Yes we coudl do an experimentation very important. I am one of the author of the "Elementary Pragmatic Model. In the next number of " Il Giornale Italiano di Psichiatria" will be publish a paper (in English) that shows that the core of the mind is sintetized in three states. The "void"(F0), The "caos"(F15) ant the "maintenance of his own world"(F3). We have discovered that while the normal people shows high F3 and low F0 and F15, people with mental disorders show less F3 and more F0 and F15: Then the main road for cure mental disorders is increase F3. We could ask to people in the net: 1) what in their opinioni increase the mainainance of his own world, 2) Gater the opiions,3) Select of them the more applicable ,4) Gather the people thar applicate them and ask if born a positive change. Publish on line the results ( for instance in Research Gate ) or in a Journal on line, or in a printed Journal.

Yours

Piero De Giacomo

www.pierodegiacomo.it.

I recommend a good review of the methodological peculiarities of this type of research:

* Birnbaum MH. Introduction To Psychological Experiments On The Internet

* Kraut et al. (2004). Psychological Research Online. American Psychologist

* Tom Buchanan* and John L. Smith (1999). Using the Internet for psychological research: Personality testing on the World Wide Web. British Journal of Psychology

* Brian A. Nosek, Mahzarin R. Banaji, Anthony G. Greenwald (2002). E-Research: Ethics, Security, Design, and Control in Psychological Research on the Internet. Journal of Social Issues, Vol. 58, No. 1, 2002, pp. 161-176

you may find some thoughts and suggestions in the following paper:

'Internet use and self-rated health among older people: a national survey'

available at:

http://www.jmir.org/2009/4/e49/

Thanks, Manuel, I know those publications, they are all classics, especially Tom's and Michael's. Enrique, thanks for the pointer - this one is new to me. What suggestion in particular did you find helpful in this article?

For example, you have to be careful with confounding variables that may lead you to the wrong conclusions, as well as the adequate planning of research to avoid methodological biases...I recommend your own reading, so you can get the information more adequate for your concerns or needs

best

Keep it ethical, don't left it be a burden to others, don't let it result in spam, make sure the result of the experiment can benefit others then yourself.

I'd generally go with those for of the top of my head dos and don'ts.

As the issues that you, Enrique, mentioned are so important, should they be applied not only to Internet-based experimenting, but to any research?

Of course, as humans we do make mistakes and it is not always easy to maintain standards, but avoiding those pitfalls should be part of our job. (Only my opinion. I encourage everybody to add different perspectives.)

1. Incentives should always be offered. I have offered results to my surveys since I have conducted it as a student. I provided a space in the end that if anyone is interested in getting an analysis done of the responses on questionnaires then please mention the e-mail ids. All of them have provided their e-mails. Getting results and analysis makes a good motivator for many participants. But response rate was very low. Many of them left the procedure in between. May be because the procedure was a bit long, consisting of 5 questionnaires. I do feel that some lucky-draw kind of incentive must be there so that more responses can be expected. In my next study I am going to offer some financial benefit to participants. Since my sample would be of students so I believe that many of them would like to have a try on their luck.

2. There will always be people who are comfortable with long procedures, but since examiner is not present with them so many of them may not understand that they can take breaks and therefore many would prefer to complete them in one go. This can result in poor response quality and may effect the data. Thus, it can be mentioned in the beginning that participants must take at least one break if the procedure is long. That definitely depends on the kind of procedure to be followed. In performance tasks we can not usually ask to take breaks in between. But in survey questionnaires this practice should be encouraged.

A sort of 'Intermission' message should be displayed on screen like in a movie theater. And information about this should be provided in the beginning too so they can expect the procedure to be interval based, so that they may not feel exhausted and leave in between. The 'filler activity' can also be provided to make them feel light headed. In studies where emotional responses are expected, this practice can be encouraged.

A big "do" I learned recently is to utilize social media as a site for recruitment. When we sent to an organizational listserv of a "hard to reach" population we got a very homogeneous response in many demographic categories (a lot of middle aged white men). However, when we utilize facebook posting and craigslist postings, we got a much broader array of responses and a much more diverse sample. Since we are seeing stratification in which groups utilize which types of technology, I learned that I needed to develop a clearer strategy for reaching the groups I wanted by having a more developed plan of recruitment based on locaiton of electronic recruitment. Just my two cents.