gilbertrollins

Dup budup - let the profs be profs

That's incredible. One course? What are a normal TA's responsibilities outside grading papers?

I'm just going to have to get smart on what the measurement issues are in the nascent empirical fields I'm looking to break into, compare those with syllabi from the different courses, and discuss all of that with the department, yeah.

^Maybe a general courtesy. Maybe a good sign. It won't hurt you to take it as a good sign.

"and lets say you missed something completely basic in linear regression that wasn't covered in the econ department.. and showed up at another campus for a job talk and everyone was like "omg, so and so doesn't know OLS regression"" That's impossible. Econometrics courses proceed at the same level of mathematical rigor and generality as graduate statistics courses in the stats department. There are issues specific to sociology, like regressing survey data, multi-level modeling, and no doubt several estimation corrections in sociology that have been developed by sociometric theorists that I will need to understand, that I would need to and want to learn independently. Understanding statistics in a comprehensive manner -- that is understanding the theory behind it rather than just how to read ANOVAs coming out of Stata -- is precisely how one prepares to learn and apply statistical techniques in new ways, which is necessary for novel research. " Departments generally prefer that student take classes in the department unless it's something that the department doesn't offer and is something that would enhance a students sociological work." I explained earlier that I was to get a better foundation in statistical theory than an applied course will offer, precisely because I want to measure data for which there are few standard methods yet developed -- i.e. content and corpus analysis, and network analysis. Those are sociological all the way down. I have no interest in "doing economics from the sociology department." It would be helpful if someone could link to a syllabus or two from graduate sociometrics courses.

Any of the people I know whom are actually prepared for T10 econ programs have taken several semesters of analysis, calculus of variations, and usually a handful of graduate micro theory and metrics courses already. It's turned into a situation much like that in physics, engineering, or mathematics, where if you don't start tracking toward graduate degrees by your Junior year (if not sooner) -- you're going to have a hard time cracking T20. The applied stats courses I'm taking in the stats department aren't proof heavy. In fact the one I took two semesters ago only did a little bit of verifying some equalities of expectation, derived a couple estimators using calculus (taking partials and equating to zero -- tough stuff!). The one next semester gets to multivariate regression (which I've covered in an applied setting in econometrics, but in a really crappy way -- like "look at the Excel plot for heteroskedasticity and serial correlation" stuff) but I don't think we even do method of moments. This is sort of a T20 standard econometrics text: http://www.amazon.com/Econometric-Analysis-7th-William-Greene/dp/0131395386 Pretty sure most programs don't get to look at a bunch of applied papers until their field courses (development, labor, industrial organization, environmental, etc) in year two. And I don't think there's a lot of hand holding on how to use the packages -- these kids will be using MatLab and Mathematica in theory courses.

I don't know what I'm getting myself into, really. I left economics precisely because I want to do a good deal of theory and do not like writing deductive proofs. Some of that is based in the lack of motivation to learn various theories of optimization and games -- as these are done to death and for a variety of reasons not particularly useful to questions i want to ask. Some of it is I think based in the usual math phobia and a lack of taste for it. On the other hand, complex adaptive systems pose I think the most compelling compromise between agency and structure to hit social science in decades, network theory is incredible and advancing, and like you said NLP is becoming a viable way to measure culture directly. Do I want to write theorems about semidirected graphs, and code Python all day? Hardly. Maybe it'll change when I get further away from econ, where the mathematics in a lot of ways feels like a rat race and signaling game, more than a collection of "tools" as it's presented. Frankly a lot of people applying comp sci to natural language processing and text analysis, building agent based computational simulations, proving propositions about graphs, and so on are just fascinated with the whiz-bang technicality of the exercise itself. I don't get those kicks, personally. And that's not to knock it at all. I'm friends with a lot of these guys and gals, and we need people like them. I'm thinking long run my comparative advantage will land somewhere in qual theory in combo with applied networks and stats to do empirical work, but not sure yet. I've been waiting for two years to wake up and love math and deductive logic. Hasn't happened yet.

