Coglanglab writes (in abbreviated form):

The New Scientist blog http://www.newscientist.com/blog/shortsharpscience/2007/03/virtual-labs-is-there-wisdom-in-crowd.html profiles a new Web-based psychology research lab http://vacognition.wjh.harvard.edu/. It seems that Internet-based research is booming, with papers published in the flagship journals Science and Nature. The vast majority of the hundreds of active Web-based studies are surveys, though a few, like the new site, are capitalizing on increased bandwidth to run real-time perception and memory studies. OK, I actually guest-blogged for New Scientist and wrote this article. But it's worth reading.

The Internet may soon change how the cognitive and social sciences are conducted, much in the way the open source movement has changed software, wikipedia has changed information gathering, etc. In the traditional lab-based model, scientists pay a small number of people to be subjects, subjecting them to a lot of testing. (By "a lot", I mean typically 1 hour, but it can last much longer.) The experiments are very carefully controlled in the lab. Now, as bandwidth has increased dramatically, it is possible to post a short (i.e., 1-5min) experiment online and get thousands of volunteers to participate, gathering just a little data from each. Of course, the controls scientists use in the lab (controlling the size of display, the timing, etc.) are much less exact. But for many experiments, the sheer amount of data more than makes up for these extra random factors. This is not only a different way of doing cognition and social science -- one that makes the process much more open and accessible for everybody -- but is also allows for new types of experiments. In the past, an experiment that involves asking subjects only one question was often hard to do because you may need very large numbers of subjects. Again, the Internet solves this problem. I'm not arguing that Web-based experiments will make the laboratory obsolete, but they are going to make huge change in the feasibility of many types of experiments and in the way in which they are conducted. That's the theory, anyway, behind my own new Web-based experiments. A few other labs are also doing on-line experiments, but we may be the first to put vision experiments on the Internet. (Many have thought vision experiments require too much control over timing, display size, etc., to be effective on the Internet. Most Web-based experiments are surveys.) Those of you who object to blatant advertising should ignore the link, but both of you who want to read more and/or participate in my study, go to http://vacognition.wjh.harvard.edu. I apologize for the blatant advertizing, but my excitement about the concept is very sincere, and I think it's something that would be of interest to slashdot readers...