Scraping Data and Making “Big” Inferences
Arnold F. Shober
Abstract: “Big Data” does little to explain the human condition, but it offers unprecedented opportunities to model how people choose. Professor Shober will describe how Google and Amazon know what you want with uncanny accuracy, and how in his research program he uses similar tools to examine how journalists cover politicians. He will also discuss some of the practical and statistical difficulties when analyzing billions of data points.
The talk is March 6 at 11:10 a.m. in Steitz Hall 102.
UPDATE: A very good talk. Unfortunately, we did not get video for his one.
Government department Professor Arnold Shober is in the news, handicapping the fall elections for the Appleton Post-Crescent. Here’s a portion of the transcript in response to a question on whether Wisconsin Democrats are going to be taken to the woodshed this November:
The answer to that question is really how well Russ Feingold does. He’s the top race in the state. We think about governor, but Feingold gets a lot more attention. If he has trouble — and a series of recent polls suggest that he really does — then it will very bad. If he can’t carry reliable Democrats into the polling booth, nobody can.
Polling out last week suggested that, in northeast Wisconsin, Ron Johnson is up 60-40. That means Steve Kagen (in the 8th Congressional District) is certainly out. Penny Bernard Schaber (in the 57th Assembly District) is most certainly out. Tom Nelson’s seat (in the 5th Assembly District) will probably flip, and on and on and on, because that’s a big margin.
Certainly, everyone voting for Russ Feingold is going to be voting for a Democrat. He’s a maverick, but he’s definitely on the left. So if he does poorly, Democrats across Wisconsin are going to be in big trouble. If he can hold it even, it may not be so bad.
Emphasis is mine (Prof. Shober typically doesn’t speak in bold font), and included so I could use the pun in the title.
You can see the latest polling results outside of Professor Shober’s door right here on Briggs 2nd.
This looks interesting. The best way to foment effective policy outcomes is to allow administrative agencies to do their thing unfettered. How do they solve the agency problem? I guess we’ll have to read the book and find out.
Splintered Accountability: State Governance and Education Reform
Arnold F. Shober
The No Child Left Behind Act declared that improving education in every school in the United States was a top national priority. However, this act did not acknowledge how state departments of education have successfully constructed reforms for the past few decades, despite the power struggle between governors, legislators, school districts, and state boards of education. Drawing upon archival sources, state budget documents, interviews, and statistical analysis, Splintered Accountability amply demonstrates that sustained education reform is best left in the hands of the relatively autonomous state departments of education in order to maintain curriculum standards, school finance, and teacher licensure systems. Comprehensive and successful education reform originates from within state education agencies, propelled by savvy state superintendents.