I have been involved many times in experimentations which required human subjects. That’s recurrent for each HCI researcher. As I have been working both in the States and mainly in Europe I also noticed great differences in the practices around these experiments: while in the States there are lots of protocols in place, in Europe things are a bit more, … hum …, fuzzy!
EPFL, for instance does not have proper guidelines. Probably because our lab pioneered HCI research into this technical school with no tradition around these themes. I like the American way, in this regard, and I would love to have some more “structure” around these experiments.
To start off the discussion I think is important to have a proper Consent Form whenever we host an experiment involving humans. Each participants should be made aware of the following:
1. The purpose of the experiment;
2. The aim and procedure of the experiment;
3. The risks related to the execution of the task;
4. The time required;
5. The conditions of participation and withdrawal;
6. The benefits obtained in participating;
7. The confidentiality of the results;
8. How and where the results will be published;
9. The person/lab responsible for the experiment.
I tried to google around to find some templates but with no satisfactory results. I found an example from Harvard, one from Grinnell university and one from myself 
Here are some nice guidelines.
Tags: human computer interaction, research, research methodology
I’ve just fallen foul of the “Informed Consent” rule with our Ethics Committee (sorry, no public guidelines). I took “Free and Informed Consent” to mean that a) the participant was not forced, threatened or coerced into the experiment and b) they were clearly informed about all aspects of it. In fact, the crux was that the participant should opt-in and not opt-out — not an easy thing when you want to involve a large group of people, some visitors, etc and you want to be as unobtrusive as possible so they don’t change their behaviour. So anyway, I’m off to think of a revised experiment and rewrite my protocol. Grrr…