We examine the effects of students' physical appearance upon examination results. We f[o]und evidence that beauty has a significant impact on academic performance, a result which is consistent with and comparable to the impact found in labour market literature.

Correlation is not causation.

I'm going to have to read this paper to see whether they have controlled for the examiner knowing the examinee.

By laying out pros and cons we risk inducing people to join the debate, and losing control of a process that only we fully understand. - Alan Greenspan

An i a professional world, for what i ve seen, most of lawyers/top managers/executives are tall and "good looking" as well.

I've read a description of an experiment where a person was introduced to audiences of students under different academic titles.
When asked to estimate the height of the person, the audiences on average considered the same person taller if he had a PhD than if he had a Bachelor degree. The lower the status the shorter he seemed, and vice-versa.

It really is striking just how good looking many of the students are, especially compared to students from the local comprehensive. And given the educational values of this school it's a given that most of the students will be aiming for Oxbridge and high-flying careers in business, law and the media.

So I'm not totally skeptical about a correlation. Although working out the cause is much harder. I'd expect a strong element of class advantage in exam performance. And presumably, in a very simple Darwinian sense, more status and more money means you have a wider pick of potential partners, and are more likely to end up with someone considered visually desirable. So there's potential for a positive feedback effect to trend towards gene pool stratification.

(Which is not to suggest that good looks cause or always imply intelligence, or that lack of good looks suggests stupidity, of course.)

1. good looks in fact go, genetically, with higher intelligence;
   
2. good-looking people are the object of positive discrimination;
   
3. good-looking people gain self-confidence and high self-esteem from their mirror-image in other people's eyes.

(1) is way short of any scientific demonstration, and anecdotal evidence would suggest there are good-looking stupid people as well as less good-looking intelligent people.

(2) is something that surely happens. Good-looking (and tall) people get offered better jobs and better pay, for example. But here we're talking about written exam results (though if the writers of this piece didn't control for examiners knowing candidates, as Migeru says, that would seem to shoot their work in the foot.)

(3) Our status as individuals depends a lot on our perception of ourselves as others see us. Good-looking and tall people (whatever their social background) receive positive feedback from others, build high self-esteem, and perform better as a result in ranking trials. (Another interesting control would be to test for results when candidates sit an exam separately rather than as a group).

I won't have succeeded in hiding my preference for (3) (with, certainly, an admixture of (2)). What would really be interesting would be to study  this cross-culturally, where canons of beauty differed. Would an individual accounted good-looking through our culturally-determined eyes, but belonging to a culture where s/he is accounted ugly, do well in exams (in either culture)? (Practically, unfortunately, not an easy study to set up...)