A good example would be Fligstien's use of the term "isomorphism." Based on the Greek etymology, he's in the right ballpark. Based on its most wide use elsewhere in academia -- mathematics (and it is quite a common concept here) -- he's being quite confusing. An isomorphism in mathematics is a function that preserves the structure of an ordered field -- that is, all of the usual multiplication, addition, division, and ordering axioms continue to apply. So for instance, a function that maps operations on matrices to many operations in calculus is often isomorphic. One can think of these as the "mechanics" or the "joints" of a social structure, and claim that when one social structure is isomorphic to another, they have identical structure. And one can metaphorically relate "linear algebra" and "calculus" to two different social structures. But an isomorphism in mathematics is not merely "something that looks different than something else, but has the same basic underlying mechanics." It would be much clearer to say "organizations are operating more and more the same way, e.g. little leagues for XYZ reasons now mimic the modern business corporation" than it is to say "organizations are increasingly isomorphic to one another." On the other extreme, definitional constance via mathematical formalism as in economics can be and often is equally as detrimental. But I do think greater definitional consensus is in order. (disclaimer: I read Fligstein's paper recently and have a small sample I'm working from -- not trying to pick on him -- he's asking invigorating and important questions)

Everything you listed is covered in most undergraduate intermediate econometrics sequences. (Removed at Users Request) @SOCgrad987 An econometrics course has very little to do with economic theory. Well i shouldn't say that. They are assuming basic stuff about production functions and supply and demand, etc. But econometrics is basically a slightly-tweaked version of applied stats from the stat department itself. From my little wikipedia skimming of multi-level models, these look like techniques to control for endogeneity among RHS variables - which is basically what instrument variables are for (popular in econ) and a number of other corrective estimators. @Splitends I've been reading a ton of soc theory and learning a ton. I didn't dismiss theory. I dismissed overt verbiage and neologisms as unnecessarily obfuscatory and prestige-hunting. This is a common complaint throughout academia. And I related it to the common complaint about formalism in economics. I don't understand why theorists get a prima facie pass to use hyper technical language and derive entire systems of thought that require enormous work of their reader, when theory can be made clear and accessible, and moreover testable. This was C. Wright Mill's eviscerating complaint in Sociological Imagination, no?

Oh. Yeah I'm getting the impression the sequence is structured really differently than in economics. Your job in economics is to make it through first year courses (micro, macro, metrics, and more general math for economists) and pass comps. If you don't pass - you either stay without funding and pass, or you leave. Then you take field courses second year, and have to pass comps for those. Then they give you the MA. Then you start working with an adviser. Sounds like in soc you start from day one working really independently, and forming your project. I like that a lot.

I've done 20+ hours of independent study slash research assistance, and the As there are pretty much guaranteed, so I'm actually already struggling with the sort of "third year woes," like how to stay productive and manage myself independently, how to battle my own insecurity about formalizing my ideas without someone right there to cheer me on, etc. Sociological theory only makes me want to cry because these guys refuse to get their definitions straight, love neologisms, etc. I love the theory itself. I do not, alternatively, like writing deductive proofs. Since leaving econ my first thought when I read any theory is "how do we test this," not "ooo -- that sounds intuitively correct." Frankly I think verbiage (sociology) and mathematics (economics) end up serving the same masters: obfuscation and window dressing. I'm taking an applied stats course and probability one this next semester, so maybe I can test out of the soc stats course? Will that be rude to ask? I don't want to do sociometric theory itself, just get really under the hood. So maybe I could take the econometrics sequence? They're concerned about a lot of the same things; I'll just have to interpret it differently. I don't think I'm going to be doing strat, or surveys, at least while my tenure clock is ticking (maybe later), so I could maybe skip the multilevel modeling stuff? I don't want to give anyone the impression like, "So I hear your stats sequence is a real joke -- Ima go over to econ thanks." I'm genuinely interested in learning how to test sociometric problems like cascades and other structural/interdependency issues (especially with language data). I was thinking it'd be a good idea to have my classics read before I start, like Durkheim, Parsons, Weber, etc etc. I really like the symbolic interactionists. Has their impact faded some?

