We had our first annual statistical teaching, application, and research conference here at Columbia last Friday. The goal of the conference was to bring together people at Columbia who do quantitative research, or who teach statistics, but are spread out among many departments and schools (including biology, psychology, law, medical informatics, economics, political science, sociology, social work, business, and many others, as well as statistics and biostatistics).

Both the application/research and teaching sessions went well, with talks that were of general interest but went into some depth, and informed and interesting discussions.

Applications and research

The morning sessions on applied statistics were fascinating. We did not plan a theme but it turned out that all four talks involved networks. Our outside research speaker (from AT&T Labs in New Jersey), Chris Volinsky, talked about detecting telephone fraud using telephone call networks. The datasets are large and it’s a real-time problem, so they have to store the network information efficiently in a way that allows them to update it daily. Tor Wager (Psychology) discussed statistical methods for brain imaging (FMRI), which of course involves real neural networks. Andrey Rzhetsky (Medical Informatics) discussed his research program for automatically parsing the thousands (millions?) of biomedical research articles to obtain a consensus on scientific research on links between molecules. Finally, Elwin Wu (School of Social Work) described a study of the effectiveness of a method of safer-sex education that involves communication with both members of an at-risk couple.

There was lively discussion of all four research talks; for example, Gueorgi Kossinets and Matt Salganik, two sociology graduate students, were interested in the social implications of the telephone network data (How many people do you typically call each month? How dense is the network structure of telephone calls?) and in what it takes for biologists to reach consensus–possibly false consensus–as revealed by the biomedical research findings.

Teaching

The afternoon session on teaching began with a lively presentation by Dick De Veaux, award-winning statistical educator and professor at Williams College, on the teaching of introductory statistics. The talk was called, “Math is music; statistics is literature,” and discussed how statistics, unlike mathematics, is informed by life experience–and how we can use this perspective in our teaching. Dick answered questions and then we had a lively roundtable discussion with David Epstein (Political Science) on the challenges of teaching introductory statistics to different audiences.

Posters

Finally, we had an interesting selection of posters from Ph.D. students, on topics ranging from probability and statistical computation, to environmental measurement, network sampling, statistical education, and computer science. Each poster participants received a small prize, and the grand prize (a $200 Barnes & Noble gift certificate) was won by Fanesca Young, for her poster, “Familial co-aggregation of two disorders: a common risk factor or a causal relationship.” Congratulations to all the students for excellent posters. After a few more of these conferences, the walls in our new building will be full of interesting research!