research methods

Research Methods Resources: quantitative analysis

Here at Le Moyne I regularly teach a research methods course that serves criminology, sociology, and political science students. It is a survey course, by which I mean I cover a broad range of topics. I typically start with the epistemology of the social sciences, move into research design issues like measurement and connecting empirical work to theory, and then spend the second half of the semester focusing on specific analysis procedures. The course covers both quantitative and qualitative methods, so very little time is allowed for each.

I spend about 3 weeks on quantitative methods, hoping that by the end students are relatively comfortable with understanding and producing descriptive statistics, crosstabs, t-tests, correlations, and regressions. This year I covered logistic regression in more detail than normal. It’s a lot to cover in that time, so over the years I’ve created some handouts meant to help students do the work we have limited class time to practice.

I’m going to put two handouts here. One is meant to help students work in SPSS to do some basic quantitative analysis. The other is meant to walk them through using the SDA resource from Berkeley. One year I taught SDA exclusively because it’s free and gives students access to some high quality data with lots of social science applications.

I offer these resources ‘as is,’ and am confident that there are mistakes – I catch some every year and surely make more as I expand them. It’d be great if others found these useful, and I’ll only be reasonably embarrassed if you point out errors I’ve made (hopefully nothing super-serious!).

SPSS handout (MS Word)


SDA handout

SDA.Handout (MS Word)


Bad, politically motivated sociology is being used for bad political purposes? Regnerus is appalled!

The Regnerus Affair will not go away. At least not until the piece of bad sociology that started it is retracted. It certainly appears that the odds of retraction are slim because there are powerful sociologists defending our institutions and practices in the abstract. This is so even when it looks clear that this is more than bad sociology, but in fact a result of a certainly failed, and arguably corrupt research and peer review process. This is particularly shameful because the research is regularly being used by those who wish to deny equal rights to gays and lesbians across the globe.

Regnerus keeps talking about it too, which looks to be good for his career. His latest comment on the affair showed up here on the Atlantic Wire. Regnerus is apparently upset that his piece of bad sociology, which sloppily contends that children of same sex parents face negative outcomes, is being used by Russian politicians to argue that ‘science shows negative outcomes for children raised by same sex parents.’ It’s shocking, isn’t it? Who would have thought politically charged research could affect politics? Why bother to do that sociology well, or put it through the same rigorous peer review process as the reams of politically inconsequential sociology that is rejected every day? No reason according to SSR, I guess. Using my sociological imagination, it occurs to me that maybe Regnerus’ (bad) paper is actually the cause of more instability for same-sex headed families, and therefore harmful the parents and children in those families? Dr. Tey Meadow said this better than I ever will, right here.

Like other examples of Regnerus’ writing that I’ve responded to here at the Morass, there are a few sentences that jump off the page. I’ll quote them:

This may come as a surprise to those who have spent the past 15 months tagging my study as discredited or “debunked,” a silly and simplistic moniker given that the data is public and the analyses in the article are rather straightforward.

Regnerus is right that a lot of sociologists and activists have spent a fair amount of time critiquing his methods, his findings, and the peer review process for his original paper. He’s smugly wrong to call that ‘moniker’ silly or simplistic. There are good reasons for scientists and activists to discredit the research.

That’s bad, but, here’s what really got me. When he defends the legitimacy of his study by saying the data are public and his analysis is straightforward. I don’t recall the public availability of the data ever being central to any criticism of the study, and his measurement of same-sex parents was anything but straightforward. His measure wouldn’t pass the ‘face validity’ test I just taught in my undergrad methods course last week. That his paper sailed through peer review at what I would have considered a respectable journal makes this marginal sociologist want to shut down Stata and call it a career.

Regnerus’ ability to flatly ignore the very sound criticism of his study, and the process that gave it to Russian politicians and the U.S. Supreme court on a silver platter, is either the blindness of privilege, or simple disregard for those who dare question his work and position. It’s hard for me to see it any other way.

And then that paragraph got more maddening:

Isn’t it hypocritical to blow the whistle on this use of the data while supporting other such uses, such as my own participation on an amici brief to the U.S. Supreme Court? No, it is not, because I oppose same-sex marriage and lawmaker Andrei Zhuravlyov’s draconian legislation for the same reason: every child has a mother and a father, and such kinship matters for kids. To be stably rooted in your married mother and father’s household is to foster the greatest chance at lifelong flourishing. It’s not necessary, of course. It just has the best odds.

Here, in passing, he blithely presents one of the least sophisticated arguments against same-sex marriage there is. He writes “I oppose same-sex marriage and lawmaker Andrei Zhuravlyov’s draconian legislation for the same reason: every child has a mother and a father, and such kinship matters for kids.”

Marriage is not about kids. Marriage is about adults who want to marry one another. Marriage is about mutual care and enjoying each other’s company. Marriage is about federal rights. Yes, I do take this personally. I’m married, without kids. We’re not going to have kids. My marriage is not suffering for it. No one else’s marriages or kids are suffering for it, either. I’m not making my community less stable. And, most important, I’m not doing lousy research to protect my privileges or to promote my moral preferences as science.

I guess I’ll end with a poorly stated hypothesis: A society with equal marriage for all will be more stable, and good for kids. I think that’s true and will make for a good society. And I don’t even like kids.

Teaching the Regnerus Controversy

I’ve been thinking about how I might use the debate about Mark Regnerus’ article “How different are the adult children of parents who have same-sex relationships? Findings from the New Family Structures Study” from Social Science Research to cover the research process, the social scientific community, and research ethics the next time I teach undergraduate Research Methods (and how it might be used in other courses). I won’t retrace the controversy, but here is a good rundown from fairly soon after the article was published. I’m assuming it’s not yet over as this very good blog from Neal Caren appeared just a few days ago at Scatterplot.

From a teaching standpoint, I think the ‘Regnerus Affair’ has a number of advantages over the examples often used to teach about ethics in social scientific research. I would guess that most who teach methods use common examples like the Stanford Prison Experiment, Tuskegee, Milgram, or the Tea Room Trade. I’ll continue to use these examples, even if just for shock value (get it!), but the Regnerus case has the advantage of being not only current, but also ongoing.