Chris Uggen's Blog: multiracial victimization part deux

Friday, September 30, 2005

multiracial victimization part deux

in response to my why are those of two or more races victimized at such a high level? post, erik larson of macalester college has already come up with at least a partial answer. he explains a large portion of the gap in victimization between multiracial individuals and all other racial groups with a single variable. based on his quick calculations, age seems to account for about 40 percent of the excess victimization over the population as a whole. i think that developing this analysis might make a nice research note when the data arrive in full (just make sure to give erik some authorship credit -- dude is on the tenure track!). it doesn't necessarily kill my discrimination hypothesis, but at least it reduces the magnitude of the multiracial victimization gap from "shocking" to somewhere around "disturbing." thanks erik.

i don't know the literature well, but i'm interested in the social construction of racial and ethnic identity for personal as well as scholarly reasons. people immediately peg me as swedish or norwegian (especially outside of minnesota) because of my physical appearance, and this is generally the way i self-identify. but i certainly grew up knowing that my background was at least somewhat more diverse (e.g., cree (nêhiyawêwin) indian, italian, irish...). if people who self-identify as biracial or multiracial are more likely to be victimized net of age and everything else, this doesn't bode well for the american mosaic. am i really less likely to be beaten or assaulted because i look like i'm from one and only one racial background? i'm not advocating a melting pot or assimilation model, but i can't help thinking we'd be better off if more than .9 percent of us (the figure in the ncvs) self-identified as "2 or more races."

on a much, much, lighter note, erik's response got me thinking about another kind of diversity -- the ability to claim "2 or more methods." since erik is best known for his qualitative comparative work, his sharp quantoid contribution helps make a case that i preach to grad students early on -- that there's no such thing as pure "qualitative" or "quantitative" researchers anymore. i probably look like a quant in the same way that i look like a norwegian and i emphasize quant tools in my work. still, few of the top sociologists i know advocate a "purebreed" approach to methods these days. here's a cheeky take on the advantages and disadvantages of [sociological] purebreeds and mutts, adapted from the "pet library:"

Take two [sociologists], one a [methodological] purebred and one a mixed breed. Which one is more beautiful, smarter, a better companion? It's often a matter of personal opinion as there are those who believe that purebreds are the only choice while others steadfastly stand by the mutt. One of the most appealing features of the purebred is that they have rather predictable [scholarly] characteristics. You more or less know what you're going to get when it comes to appearance and size. They also have a fairly predictable temperament so you can get a pretty good idea of what your [sociologist's] disposition will be. And, of course, if you want to professionally breed or show your [sociologist], a pedigree is your only option.

[Methodological] purebreds are more prone to [career] problems, many of which are often due to overbreeding. It's very important when considering a purebred that you find a reputable and proven [graduate program]. Many purebreds also come with working behaviors that may not fit your ideal of the perfect [sociologist]. Behaviors like digging holes, chasing after things or nipping, which have been bred into certain [epistemologies] for centuries, often prove difficult to change. And, of course, purebreds can be very costly, running anywhere from several hundred dollars to over a thousand dollars.

Now for the mixed breeds. One of the most obvious downsides of the mixed breed is that you can't predict what a [new ph.d.] will look like or what size it will be as an adult. However, there are those who actually find this to be a positive, enjoying the surprise of realizing what their mixes grow up to look like. Along the same lines, mixed breeds are less predictable than purebreds when it comes to temperament. However, mixed breeds tend toward the moderate, with their temperaments often proving to be less extreme than those of purebreds. You're less likely to have a [sociologist] who's "very" energetic or "very" demanding or "very" stubborn, with [epistemologically]-based characteristics that often prove difficult to change, and more likely to have a [sociologist] that can adjust to a greater variety of situations. With a greater [methodological] diversity, mixed breeds are less likely to suffer conditions that affect certain purebreds as a result of inbreeding. They also tend to be a lot less expensive, usually costing around $25 to $75 at most [graduate programs]. Furthermore, by opting for a mixed breed, certainly one from the [non-elite departments], you may just be saving a life.


At 3:21 PM, Anonymous Minor Threat said...

Chris, I strongly agree with the sentiment in your post (and strongly prefer mutts to purebreds with actual dogs as well) but I worry a bit about doing each method well. My favorite books in sociology and criminology employ "mixed methods" and I aspire to it myself but I also worry about one scholar doing each method well, rather than including one as an "add-on."

Though my "favorites" use mixed methods, I can think of innumerable examples of great quantitative work that clumsily includes snipets of interviews with no way for a reader to assess whether they were chosen for convenience, relevance, or representativeness. Alternatively, I have read great qualitative ethnographies with goofy stats thrown in at the end and which may or may not be correctly specified, but I often have my doubts.

I think too of the ways in which graduate programs teach methods -- though my experience is small (n=2), it seems that most departments encourage specializing in one method or the other. What's the incentive to be good at both? It also makes choosing a publication outlet a little tricky. So, Mr. Soon-to-be-Chair, how do you train more erik larsons? Feel free to start with me... What should I be doing?

At 3:36 PM, Anonymous chris said...

good point, mt. your "add-on" critique is on target (ouch!). one thing that seems to be changing is that people seem more sincere in appreciating mixed methods or methods other than those they use themselves. a (scholarly) generation or so ago many just gave lip service to these ideas, and a generation before that open resistance was not unusual.

i can offer a few suggestions on your (how?) question:
(1) let the research questions be your guide;
(2) don't flinch at the start-up costs;
(3) seek as much methods training as you can get;
(4) seek such training as you are chasing a question that you've just gotta know the answer to; and
(5) find good and generous collaborators with training and skills that complement your own (e.g., ms. blackstone taught me much on my s.h. project).

At 4:32 PM, Anonymous Minor Threat said...

All good suggestions... I guess I worry mostly that although most of us don't do both well (irrespective of our sincerity in arguing that we should), we are still expected to review or critique others who do both, one, or the other.

A semi-autobiographical semi-hypothetical example... My dissertation group has five members; two highly quantitative, two highly qualitative, one a great example of "mixed". The two highly quantitative members often remark (quietly and only to one another) that we're not *quite* sure how to go about evaluating the comparative historical, ethnographic, or archival research of the other members on methodological grounds. What makes it good? What makes it bad? How many documents is enough? How many sites is reasonable? (see, we also tend to think in terms of 'how many', which may be part of the problem)

What we usually end up saying to one another (much more quietly) is that we know person X is smart and careful and a good person, thus we think they use responsible methods. This is not a good thing...a little institutional support for expertise in "2 or more" methods would be.


Post a Comment

<< Home