Quant. Sociologists - please recommend readings

[realmadrid7]

Hi guys,

I'm entering sociology from a different undergrad major (econ.) and I was wondering if there is some contemporary work (in quantitative sociology preferably) that you could recommend to me to help me get a better sense of the field. I figure many of you will be more knowledgeable than me at this point about the research happening in sociology right now, so I would trust your recommendations over random google searches. (I am not as interested in the topics as getting a feel for the methods used) Thanks!

[TheBoxInspector]

I'm not sure about how much I would be able to tell you about quantitative stuff, but for economic sociology you might be interested in the following:

An Engine, Not a Camera: How Financial Models Shape Markets by Donald Mackenzie--constructivism.

Getting a Job by Mark Granovetter (including the appendix on "Economic Action and Social Structure")--the problem of embeddedness.

Structural Holes: The Social Structure of Competition by Ron Burt--network analysis.

The Architecture of Markets by Neil Fligstein--field theory.

[thingsandorders]

Yes, I also would like to have some offers on what to read on quantitative methods.

[Ladril]

Hi guys,

I'm entering sociology from a different undergrad major (econ.) and I was wondering if there is some contemporary work (in quantitative sociology preferably) that you could recommend to me to help me get a better sense of the field. I figure many of you will be more knowledgeable than me at this point about the research happening in sociology right now, so I would trust your recommendations over random google searches. (I am not as interested in the topics as getting a feel for the methods used) Thanks!

Sociology mostly uses multivariate statistical models and quantitative methods for social network analysis. However nearly all areas of mathematics have been found to have sociological applications (set theory, boolean algebra, differential equations, to name a few). I'm not sure there is an area that can be called "quantitative sociology" per se. Most subspecialties make use of qualitative and quantitative methods.

I assume you're mostly looking for a selection of texts which may allow you to gain a feeling for the most commonly used methods in the discipline. I would suggest you take a look at the SAGE series "Quantitative Applications for the Social Sciences". Complete list here:

http://www.sagepub.com/productSearch.nav?seriesId=Series486

You should also take a look at the work of professor Charles Ragin here: http://www.u.arizona.edu/~cragin/cragin/publications.shtml. He has written a lot on innovative applications of mathematics to social research.

[Ladril]

Further detail: there is a subfield called "mathematical sociology" which at this time focuses mostly on agent-based simulations of social behaviour. Obviously it involves a lot of math. One of the most accessible texts I've found on the subject is:

http://books.google.com.mx/books?id=TzeGLSUheOoC&dq=billari+simulations&source=gbs_navlinks_s

[quantitative]

I'd just suggest really learning the various models/different statistical coding languages. By really understanding what components that work together in a model from a statistical standpoint, you can gain a lot of insight into why certain models are used in certain situations and be able to interpret the results from academic papers (using models you know). For me, that was a very good jumping off point. There's a lot of quant stuff going on in sociology now. Social network analysis is the new big thing, it seems. Personally, I'm very excited about event-history modeling since it allows you to really utilize longitudinal studies. Obviously, that is more on the graduate level, but I'd just suggest immersing yourself in understanding/being able to really interpret the results that you see. Learning SAS/SPSS/Stata/R is a valuable tool as well.

Edited April 26, 2012 by quantitative

[realmadrid7]

Thanks guys, this is exactly what I was looking for.

[smokeypup]

If you want a good book on stata, which seems to be the norm in top departments, I would recommend this book:

http://www.stata.com/bookstore/regression-models-categorical-dependent-variables/

It gives you the logic of the models and the stata commands for categorical models, post estimation and presentation of data.

[realmadrid7]

If you want a good book on stata, which seems to be the norm in top departments, I would recommend this book:

http://www.stata.com...dent-variables/

It gives you the logic of the models and the stata commands for categorical models, post estimation and presentation of data.

Thanks! I'm reasonably familiar with how STATA works, since I've used it quite a bit in economics. What I'm most interested in is, I suppose, examples of sociology applications of the kind of regressions I'm familiar with in empirical economic analyses. (this link looks like a useful manual to have on hand though)

Edited April 26, 2012 by realmadrid7

[Ladril]

Thanks! I'm reasonably familiar with how STATA works, since I've used it quite a bit in economics. What I'm most interested in is, I suppose, examples of sociology applications of the kind of regressions I'm familiar with in empirical economic analyses. (this link looks like a useful manual to have on hand though)

The main difference is that much of econometrics deals with continuous dependent variables. In sociology, categorical dependent variables are more common. Also, some types of models - such as instrument variable models - are more common in economics, while others - such as structural equations models - are more common in sociology. It's a combination of the types of questions asked in each discipline and also simple preferences within specific scientific communities.

