
- andy: "High returns to additional schooling are noted. In the US these returns have grown dramatically over the past 20 years, particularly for a college education...."
- andy: so that author is using the statement above as evidence for this assertion
- Zia: causation can't "start", I have no idea what you're referring to...
- Zia: okay, no causation still
- andy: do you not think that he's trying to create a causation between level of schooling and salary?
- Zia: it's whether I think it or not, he just isn't
- Zia: between those two statements, all that exists are correlations
- andy: ok i agree with you
- andy: i feel that correlations are all that exist
- andy: but he's definitly trying to show a causation
- andy: you think he's only trying to show correlation?
- Zia: he's not trying to show anything
- Zia: he IS showing a correlation
- Zia: that is it
- Zia: period
- andy: see, to me, he's trying to show causation where there's only correlation
- Zia: try to explain to me how there's a claim for causation
- andy: he's arguing that high wages are caused by college education
- andy: no?
- Zia: no
- Zia: he's saying that the data shows a correlation between high wages and a college education
- Zia: not that it's the cause
- andy: high returns to additional schooling are noted
- andy: additional schooling creates high returns
- Zia: you wrote that last statement
- Zia: data shows that there are more ice cream vendors in the summer months than in the colder months... therefore, it's the ice cream vendors that cause warm weather
- Zia: now, if you showed me another part of the article where he makes some kind of a case that there is causation, I wouldn't be shocked
- Zia: but you haven't
- andy: he says that high return are associated with additioning schooling only
- andy: he's not including any other factors
- andy: he's implicitly arguing for causation
- Zia: you can't implicitly argue for causation
- andy: ok
- andy: it would seem rather pointless otherwise
- Zia: correlations are simply indications that there's a pattern, that's useful, but if you're trying to prove a point, it's not
- andy: the way i see it...he's creating a relationship between two variables in order to prove a point; to me, that means he's creating causation
- Zia: to statistics, it means he's creating a correlation
- andy: how so?
- Zia: correlation: the relationship between variables.
- Zia: that's how so
- Zia: there's no point he's trying to prove, not in those two statements
- Zia: it's like if an alien came to earth on a rainy day and saw everyone with umbrellas, and the next day was sunny... he might say "hmm... studies show that when the umbrellas were out, it was raining, and when the umbrellas were away, it was not raining"
- Zia: there's no argument there
- Zia: if he said "well, obviously, it's the umbrellas that caused the rain since the rain wasn't there when no one had umbrellas" that's an argument
