I am a statistician and when I look at the data...

I am a statistician and when I look at the data, I have to say, there is something there.......

This is how an ABC special on Extra-sensory perception [ESP] begins. There are psychics and politicians; there's the CIA and some statistics.

Entertaining....... Get the video here

Job Opportunity for Graduate Students

I do not know who is running this site. But graduate students are always starving and they should not pass on money. Check this job offer. It seems they pay grad students for (short) reviewing papers.

Interesting Seminar here at UCI

The Department of Statistics here at UCI is having Peter Grünwald for a talk on Wednesday, January 31st. His research seems very interesting.

Here the abstract:

We show that forms of Bayesian learning that are often applied to classification problems can be *statistically inconsistent* when the model is wrong . More precisely, we present a family ('model') M of probability distributions, a distribution P outside M and a Bayesian prior distribution on M, such that - M contains a distribution Q within a small distance \delta from P. Nevertheless: - when data are sampled according to P, then, no matter how many data are observed, the Bayesian posterior puts nearly all its mass on distributions that are at a distance from P that is much larger than \delta. The classifier based on the Bayesian posterior can perform substantially worse than random guessing, no matter how many data are observed, even though the classifier based on Q performs much better than random guessing. The result holds for a variety of distance functions, including the KL (relative entropy) divergence.

Misspecification themes are usually overlooked by the Bayesian statistical literature and it could not be differently. As long as one is concerned with actions to be taken and these action are taken with respect to the likelihood principle misspecification is noting more than a nuisance. Clearly the paper above add a (welcome) frequentist perspective.

I am looking forward listening what Peter has to say.

iPods set the tempo in currency

The Economist's BigMac index has had his fortune as informal way of measuring the purchasing power parity (PPP) between two currencies.

Now Commonwealth Securities has launched an iPod index, based on the cost of the 2 gigabyte iPod Nano player as a way of showing if a currency is overvalued. BBC has a report on it.

P.S.: If you plan to buy a iPod and you leave up north you should head your way into Canada.

Graph your way into the new millenium

This graph is very interesting. It plots income per capita for many countries over-time.

Heckman: a myth debunker

In a recent paper in Sociological Methodology, James Heckman refers to "the myth that causality can only be determined by randomization, and that glorifies randomization as the ‘‘gold standard’’ of causal inference."

Heckman's article is a response to this article by Michael Sobel, who argues that Heckman's methods are actually not so different from the methods commonly used in statistics.