school is amazing me right now. there are only two weeks left, and then i may never take another class in my life.
during my first full-time semester of university (fall 2004), it was difficult for me to see the dignity and importance of each individual class. the marginal homework problem didn't always matter much. but i was filled with such enthusiasm for my resultant education. i looked forward to a degree with multiple majors.
now that it's almost all over, there's nothing to look forward to. i have no career plans whatsoever. i look at the work i have to do over the next 13 days. i probably couldn't care more. these marginal course materials do matter! my knowledge can change a letter grade, my understanding of a concept, the impression i leave my professors with, my letters of recommendation, whether i get more funding for the rest of the year. a lower letter grade can significantly change my graduate gpa, which can result in fewer admissions to future programs or in fewer job offers.
plus i've come to realize how much i really do like economics. it is true that a lot of people screw it up.
statistics really just formalizes what goes on in my mind, like that you need a sufficient sample size, or that you need sufficient variation across all variables with respect to all other variables. let's not forget the gauss-markov theorem.
good economics training will teach you to avoid spurious correlation (which seems prevalent in the media--global warming?) and concepts like price gouging (common political rhetoric for the law of supply).
econometrics is an important addition to statistics. it allows you to deal with naturally occurring experiments, and not just controlled experiments. the use of instrumental variables and other techniques is brilliant.
so i've given up on my crusade against statistics and econometrics. instead, i'll take up arms against people who say things like "accept the hypothesis" instead of "fail to reject the hypothesis." after all, inference doesn't let us know anything. but data that satisfy all of the assumptions can lead us to reject or fail to reject a hypothesis with some degree of confidence. however, we never have perfect confidence for predictions. (or, man is not god.)
i believe that as long as we use the language "fail to reject" instead of "accept," type I errors are much worse than type II errors. let us never overstate our knowledge.
during my first full-time semester of university (fall 2004), it was difficult for me to see the dignity and importance of each individual class. the marginal homework problem didn't always matter much. but i was filled with such enthusiasm for my resultant education. i looked forward to a degree with multiple majors.
now that it's almost all over, there's nothing to look forward to. i have no career plans whatsoever. i look at the work i have to do over the next 13 days. i probably couldn't care more. these marginal course materials do matter! my knowledge can change a letter grade, my understanding of a concept, the impression i leave my professors with, my letters of recommendation, whether i get more funding for the rest of the year. a lower letter grade can significantly change my graduate gpa, which can result in fewer admissions to future programs or in fewer job offers.
plus i've come to realize how much i really do like economics. it is true that a lot of people screw it up.
statistics really just formalizes what goes on in my mind, like that you need a sufficient sample size, or that you need sufficient variation across all variables with respect to all other variables. let's not forget the gauss-markov theorem.
good economics training will teach you to avoid spurious correlation (which seems prevalent in the media--global warming?) and concepts like price gouging (common political rhetoric for the law of supply).
econometrics is an important addition to statistics. it allows you to deal with naturally occurring experiments, and not just controlled experiments. the use of instrumental variables and other techniques is brilliant.
so i've given up on my crusade against statistics and econometrics. instead, i'll take up arms against people who say things like "accept the hypothesis" instead of "fail to reject the hypothesis." after all, inference doesn't let us know anything. but data that satisfy all of the assumptions can lead us to reject or fail to reject a hypothesis with some degree of confidence. however, we never have perfect confidence for predictions. (or, man is not god.)
i believe that as long as we use the language "fail to reject" instead of "accept," type I errors are much worse than type II errors. let us never overstate our knowledge.
instead, i'll take up arms against people who say things like "accept the hypothesis" instead of "fail to reject the hypothesis."
:D
also, "let us never overstate our knowledge" is a serious reflection of how I try to live my life.
It's good to see you full of passion for the things you've learned and are still learning!
"scientists explain that they no longer know things" —edan
:)
"good economics training will teach you to avoid spurious correlation"
in my personal anecdotal world view, other disciplines tend to think that economists live with their heads in the clouds.
in my opinion, very few economic arguments don't begin, "in the perfect world,..." but usually the words "in the perfect world" are omitted while the concept remains. the whole of economic theory is no more than a sequence of thought experiments, perhaps supplemented by real world data, but not necessarily. thus economists don't believe in the possibility of price gouging.
does training as a economist make you any more qualified to make judgments about climate change than laypeople?
