Research On-Line: Human Participants Ethics Issues
created by John Mueller, Educational Psychology
Background
Many areas of inquiry can benefit by extending their data collection activities to take advantage of an Internet connection. For example, surveys and similar endeavors can gain access to a cross-cultural sample, or merely save the expense and delay of regular postage or other distribution, with the added benefit that the data will be stored in a cumulative data file ready for analysis with SPSS or SAS at any time. In addition, because "the web never sleeps" the data can roll in 24 hours a day, without a research assistant or physical laboratory space. An example of such implementations can be seen by pointing your Netscape to http://www.ucalgary.ca/~mueller/tai-consent.html.
Of course, the usual considerations for ethical treatment of participants apply. By and large these are simple enough to resolve in the on-line format. These comments are based upon my own experience, with many suggestions also provided by Karen Pasveer.
Informed Consent
Informed Consent is readily obtained by staging the survey so that the first page is an information and consent form rather than the survey proper. However, this is usually concluded in a different manner than the hardcopy format. With on-line surveys, it isn't feasible to get an actual signature, so the respondent will be confronted with an on-screen button that says "I Agree", and by clicking this link the participant accepts the terms of the consent form and proceeds to the actual survey.
Although this could be seen as problematic, in fact, this format is widely used in lieu of a signature, for example, when entering credit card information on-line, or when accepting the terms of a software license prior to installation, so there are ample precedents for its use.
Another consideration with the consent form is its format. Length is your enemy: keep the consent form as concise as possible, and free of jargon and acronyms. There are several reasons for this. For example, prospective participants with a slow modem connection will have little patience for scrolling through several screens of preamble just to get started, plus with a world-wide reach some of your prospects will be reading in a second language. Keep the consent screen simple and uncluttered, as this enhances the likelihood that it will actually be read and that prospects will actually continue to become respondents rather than window shoppers.
Voluntary Participation
Voluntary Participation is still assured because prospective respondents are free to simply "surf elsewhere" if they don't wish to do complete the survey. In truth, as they confront your web page half-way around the world they would feel very little pressure to continue. Normally, participants tend to continue due to face-to-face social pressures, or the fact they have already paid to park their car, and so on, but none of those factors work in your favor with a virtual encounter. Another component of participation also applies, namely that withdrawal be without penalty (as might be possible if identity information was required). As long as their participation is anonymous this will be perceived as a credible option by the participants.
Anonymous Participation
Anonymous Participation is as important as ever. Respondents need to be informed of this component, and should not be required to provide any more personal information than is absolutely required for the research program. In general the web page software will log as header lines the IP address of the machine the respondent accessed your page from, e.g.,102.403.506.807, but otherwise only the demographic information the respondent explicitly enters will be stored.
However, it is true that some of the domains associated with an IP address are personalized, and may provide too much identifying information. Therefore it is a good idea to edit the data file periodically and erase these headers.
It is best to not ask for an e-mail address that will be stored in the main data file, because some userids are fairly specific. Participants may wish to be informed of the general results later though, so the best way to accommodate this request is to transfer them from the survey page to a "thank-you" page, and on the latter you have a place for them to enter an e-mail address that will go into a separate data file.
Data Confidentiality
Data Confidentiality is typically achieved through some restriction such that "the data will be kept in a locked file cabinet accessible only to the researcher." That refers to hard copy, which doesn't exist here. For an on-line data file, confidentiality is achieved in that the data are stored on a computer in a personal account that is accessible only to someone who knows the account userid and password, which should again be just the researcher. (In truth, cracking a Unix account password is probably much more secure than a file cabinet that yields too often to a bent paper clip!)
Feedback
Feedback is achieved by having the "submit answers" routine transfer to a closing page that acknowledges the submission, thanks the subject, and so forth. If applicable, this page should quickly inform them that no individual scores are to be provided. Participants often expect a score at the end of a questionnaire, and need to know that such is not the case so that they do not go back and resubmit. The feedback page should also provide contact information, such as an e-mail address for feedback to the researchers, as per the Consent Screen.
Other Concerns
Finally, on-line research does pose some different "ethical" considerations having little to do with the respondents. Once your webpage survey is ready to go, you need to announce it somehow. One strategy is to visit the assorted web search engines, e.g., Web Crawler, and enter the address for your page so that it can eventually be entered into that database. Some day people will come visit you as your web page begins to turn up when they search for something. Such seeding is an expected activity on the search engines, so there are no ethical problems here (other than perhaps entering your web page address just once).
As you wait for this windfall, you will want to announce it beyond the passive search engines, as in personal e-mail or posting it to a LISTSERV or USENET discussion group. You should do this so as to not be confused with the junk e-mailers or spammers, notably by judiciously announcing the project only to people/groups who would reasonably be expected to be interested. This is more a matter of "netiquette" than the usual ethics review issues. Compose a very short description of the project, compose a clear Subject Line entry (as opposed to a teasing come-on), and the ne-mail or post to appropriate groups rather than all you can find. There are several on-line sources for additional netiquette information, such as http://www.dtcc.edu/cs/rfc1855.html. The Wikipedia page for Netiquette lists several other resources.
The ultimate purpose of an Ethics Review is to protect the participants from any risk greater than that of everyday life. It is difficult to imagine the circumstances whereby an on-line survey could pose any risk, provided that the guidelines above are followed.