What about the derivations of estimators and corrections? The theory behind everything? I want to get under the hood, and am wondering if taking the first year econometrics sequence at my prospective school will be dramatically more rigorous than the sociometrics sequence or not.

Are there qualifying exams at your programs that cover your core material?

Don't you think most students would need a bridge to digest Gintis' little document? He's very formal. And very few soc students will write formal proofs, ever, no? The section on probability is probably useful preparation for the metrics sequence, but again someone who's only had a few semesters of calculus, if that, is going to have quite a bit of trouble reading Gintis. He's going for mathematical elegance as against hand-holding. I thought his title was laughable (and that's coming from someone who loves most of what Gintis does).

It's strange. I listen to Cato Institute podcasts when I'm taking breaks from more serious work (which is too often), but I still listen to Earth Crisis and Dead Prez. Every time I listen to reggae I think "all this stuff about burning Babylon down is the politics that keeps the Caribbean poor," but I still love it. (Removed at Users Request)

So, I've heard mixed reports on here about the difficulty of first-year courses. There is a syndrome (not permanent) called Law Student Blindness, where people literally go blind from reading too much and not sleeping. Economics programs are famous for their first year courses being soul-crushing hard. And I presumed this was the case for any graduate program -- thousands of pages of reading per week -- maybe a couple few hundred pages of writing by the of the semester. I was looking forward to that. Can anyone elaborate on the difficulty of their first year courses, and whether they felt challenged. I was really looking forward to the boot-camp aspect, breaking you down and building you back up again with knee-jerk training in how to pull the trigger on your ideas no matter how terrified you are.

I've never read any other Prickly Paradigm other than Secret Sins, and I do think it is vital that scholars feel free to make some mistakes publicly, some unsupported conjecture, and vent about the conditions of their discipline. Graeber is a wicked smart guy, though. He's not naive about the bad name anthropology has gotten for Noble Savage myths, for doing straight up political advocacy, and for carrying the banner of the French Structuralists even more than the English department at times. His solution to those problems? Give those people the middle finger, by just ignoring those complaints. Well you can do that if it makes you feel like a revolutionary and reifies your sense of intellectual integrity, but it's not a particularly persuasive way to confront your interlocutors. I agree with herbertmarcuse and Graeber in that social science ought to be relevant, and that we should watch out for Arm Chair Ism. Understanding history of thought, though, is incredibly important to any Science that takes for granted that frameworks of thought compel a great deal of action. And I did not find Graeber's blurb on anarchist scholarship being about action and politics in process (how to organize a direct action meeting), versus Marxism being all talk and no walk, very compelling. I have not been convinced (obviously) that his, Mauss', and other anthropological work doesn't constitute hunt-and-peck Marxist scholarship, in the same way that a majority of 20th century history was done. A major difference between McCloskey's essay and Graeber's, is that McCloskey says "Here are the reasons most people think economics is bullshit -- and the economists are right that those are bullshit reasons to think economics is bullshit. Here's the real reason you should be skeptical." Graeber says, "Here are the reasons most people think Anthropology is bullshit. Fuck you; I'm a revolutionary, better even than all the other scholastic revolutionaries, and here is more Anthropological opining in the traditional mode to prove it."

^I was picturing skinny white kid with stretched ears, a NOFX t-shirt, and a denim jacket. God bless stereotypes.