"There is very high confidence that the global average net effect of human activities since 1750 has been one of warming, with a radiative forcing of +1.6 [+0.6 to +2.4] W/m2" - Intergovernmental Panel on Climate Change 2007: The physical basis. Summary for policymakers 2007.
in my personal anecdotal world view, other disciplines tend to think that economists live with their heads in the clouds.
in my opinion, very few economic arguments don't begin, "in the perfect world,..." but usually the words "in the perfect world" are omitted while the concept remains. the whole of economic theory is no more than a sequence of thought experiments, perhaps supplemented by real world data, but not necessarily. thus economists don't believe in the possibility of price gouging.
does training as a economist make you any more qualified to make judgments about climate change than laypeople?
"There is very high confidence that the global average net effect of human activities since 1750 has been one of warming, with a radiative forcing of +1.6 [+0.6 to +2.4] W/m2" - Intergovernmental Panel on Climate Change 2007: The physical basis. Summary for policymakers 2007.
> does training as a economist make you any more qualified to make judgments about climate change than laypeople?
sure. if one studied econometrics well, one should know that one needs to look at a regression of human impact on climate, not time on climate data.
most scientific disciplines will teach the students to be more scrupulous and skeptical of assertions.
sure. if one studied econometrics well, one should know that one needs to look at a regression of human impact on climate, not time on climate data.
most scientific disciplines will teach the students to be more scrupulous and skeptical of assertions.
> one should know that one needs to look at a regression of human impact on climate, not time on climate data.
sample size 1?
so unfortunately we can't regress time or human actions on climate change. all we can do is measure how much global warming potential (GWP) a particle has and then measure how many particles are in the atmosphere or are being emitted and figure out what the expected change in the climate would be. assuming that particles will have the same effect on the atmosphere as they have in laboratory experiments is an assertion, but it seems a reasonable one. GWP rates and calculations:
http://www.grida.no/publications/other/ipcc_t … g1/248.htm
however, assuming that global warming is a concern because some researcher somewhere found a correlation between time and average global temperature changes seems a pretty awful assertion.
sample size 1?
so unfortunately we can't regress time or human actions on climate change. all we can do is measure how much global warming potential (GWP) a particle has and then measure how many particles are in the atmosphere or are being emitted and figure out what the expected change in the climate would be. assuming that particles will have the same effect on the atmosphere as they have in laboratory experiments is an assertion, but it seems a reasonable one. GWP rates and calculations:
http://www.grida.no/publications/other/ipcc_t … g1/248.htm
however, assuming that global warming is a concern because some researcher somewhere found a correlation between time and average global temperature changes seems a pretty awful assertion.
ehh sorry, if i was overly offensive. just, you know i have spent a lot of energy doing research much of which would be meaningless, if global warming was just the conjectures of a short-sighted statistician. and as for the economics stuff, my hairs start to rise when economists start saying things are this or that way. but you know, you are the last person i would direct such criticism at.
> sample size 1?
what's a sample size of one? i'm not making any inferences, so i don't need a sample. i'm just making an argument for skepticism of potentially spurious inference.
> assuming that global warming is a concern because some researcher somewhere found a correlation between time and average global temperature changes seems a pretty awful assertion.
i never said anything like "the entire foundation of global warming rests on spurious correlation." i just said that the arguments seem prevalent to me--they do (again, to me). i often hear climate impacts related to time, then time related to industrial impacts. but i certainly have not evaluated the proper academic literature, and i'm not claiming that i have.
what's a sample size of one? i'm not making any inferences, so i don't need a sample. i'm just making an argument for skepticism of potentially spurious inference.
> assuming that global warming is a concern because some researcher somewhere found a correlation between time and average global temperature changes seems a pretty awful assertion.
i never said anything like "the entire foundation of global warming rests on spurious correlation." i just said that the arguments seem prevalent to me--they do (again, to me). i often hear climate impacts related to time, then time related to industrial impacts. but i certainly have not evaluated the proper academic literature, and i'm not claiming that i have.