Cognition experiments on the web

**coglanglab** · August 12th, 2007, 10:43 PM

Hello,

I advertised some Web-based experiments here in the past. I have a new site, with new experiments: www.coglanglab.org. There are experiments dealing with language processing, with birth order effects on personality, as well as other issues.

It might be interesting to discuss the methodology. It's a controversial one, though one I'm in favor of. Just to get it started:

"Cons":
1) You don't have control over what the participants are doing.
2) You don't know if the participants are lying.
3) You don't know how representative your sample is.

Those are the ones people usually bring up.

Replies:
1) When participants come to the lab, you can't control what they are doing. I have tested many participants in the lab, and I can attest to that.
2) Ditto. People who want to lie will lie. In fact, participants online may be less likely to lie. There is more anonymity for one. For another, lab participants usually don't pay attention or lie because they get bored. A participant online can just quit if they are bored.
3) Psychology and cognition experiments don't generally require "represtentative" samples. I'm not going to get into it here, but just rest assured that if you got through a random psychology journal, you are unlikely to find a single study with a "representative" sample of participants. So if you do believe it is necessary, it's not a problem with Web-based experiments per se.

"Pros":
1) You can test large numbers of participants. I had one study with nearly 5,000 participants. I can't do that in the lab -- it's not that big! Some questions can't be asked unless you have very large numbers of participants.
2) It's cheap.
3) It's convenient for participants (no trudging off to some university).
4) It allows more people to get directly involved in science.
5) You can reach hard-to-reach populations. If I want to know if moral intutions are similar across many different cultures, in the past I would have to fly to several different countries and test people in each. Now, I can just put the experiment on the web. (The Moral Sense Test is a good example of this; use Google to find it.)

Thoughts?

------
Please try my web-based experiments at http://coglanglab.org