Saturday, March 29, 2008

Open Reviews?

The PLoS Computational Biology journal has an editorial about open access biology papers. Most of it is the usual Web 2.0 aspirations: We need more structured data, more semantics, more killer apps taking advantage of it.

Not that I disagree, necessarily, but I've heard it all so often that it's not really registering anymore. Maybe the killer apps will come along, and maybe they won't. Web 2.0 thinking has potential for science publications, but not everything that has potential gets realised.

But towards the end of the article, there was something that caught my attention: Open review.
Certainly rating a paper would seem reasonable when done by the Faculty of 1000 (, but it is not a generally accepted practice. We challenge you to rate this Editorial too. In some ways the reluctance to rate a scientific paper is strange since we suspect the same person may well rate a book on Another option would be to add a Digg or button ( or to incorporate conventional media ranking tools into an academic journal Web site. If one finds an interesting article, one could immediately flag it with these tools.
Now this is interesting, because peer-review is a nearly sacred notion in science. Your paper has not proven merit until it has passed the peer-reviewing process and been published. It basically says, "other scientists thought this was worth reading".

So what happens if the reviewing suddenly become open to all? Well, it essentially become a popularity contest. Digg is a perfect example: The stories that end up on the front page are the ones that a lot of people liked. Very democratic, isn't it? Only it means that today the stories included "Pranks to pull on your Co-workers" and "The 10 Most Mismatched Movie Couples".

Oh, there were plenty of interesting stories as well, but my point is that what's popular is not necessarily what's best. By all means, let people comment and review papers, but make sure we know which reviews come from scientists, and which come from your average schmoe. I know it sounds elitists, but it's not much more elitist than demanding that the person who takes out your appendix have a medical degree.

Friday, March 21, 2008

Some Reading for your Easter Weekend

I can't seem to find any interesting science stories today. (Maybe everybody's taking an extended weekend off?) So instead, I decided that it was time to round up some of the blogs that I read, but don't cite here so often.


Reading FSP makes me wish that I'd had somebody like her as a lecturer in my undergraduate years, rather than a series of boring white guys. (Not you Dr. S! Nor you Professor W!) She clearly cares about her students and loves her job, always a winning combination. Plus, I find her stories about careless misogyny in academia endlessly fascinating (in a horrifying, hope-I-never-act-like-that way).

Bioinformatics Zen

Not updated very often (kind of like this blog, huh?), but when it is, the articles are always worth reading if you're interested in the nitty-gritty of bioinformatics. Anything related to the field can come up here, whether it's the intricacies of programming, tips on how to get a PhD or humorous characterisations of stereotypical bioinformatics people.

Minor Revisions

This blog is a more recent addition to my reading schedule, but a charming one. Katie gives us a glimpse into the life of a biomedical engineering postgrad, and a very personal glimpse at that. I'm always impressed with people who are willing to share their ups and downs on their blog; it's a (skill? trait? strength?) I don't seem to possess. Katie likes having lots of subscribers to her feed, so go subscribe!

Saturday, March 15, 2008

Talking the Talk

And here's another post that I'm stealing from Of Two Minds: How to Give a Bad Science Presentation.

Of course, the advice they give applies to presentations given to fellow scientists, with the objective of introducing your work to them. And in that particular scenario, I probably agree with everything they say.

However, what if the aim of the presentation is not to inform, but to educate? In other words, what if you're giving a lecture? This is very topical for me, as I've just finished a course where students were giving presentations on papers, and I've had to do one of the presentations myself. We disregarded most of the rule they came up with. Were we right to do so? Well, let's look at the rules:

- Be able to give the presentation without support of the slides.

That one's a tricky one, because we were explaining a technique. In my part, I was heavily relying on examples to explain what was happening, and those examples were all on the slides. Could I have done it on the blackboard? Probably, but not without taking considerably more time. Still, we did rehearse a few times, so I think we could have brought the point across even without the slides. Overall, this rule holds.

- No outlines on the slides

Now this I can't completely agree with. Sure, giving an outline is slightly superfluous when you're repeating what it says on the slides. But if you're trying to get an unfamiliar topic across to an audience, reinforcement helps. During the presentations by other groups, I often found myself referring back to the slides when I hadn't caught what they were saying. I think outlines have their place in lecture slides.

- The less text the better

Two problems with this one: The first is the point that I just raised that it helps to refer back to the slide if you missed or were confused by what the speaker was saying. The second is that sometimes, the slides are made available to the audience as a study help before or after the talk. They effectively double as lecture notes, and so it is helpful if they contain enough detail so that you can understand them without the help of the speaker.

On the other hand, too much text can indeed be distracting during the presentation. So I'd advocate a compromise solution here: Keep the slides sparse, but provide detailed lecture notes at the end. Unless you're confident that your speaking ability is good enough to allow your audience to follow along easily and take notes while they do.

- Let us see the data

No argument there: Figures should be clear and big enough so that the audience can get a sense of what it is you're trying to show.

Monday, March 3, 2008

And Now For Something Completely Silly

For his inaugural post at the new Of Two Minds blog, Steve Higgins chose one the most important scientific issues of our times: Could Superman's x-ray vision really exist?

There are three basic conditions that a superman x-ray system must meet to be plausible.

1. Transparency:

The rays must be such that all objects but lead are entirely or almost entirely transparent to them. Lead is always entirely opaque to the rays.

2. Color:

The rays and processor must result in Superman perceiving the same colors as would an Earthling viewing the scene in ordinary sunlight.

3. Exclusivity:

The rays must permit Superman, but not Earthling standing in line with the reflected rays, to see through normally opaque surfaces.
Steve wrote his article tongue-in-cheek, of course, but he bases it on a very real article that appeared in 1985 (!) in the journal Perception: "On the plausibility of superman's x-ray vision" by J.B. Pittenger.

And they wonder why everybody thinks scientists are a bunch of nerds...