Tuesday 27 January 2009

How R you at statistics?

[Disclaimer] This post is little more than an exclamation of science-geek joy at some media attention for the stats program I use. Non-geeks are free to stop reading now, lest you be contaminated. Also, I must apologize that it is a little dated now, the original news item that sparked it having come out 3 weeks ago.

Research in science invariably includes running statistical tests. The more complicated the experiment, the more complicated the test to determine whether my results are significant. I am delighted that the research community has a tool like R, which in addition to being fully-featured, extensible, and producing pretty figures, is free software (as in free beer, but also as in free speech). There are other tools out there that have comparable features (so I hear, anyway), but as a student I could not have afforded to put them on my personal computer.

So I'm a big fan of R, and am delighted to report that it is gaining headway in my local linguistics department against the old commercial packages.

Imagine my glee, then, to learn that the New York Times published an article earlier this week on the merits of R for data analysis!

Of course, the R discussion lists are all abuzz with the news. People are pleased that it's getting some wider attention, but also (of course) noting the apparent inaccuracies of the piece. None of them are catastrophic, and the writer of the article addresses some of them in his blog follow-up. Which has itself been responded to.

More on R and other free software projects in future posts - for now I'm just enjoying the glow of almost-fame for my little corner of geekdom. Ahhh.

Tuesday 20 January 2009

Mooning over creationists (sequel)

In a post last year, I noted an intriguing question asked on the website of the local Edinburgh Creation Group (ECG).

They asked how the moon got into orbit.


The standard answer emerging from science is that some large object about the size of Mars (sometimes called “Theia”) hit the Earth very early on in the formation of the Solar System. That impact threw up debris which coalesced into the Moon. This hypothesis explains many otherwise-odd facts about the Earth-Moon system, such as their relative composition and their angular momentum.

However, the ECG folks claim to have found a problem. It’s very simply and persuasively illustrated by this simulation of orbital mechanics, which was linked from the ECG's website*. Play with it a bit yourself, to get a feel for the principle involved. Basically, once other forces (rocket thrust, planetary impact, atmospheric friction, etc) have stopped acting, only gravity is left. And when two objects (say, a fragment of debris and the Earth) are interacting only by gravity and are not substantially affected by external forces, there are three possible outcomes:
  1. The smaller object already orbits the larger (or they orbit each other), and will continue to do so indefinitely.
  2. They are on a collision course, and will impact (probably not parting again).
  3. The smaller object meets or exceeds escape velocity, and leaves.
There is no trickery here – the simulation linked to is a faithful representation of two bodies interacting gravitationally. Now, I am not a creationist. I tend to distrust claims from people with a vested interest, unless they can demonstrate that they've tried to set aside their own bias (for example, by doing actual science). I therefore put more trust in the scientists who report in a peer-reviewed forum, suggesting they've worked out the bugs in their hypothesis. But I do like to understand things. The creationist question is a valid one, a question I wanted to know the answer to. So in my earlier post, I asked my readers to help me out and let me know what the simulation is missing. One of them pointed to a researcher who has some animations on her website. However, the details of exactly how the creationist argument falls down were not obvious there. So (after a thesis-induced hiatus) I finally decided to look it up myself. Here is what I found.



First, I learned that scientists occasionally call it the Giant Splash rather than the Giant Impact, because at that scale even bodies of solid rock behaves more like a liquid when thumped together.

Second, when a bunch of debris is flying through space above the Earth's surface after such an impact, there is a very obvious force present that is adjusting the orbit of the fragments: the other fragments! Collisions and gravitational interaction between the bits of debris (which, remember, must together have massed at least as much as the moon) caused some of the fragments to settle into a stable orbit, where over time they accumulated into the single satellite we know today.

All this is a very abstract description of the answer. The details take massive computing power for simulations of all the particles thrown up by a big splash. Researchers have simulated many different types of impact, with different impactor sizes, different angles of impact (from head-on to glancing blow), and so on. Not all of the questions are resolved, but the problem highlighted by the ECG is certainly a non-issue.

If all that is too cerebral for you, I'll repeat something I said in the earlier post: I love games that use orbital mechanics. KSpaceDuel is one that models only a single centre of gravity, so if you don't use thrust you'll just stay in your initial orbit (until your opponent blows you up, anyway). [Note that it's a Linux-only game - yet another reason to start shedding that Windows or Mac addiction.] And for those of you who haven't yet converted to Linux, Orbit is an online flash game that very neatly illustrates how complex the gravitation effects become when you’re dealing with more than a couple of objects.

References:
Notes for an undergraduate course at UCL (detailed)
New Scientist, 2006 (summary)

* I am currently unable to find the link - they seem to have redesigned their site and left it out. I therefore cannot be certain whether or not the claim is supported by the members of the ECG.

Thursday 15 January 2009

Truth about dyslexia

Gentle and thoughtful Cath is on the warpath - and justifiably so.

Some loon (Graham Stringer, MP for Blackley) is trying to deny the existence of dyslexia, a well-evidenced and widespread disorder, using false claims, fallacious reasoning, and other tactics familiar to rational people.

I cannot add anything to her very lucid demolition of his article. I'll just give a couple of potentially useful links for further reading.

Here's an international list of Dyslexia research and support organizations that a quick search brought up, in case anyone wants to read further: Dyslexia Parents Resource.

And if you're from the Blackley area, check out Stringer's contact info, voting record, etc. on this page.

Tuesday 13 January 2009

Primo Levi - quote

From Primo Levi's afterword to the combined volume of If This is a Man and The Truce, which relate the author's experiences as a prisoner at Auschwitz and on his journey home.
It is, therefore, necessary to be suspicious of those who seek to convince us with means other than reason, and of charismatic leaders; we must be cautious about delegating to others our judgment and our will. Since it is difficult to distinguish true prophets from false, it is as well to regard all prophets with suspicion. It is better to renounce revealed truths, even if they exalt us by their splendor or if we find them convenient because we can acquire them gratis. It is better to content oneself with other more modest and less exciting truths, those one acquires painfully, little by little and without shortcuts, with study, discussion, and reasoning, those that can be verified and demonstrated.

Tuesday 6 January 2009

One step closer

I have submitted my PhD dissertation. It is right now in the hands of my examiners, who are reading it and deciding whether I'm worthy of three more letters after my name. I'm one step closer to being a fully-credentialed scientist. Pretty cool.

This also means I'll have time to get back to a somewhat more regular blog schedule, and to participating in the online humanist community a bit more. But I do still have a viva coming sometime soon, for which I need to prepare; I do still have a post-doc jam-packed with paper-writing for the next year; and I do still have a gregarious daughter and a patient wife who I want to spend the bulk of my free time with.

So please bear with me as I look for the right balance between family, work, and you. And, if you have any particular hopes or aspirations for this blog, let me know in the comments for this post.