I find Graeber's style disturbing and at the same time really interesting. Here's a guy who unashamedly thinks social science ought to serve and form a political agenda, without a single nod to positivism. And that, to me, is scary. But at the same time, he attempts to infiltrate an institution built largely on the pretenses of positivism, in order to preach theoretically about how anarchism is a-theoretical. I've got to read more of this guy. He says: "In the end, though, Marcel Mauss has probably had more influence on anarchists than all the other ones combined. This is because he was interested in alternative moralities, which opened the way to thinking that societies without states and markets were the way they were because they actively wished to live that way." As I understand it, Noble Savage sentiments emerged from the origins of social science in the 18th century. These guys were fascinated with the discovery of North and South American Indians, and the argument was largely over whether or not these "original state of nature" people actually represented a Hobbesian "warre of all against all" or not. The first anthropologists didn't need a ton of encouragement to take 18th and 19th century socialism and go paint it on their ethnography (and to their credit, calling Savages noble at the time was a radical, contrarian, and insightful question rather than received dogma as it is today). Sahlins was the Original Affluent Society guy who really nailed it home. I think it's incredible that he doesn't discuss market anarchism in fifty pages of opining about anarchism - that just doesn't seem scholarly. Yet he's totally content to tell these stories about how indigenous people "consciously rejected most of what we would consider the basic principles of economics" without verifying it on any empirical criterion other than his second-hand reading of cherry-picked ethnographies. I mean, I could write a nice long essay about how most of the interactions within a modern firm look like "gift economies," and that these are "not based on calculation, but on a refusal to calculate" because of the incredible cooperation, sacrifice, reciprocal giving, and so on that takes place. But I wouldn't have done any science at all unless I backed up statements like: "They would have found the very premise that the point of an economic transaction . . . was to seek the greatest profit deeply offensive," with at the very least a question put to the workers in the firm "do you find X offensive?" I have a hard time believing anyone here is as offended as Graeber. Edit: and I read more of the essay. Graeber's reading of reciprocal giving is just incredible. I mean, i learned in a 100-level cultural anthropology course I took at a community college that reciprocal giving is often propagated between groups by ceremonial Big Men. Incredibly Graeber still continues to claim that these systems necessarily delimit and undermine hierarchy and power. Per capita death rates due to extraordinary inter-group violence are enormous among most indigenous populations, and are verified in the archeological record. That, to me, does not strike me as an anarchic world blissfully and purposefully removed from power and "symbolic violence."

^Great stuff. I find Marxism extremely obnoxious, but I try to be scholarly about it and understand what I mean to argue against. In any event, that book looks great. These guys seem to be recognizing that you can't add up "independent observations" of opinions to derive public opinion any more than you can (in economics) add up "independent utilities," to derive social welfare. Fundamentally people's opinions are recursive functions of one another, subject to the usual knock-on cascading/network/herding effects. So there is indeed a great deal of endogeneity surveyors need to account for in understanding macrosocial mores. Great stuff. Will read thanks.

Rent controls have a documented history of increasing non-pecuniary discrimination, and housing projects were by most measures no less a manifestation of institutional racism and the corralling of the poor like cattle, than the war on drugs is. If you want to help the poor, write them checks, and tie the gifts to their wages by matching income privately earned. That incentivizes work, and lets poor people, whom are smart enough to figure out how to bring themselves out of poverty, decide what to spend their money on. Housing, food, and other directed subsidies in combination with the war on drugs do nothing but exacerbate the degradation of property and the economic climate in poor neighborhoods. The solution to the problems of the developed world's poor is not even more government paternalism and fumbling around, trying to litigate the ethics and decisions of these people. Gentrification, that is, improving property, is a good thing. The neighborhood I live in was recently ranked the most dangerous neighborhood in the country. Fifty years ago it was an affluent Jewish neighborhood. Neighborhoods change rapidly in cities. It's not a simple matter of affluent neighborhoods spreading-from-center, pushing minorities further and further toward the edges of the city and eventually off the edge of the world. Anti-gentrification sentiments inspire protectionist economic policies -- or rather I should say local alderman in my neighborhood not letting Arabs and Whites open businesses because doing so "keeps money in the Puerto Rican community." No -- preventing people from trading one another, like say whites or arabs setting up businesses and offering Puerto Rican consumers competitively priced goods so that they can invest those dollars saved into more productive margins like, say, their children's education -- is how you make everyone rich, Puerto Ricans and Arabs and Whites alike.

Pretty sure only marginal candidates get interviewed, and mostly to track them towards a reject. My man got interviewed for a barely T50 economics program last year, and got in. He reports that he's at the bottom of his cohort, and is actually paying for his degree (he doesn't mind - he'll end up in private industry paying off loans no probs).

Ballin' and shot callin'.

Money is all that matters, always.