Friday, October 19, 2007

Component analysis, causal inference, and general intelligence

The aim of astronomy is astrophysics - we observe to Universe with the hope of using the resulting data to understand the fundamental physical processes that give rise to its observed properties.

As with many sciences the data obtained from observation (experimentation, in other sciences) itself does not uniquely tell you the physics or what caused what. Instead one normally investigates to look for correlations between different aspects of the data.

For example it is known that the surface brightness, effective radius and velocity dispersion of the stars in elliptical galaxies are strongly correlated, a result now called the fundamental plane. Another example is that in starburst galaxies the soft X-ray luminosity is linearly proportional to the galaxies far-IR luminosity because, causually, the FIR traces the formation rate of massive stars, the same stars that very rapidly die and whose supernovae heat the ISM to X-ray-emitting temperatures.

Various methods of investigating correlations between multiple variables exist (e.g. principal component analysis), now often referred to as "data mining." The problem is that these methods, while useful at recasting the data in ways that aid visualization of any correlations in the data variables, do not necessarily tell you what caused what.

An interesting discussion of these often-forgotten issues and complexities, one is applicable even to astrophysics, can be found in Cosma Shalizi's article on the myth of g, the so-called general factor of intelligence. Indeed, he argues that while factor analysis is perfectly valid for data exploration or model testing, as a method for finding causal structure it is not reliable (it can be right, but often its completely wrong and can fool you).

All very interesting, and rather important to understand in the wake of a certain elderly Nobel-prize winner's recent counter-factual comments.

Wednesday, October 17, 2007

Rumors of the LTSA program returning

Steinn Sigurdsson reports rumors that the NASA LTSA (Long Term Space Astrophysics) funding program may return. The LTSA program has been not offered for several years due to budget cuts.

This would be good news for younger researchers in the space sciences, in particular those using mainly HST, Chandra, XMM-Newton or Spitzer data, who are ineligible for the currently remaining funding programs like the ADP. Let us hope the rumor is true.

Monday, October 15, 2007

LibraryThing

I spent a large part of Sunday adding the non-fiction part of the home and work book collections to LibraryThing, an online book collection web site. You can view my library here (or click the link at the bottom of the blog). A short introductory guide to LibraryThing displays many of its features.

I have wanted an easy-to-use book cataloging tool for some time now, largely to have a list of my book in case of accident or theft, and also to prevent the purchase of duplicate copies of books (which can be an expensive mistake when you're buying high level science textbooks).

Many of the available open-source list or catalog making software is not book-specific, and hence leaves much to be desired. In contrast, LibraryThing offers many of the features I've wanted and a few I hadn't even considered. Item entry is very easy, you can pretty much rely on entering the ISBN number and LibraryThing recognizing the correct book and edition (although the cover images are occasionally incorrect).

It also offers a variety of tools for blog integration - for example a random collection of covers from my collection should appear in the side bar to the bottom right of this blog. So far it appears random but static, as the covers do not change when I reload the site.

One drawback is that the free accounts are limited to 200 books, rather restrictive. A lifetime membership with unlimited books is $25. So for now I've limited my catalog to non-fiction history or science books, everything else will go in another free public LibraryThing when I get around to it.

Swara deserves thanks for bring LibraryThing to my attention.

Friday, October 05, 2007

Judge, fearing "Sleeper Cells", rules in favor of universal background checks

Judge Otis Wright has ruled in favor of NASA and against a group of JPL employees in a case regarding invasive background checks that would allow the questioning employees of sexuality, medical records and finances. These were ordered by presidential directive in 2004 for all national laboratories, of which JPL is one. "...I want the security of this nation preserved,'' Wright said Monday. "I don't want any sleepers infiltrating NASA or JPL.''

Apparently a bunch of uneducated turbaned guys hanging out in caves in Pakistan might be able to put "sleeper cells" into positions in NASA or JPL that don't even deal with classified information but that will still threaten the "security of this nation", even though this was something that clearly wasn't even a possible threat from the Soviet Union during the Cold War.

Link: SPACE.COM.

Wilkins on Feyerabend


John Wilkins at Evolving Thoughts has an interesting essay-length article on the philosopher most guaranteed to make a scientist role their eyes in scorn: Feyerabend. Well worth reading, especially regarding the origin of Feyerabend's ideas and their consequences in todays era of special interest denialist "think tanks".

[Image of Paul Feyerabend from https://webspace.utexas.edu/cokerwr/www/slides/philosophers.html]

Thursday, October 04, 2007

Dark Energy and Straw Men


On today's set of astro-ph preprints is a conference proceedings article by Andreas Albrecht (astro-ph/0710.0867) entitled "The case for an aggressive program of dark energy probes" caught my eye. Specifically the abstract attracted my attention:

The observed cosmic acceleration presents the physics and cosmology communities with amazing opportunities to make exciting, probably even radical advances in these fields. This topic is highly data driven and many of our opportunities depend on us undertaking an ambitious observational program. Here I outline the case for such a program based on both the exciting science related to the cosmic acceleration and the impressive impact that a strong observational program would have. Along the way, I challenge a number of arguments that skeptics use to question the value of a strong observational commitment to this field.

From the abstract I had hoped this would address and rebut the criticisms raised by Simon White earlier this year in "Fundamentalist physics: why Dark Energy is bad for Astronomy" (astro-ph/0704.2291). Unfortunately Albrecht's article does not address any of these substantive criticisms regarding discovery space, scientific method, whether Dark Energy experiments advance our knowledge of astrophysics, or experiment systematics, so for now I am left to believe that White's criticisms remain valid and significant.

Don't misunderstand me - IF we can confirm and constrain the nature of dark energy it may revolutionize our understanding of particle physics, but its role in explaining the Universe of galaxies, stars and planets that we live in is extremely limited. By all means fund dark energy probes out of the DoE and classics physics funding, but significant progress in astrophysics will be stifled if too much astrophysics money is diverted into DE.

Understanding the nature of the dark matter particle would also advance fundamental physics greatly, and dark matter plays a much larger role than dark energy in shaping the nature of the Universe and the specific objects we live in than dark energy does. Yet the discovery of the dark matter wasn't met with a program of dropping successful existing astronomy projects and throwing vast resources at satellites that could only quantify one number about dark matter.