This summer, editorship of Sociological Methodology transitions to Ohio State University (OSU) into the capable hands of two dynamic young associate professors there, David Melamed and Michael Vuolo. They are the 16th editorial team since the journal began in 1969, and the third duo. Their collaboration began early – both came to OSU as the first generation of a new interdisciplinary group (the Translational Data Analytics Institute) that puts them in touch with leading methodologists across fields. In particular, this position locates both in the center of cross-currents coming from other disciplines, most importantly computer science, and an awareness of the opportunities and the dangers therefrom.
David Melamed received his PhD in sociology from the University of Arizona in 2012, after earning his MA in sociology and his BA in sociology and philosophy from Kent State University. He has always had a flair for creativity and innovation, such as when, as a teenager, he set off (he claimed) “to go camping,” but returned with a wrecked car and a tattoo, after ending up drinking beer with hockey players in Niagara Falls.
Melamed became convinced in his graduate training that the weak link in our work was usually methodological, not theoretical, and he was inspired to pursue innovative methods by those who pioneered elegant, simple, and powerful approaches, such as Blau’s Inequality and Heterogeneity or Effron’s bootstrap, and the work of people like O.D. Duncan, Scott Eliason, Harrison White, and his collaborator and one of his mentors Ron Breiger. Like his inspirations, Melamed has strengths in multiple methodological approaches—in addition to classical statistics, he uses simulations and experiments, as well as various forms of categorical data analysis.
While Melamed’s main research directions turn on social psychological issues of cooperation and status, he also has published on new methodological approaches to mobility data and to cross-case comparison. (At OSU, he is affiliated not only with the Criminal Justice Research Center, but with the Mershon Center for International Security Studies.) His articles have appeared in ASR, American Journal of Sociology, Sociological Methods & Research (SMR), and Social Networks. A good example of Melamed’s creative approach is his sole-authored article in Research in Social Stratification and Mobility (2015), in which he proposes using community detection network methods to identify structures of occupational classes from a conventional mobility table. One can either see it as using network methods on the residuals from mobility tables, or a new, more theoretically informed, way of creating a modularity matrix for an eigenspectrum decomposition. As a scholar, Melamed joins this sort of creativity with a careful desire to recreate, whenever possible, plausible models for the generation of his data, combining the vision of a mathematical sociologist with the rigor of a true social psychologist.
Michael Vuolo received his PhD in sociology from the University of Minnesota in 2009, after receiving a master’s both in statistics and in mathematics there. He grew up in New Haven, CT, and, despite having no formal family connections to Yale, managed to make use of its munificent resources not only by sneaking into parties, but by finding the secret subterranean tunnels and roof passages that made for ideal hangouts. Yet, by daylight, Mike took college classes at Yale as a precocious high school student. But it was Brad Wright at UConn who turned Vuolo on to the path of research and the life of a professor.
Vuolo’s intellectual origin in mathematics shows. He never tires of going back to the most foundational issues in the derivation of statistics. His inspirations are less Blau and Duncan than Cauchy, Gauss, LaPlace, as well as statisticians like Fisher, Pearson, and Cox. And he sees the fundamental issues in the derivation of mathematical statistics, such as our need to grapple with complexity, as being relevant for qualitative methods as well.
His work brings cutting-edge statistical approaches to issues of crime, law, and deviance; he has studied both the predictors and the consequences of drug and tobacco use, as well as aspects of criminal justice procedures. Vuolo is affiliated with both the Institute for Population Research and the Criminal Justice Research Center at OSU; he has published in SMR, Demography, American Journal of Public Health and SPQ.
A good example of Vuolo’s mathematical flair in action is his 2017 SMR article on copula models. Conventional practice in sociology is to give the most cursory attention to the actual distribution of our variables: perhaps an egregiously skewed one will be logged, but that’s about it. Almost never do we grapple with the fact that many of our methods rely on assumptions about joint distributions that may fail to be satisfied by the data—and we do this because more flexible models can turn into computational nightmares. Here, Vuolo presents a range of models for quantifying the dependence between variables that make no such assumptions, and instead, take whatever marginal distributions are observed and use these to determine the appropriate quantification of dependence by considering their cumulative distribution functions. It’s an elegant, powerful, and important solution—but one that requires that we realize that there is an underlying problem, which is just the sort of recasting that Vuolo excels at.
The Future of Soc Meth as They See It
The overarching goal for Melamed and Vuolo is to make SM the obvious first choice for the submission of the best work in sociological methodology across a wide range of approaches, both numerical and non-numerical.