Results for “Watson”

84 found

In 1990 my colleague Robin Hanson wrote:

Imagine a betting pool or market on most disputed science questions, with
the going odds available to the popular media, and treated socially as the
current academic consensus.  Imagine that academics are expected to "put up
or shut up" and accompany claims with at least token bets, and that
statistics are collected on how well people do….

This would be an "idea futures" market, which I offer as an alternative to
existing academic social institutions.

More and more it looks like Robin was right on.  Consider this story from the London Times:

WHEN Ladbrokes teamed up with New Scientist magazine
in August last year to offer odds on five great breakthroughs being made by
2010, it looked like a typical silly-season stunt.

It is now expected to become a very expensive one. As soon as the book
opened, physicists began to put their money where their theories were and backed
themselves to find gravitational waves – ripples in space and time predicted by
Albert Einstein but not yet proven to exist.

Alan Watson, of the University of Leeds, was astounded
to see odds of 500-1 on a discovery that he considered a matter of when, not if,
and promptly wagered £50.

So many other scientists did likewise that by lunchtime Professor Jim Hough,
of the University of Glasgow, who leads a team seeking the waves, was allowed to
stake only £25 at odds that had fallen to 100-1. When his colleague Sheila Rowan
placed her bet in the early afternoon, the odds were down to 5-1, and when the
book was closed they were 2-1.

It’s amazing how far we have come since Robin proposed idea futures, especially given that the idea could have been implemented hundreds of years ago.  But Robin’s vision is even more radical than betting markets.  Robin proposes that betting markets can substitute for many of the funding arrangements that we use today.  Consider the part of the quote I excised above:

Imagine that funding
agencies subsidize pools on questions of interest to them, and that
research labs pay for much of their research with winnings from previous
pools. 

Imagine indeed!  We are not there yet but the odds are increasing in Robin’s favor.

I am teaching econometrics this semester and using a new book, James Stock and Mark Watson’s Introduction to Econometrics. It’s a very good textbook.

Stock and Watson use a “robust” estimator of standard errors right from the beginning. This means that they can dump an entire chapter on hetereoskedasticity and methods of “correcting” for hetereoskedasticity (these rarely worked in any case.)

They do not waste time discussing the difference between the t-distribution and the normal-distribution. Instead, they assume reasonably large datasets from the get-go and base their theorems on large-sample theory.

The book is not cluttered with examples. Stock and Watson use a handful of applications that they return to again and again as they introduce new problems and new techniques – thus simple regression is introduced with the goal of estimating the affect of the student-to-teacher ratio on test scores. The problem of omitted variable bias is then introduced and the solution of multiple regression then discussed. Later the same example is used to discuss fixed effects and so forth.

Finally, they have a good chapter on evaluating research designs for internal and external validity. In other words, they discuss how to tell the difference between a good study and garbage – really the most important asset for any reader of statistical work.

Two regrets. I would have liked an early chapter on exploratory data analysis. I would have loved a chapter on regression discontinuity design.