9 posts tagged with statistics by escabeche.
9 posts tagged with statistics by escabeche.
Displaying 1 through 9 of 9.
Happyism: The Creepy New Economics of Pleasure. Economist Dierdre McCloskey, in the New Republic, digs into the mathematical underpinnings of the scientific study of happiness. Executive summary: she doesn't like what she finds.
galton.org is an exhaustive website devoted to the life and works of the statistical pioneer and "father of eugenics" Francis Galton, inventor of the scatterplot, the correlation coefficient, fingerprint identification, and who knows what else. Almost all of Galton's books and papers are reproduced here, some in scanned form and some in searchable .pdf, from his major books to his letters to Pigeon Fancier's Journal. A short selection after the fold. [more inside]
Value-added model scores for teachers: some disturbing scatterplots. Gary Rubinstein finds a lot of noise and very little signal in the VAM scores of 18,000 New York teachers, recently released by the Bloomberg administration under a Freedom of Information Act Request. (VAM previously on MetaFilter.)
“certain styles of research were suggested to be prone to ‘groupthink, reduced creativity and the possibility of less-rigorous reviewing processes.’ Edward Wegman is a professor at George Mason and a distinguished statistician with a long career, a former winner of the ASA's Founders Award. In 2006 he testified before Congress on climate science, sharply criticizing the statistical methodology of Michael Mann's "hockey stick graph," which showed a sharp increase in global temperature in the last part of the 20th century. One section of Wegman's testimony concerned "social network analysis," and suggested that Mann's tightly-knit network of co-authors might have led to insufficiently aggressive peer review. USA Today reports that Wegman's testimony contained a substantial quantity of plagiarized material, and the peer-reviewed article derived from the testimony has been retracted by the journal that published it. John Mashey has compiled an obsessively thorough catalogue of the plagiarized text. (large .pdf.) [more inside]
"Value-added modeling is promoted because it has the right pedigree -- because it is based on "sophisticated mathematics." As a consequence, mathematics that ought to be used to illuminate ends up being used to intimidate." John Ewing, president of Math for America and former executive director of the American Mathematical Society, criticizes the "value-added modeling" approach used as a proxy for teacher quality, most famously in a Los Angeles Times story that called out low-scoring teachers by name. A Brookings Institution paper says value-added modeling is flawed but the best measure we have of teacher value, arguing that the metric's wide fluctuations from year to year are no worse than those of batting averages in baseball. (Though the weakness of that correlation is mostly a BABIP issue.) Can we assign a numerical value to teacher quality? If so, how?
Sabermetrician in Exile. Voros McCracken's radical idea -- that pitchers have very little ability to induce batters to hit into outs, and succeed mostly insofar as they can strike out a lot of hitters and give up few home runs and walks -- has changed the way baseball teams are constructed. (Heard of BABIP? That's him.) Every major league team has employees who rely on McCracken's insights. McCracken, struggling to make his rent in suburban Phoenix, isn't one of them.
Kaggle hosts competitions to glean information from massive data sets, a la the Netflix Prize. Competitors can enter free, while companies with vast stores of impenetrable data pay Kaggle to outsource their difficulties to the world population of freelance data-miners. Kaggle contestants have already developed dozens of chess rating systems which outperform the Elo rating currently in use, and identified genetic markers in HIV associated with a rise in viral load. Right now, you can compete to forecast tourism statistics or predict unknown edges in a social network. Teachers who want to pit their students against each other can host a Kaggle contest free of charge.
After five years of number-crunching and methodological controversy, the NRC's rankings of US graduate programs were released today, three years after the target date and fifteen since the previous ranking. Peruse the results at phds.org. Instead of numerical ratings, the NRC released two rankings, the "R-ranking" and the "S-ranking", each one with a wide error bar around it. Confused yet? Brian Leiter thinks the philosophy rankings "qualify as somewhere between "odd" and "inexplicable."" The University of Washington's CS department says their ranking of 15-32 is "clearly erroneous." Obviously, the only appropriate response is to compute asymptotic formulae for the number of possible fuzzy rankings.
Mercenary Epidemiology: Data Reanalysis and Reinterpretation for Sponsors With Financial Interest in the Outcome. (.pdf link) When should scientists be required to release their raw data for (potentially hostile) re-analysis? A letter to the editors of Annals of Epidemiology from David Michaels, Ph.D., MPH, public health blogger, author of the book Doubt Is Their Product, and, as of December 2009, the Assistant Secretary of Labor for OSHA, unanimously confirmed by the Senate despite the dismay of some. Michaels interviewed at Science Progress about Doubt Is Their Product (podcast, with transcript.)