I've just wasted spent a few hours looking through the open source options for software that allows you to run cognitive psychology experiments. This is one of those realms where open source/free software really shines, I think.

One big reason for this is that most of the available proprietary options suck, because its never going to reach mass appeal or distribution. There are probably only a few thousand people who migh ever need software like this, so between competition and obsolete homegrown solutions that use hypercard or something. On the other hand, the faculty member reports that they "wrote their own experimental environments in QuickBasic," which is a product of the late 80s, and she went to grad school in the late 90s. Maybe she meant visual basic and misspoke, in any case, it's frightening. I'm frightened.

I think we should all just sigh at this point.

Ok, so open source is made for this kind of situation. Not only do the proprietary options lack a certian grace and flexibility, but there's a lot of polish that we don't need and on the whole we're talking about a really rather simple program. It also helps that a good percentage of "cognitive scientists" are trained a computer scientists, not psychologists, alhough the literature doesn't cross over as much as you'd think [1]: despite this project, it's not really my field, so I'm not completely "up" on the literature. In any case, computer scientists generally have the knowledge to write these kinds of programs, and because they work for universities...

Anyway so I found one of these programs called WebExp, that uses XML to define the experiments, and outputs data in a pure XML file. The only disappointing factor is that it runs in java. I'm not a java fan, because installing all the frameworks and what not has always seemed less than desirable for me. But what? I'm a fearless OS X user. I open up my terminal window, and on a lark, I type:

$ java -version

This was the response:

Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_07-164)
Java HotSpot(TM) Client VM (build 1.5.0_07-87, mixed mode, sharing)

Which is good enough for this program. I love OS X. Now all I have to do is unzip, run, and set up the XML file to design the experiment. It will even automatically counterbalance or randomly order stimuli! While I think the intention of this program is to let people do experiments at home, there's nothing to stop you from running them locally on your own system. How cool is that?

The best things in life are free.

[1]R. Gibbs' `Embodiment and Cognitive Science <http://www.cambridge.org/catalogue/catalogue.asp?isbn=0521010497>`_, is a delightful exception.