Psychological studies implementing the World-Wide Web as a data collection medium have traditionally been constrained to data obtained from surveys or from within-subjects comparisons. Common gateway interface (CGI)scripts, however, provide researchers with a means of collecting data for between-subjects comparisons. This paper provides a description of the information flow between the user (subject) and the experimenter by referencing a completed between-subjects assignment CGI. A current online cognitive psychology experiment serves as an example of the utility of CGIs as subject assignment mechanisms. Server-side and client-side scripts are compared for effectiveness at data collection, with these types of designs in mind.The ubiquity of the World-Wide Web (WWW) has served to increase the rate of data collection for many psychologists. A testament to this observation is the increase in both degree and kind of data collection devices currently found on the Internet. Whereas survey forms (Schmidt, 1997) and within-subjects research methods implementing static hypertext markup language (HTML) pages have become a quotidian staple in the experimental psychologist's research toolbox, methods involving between-subjects designs have been less recognizable as Internet-based experimental designs.Central to the implementation of a between-subjects experimental design on the Internet is a program that is external both to the user/subject and to the experimenter. A method is necessary that instructs the subject about the experimental requirements, as well as provides accurate, consistent data reporting to the experimenter. Server-side scripts that are written as common gateway interfaces (CGIs) can serve this function. Server-Side and Client-Side ScriptsThe key differences between a server-side and a clientside script are not so much a matter ofwhat is being done as ofwhere it is being done. Although we may think ofeach as containing the instructions for an experimental protocol, there exist key differences between the two. Clientside scripts occur on the machine receiving feedback from the server. As a matter offunction, client-side scripts are shipped in their entirety to the computer (more specifically, the browser) that is downloading the experiment.Correspondence concerning this article should be addressed to R. H, Morrow, U.S.M.P.O. Box 5934, Hattiesburg, MS 39406 (e-mail: rmorrowrgzocean.otr.usm.edu). The client's browser then builds the experiment each time it is requested from the server. Results of the experiment are then shipped back to the server. The scripted actions ofthe client-side script take place entirely on the subject's/ client's machine. As a methodology for experiments, this saves bandwidth and reduces confounds that are due to temporal insensitivity (i.e., geographical proximity to the experimenter's Web server). Server-side scripts, on the other hand, perform all actions on some web server remote from the user, after interacting with the user's computer. As such, the scripts are independent of wha...

No abstract

No abstract

No abstract

No abstract