[FertMigMort]

The main difference is that much of econometrics deals with continuous dependent variables. In sociology, categorical dependent variables are more common. Also, some types of models - such as instrument variable models - are more common in economics, while others - such as structural equations models - are more common in sociology. It's a combination of the types of questions asked in each discipline and also simple preferences within specific scientific communities.

I would also add that it really depends on what sub-field you're in. We have several people who use instrumental variables almost exclusively as well as people who only use SEM. It's more related to their sub-field in sociology than when they got a degree or anything. What general area of sociology are you interested in realmadrid7?

[realmadrid7]

I would also add that it really depends on what sub-field you're in. We have several people who use instrumental variables almost exclusively as well as people who only use SEM. It's more related to their sub-field in sociology than when they got a degree or anything. What general area of sociology are you interested in realmadrid7?

Mostly urban soc. and stratification (especially in developing countries like India/Brazil)... though who knows how these interests might change in my first year as a Sociology grad student!

[jacib]

A lot of sociology uses pretty standard econometrics methodology, especially a lot of the more vanilla strat stuff... but some of it does cool, random stuff.

Networks is the first thing that is probably outside of the normal econometrics world. "The Strength of Weak Ties" and "Finding a Job" are always favorites, but there are also cool networks stuff deal with other things (Padgett's work on the Medici for instance). There are lots of articles that are lit reviews of networks stuff, so find one of those.

Harrison White does crazy things, his piece "Where Do Markets Come From" is dense with math, but great. His concepts of vacancy chains is also really cool. Everything that he's written that I understand is cool (there's a lot he's written that I can't understand. Take a look at Andrew Abbott's review of his book Identity and Control--he basically says you have to read it like poetry, not social theory. Weirdness). Ivan Chase has done some really interesting sociology with some of Harrison White's concepts <i>using animals</i>, the best of which is vacancy chains in hermit crabs and where hierarchies come from (pecking orders) in chickens. Ivan Chase's work is cool in part because it's so weird.

Andy Abbott's work about, oh G-d, I'm blanking on what it's called, but his form of sequence analysis is cool. His work is always worth checking out. Andrew Abbott is very well loved by all the quant people I love (Harrison White, then Andy Abbott). Katherine Stovel is also really cool, she's done interesting things with Abbott's sequence stuff, like using it to analyze lynchings in "Local sequential patterns: The structure of lynching in the deep south, 1882–1930".

Her adviser and frequent coauthor Peter Bearman is also really cool. He's written some of my favorite quant articles. They're famous for ADDHEALTH which is a huge database and a lot of the stuff came out of that, like "Chains of Affection". But the coolest are these little one-off, incredibly creative articles like "Desertion as Localism" and "Becoming a Nazi".

Neighborhood effects research is also big in sociology. The big name associated with that is Robert Samspon (in addition to the empirical articles he has, he has a cool lit review in the Annual Review of Sociology).

You should read the Culturomics folks, too, though they're not sociologists.

I love stuff that uses instrumental variables, but a lot of the coolest work with that is done by economists. You're probably as familiar with it as I am. With slightly different wording and the exact same results and tables, you could call a lot of the stuff looking at institutions (Acemoglu and his students are the ones I've read the most of) "sociology" instead of "economics".

Chuck Ragin, mentioned above, has his weird "Fuzzy Set" sociology and might be worth checking out.

Last thing I'd add is mobility tables that(usually) look at intergenerational mobility. I don't know if that's taught to econ undergrads. It's not done that much any more in sociology, but was really important a while ago, and it's useful to see it in case you ever are presented with data that can be analyzed that way.

Arizona makes public the readings lists for their qualifying exams. You can see what articles they see are key for stratification and especially statistics and methods.

Especially if this is just your summer reading, rather than looking at handbooks, find interesting social scientists and then look at their weirdest sounding articles. 1) it'll teach you the relative freedom you now have in sociology 2) often times they're using weird data because they want to test out some weird method they've never used before 3) a lot of this stuff is just cool. Now is your chance to see what's out there. I know even just learning about a lot of these weird methods has let me ask questions I couldn't have even asked two years ago.

[jacib]

Optimal matching. Andy Abbott's form of sequence analysis that Stovel uses is called optimal matching. Andy Abbott has done a few weird things with it, besides his famous work on occupations using it, he has a weird little article looking at diffusion of the welfare state that used optimal matching.

Sorry, the fact I couldn't remember that name was killing me.

[SocioEd]

If you are referring to quantitative sociology (not mathematical sociology), maybe also look at Stephen Raudenbush (Hierarchical Linear Models, usu. applied in Educ), Yu Xie ("Demography: Past, Present, Future").

Urban/Social strat is a big big field. International comparison sounds popular recently (which I try to do too). To start with social strat, you may be interested in scanning Grusky's "Social Stratification", which includes most classic works in the field.