The tradition of ethics in psychological and sociological research is being confronted with a new dimension in studies done using the Internet, which thwarts the conventional application of many of the principles of ethical research. The problem of exactly how modern researchers should carry out research on or involving the Internet can be subdivided into a number of problems that are fundamentally ethical in nature and those which involve technical matters. For example, obtaining informed consent from subjects is one of the fundamental tenets of research in the social sciences. However, suppose that a researcher wished to incorporate into his work a message posted to any one of the thousands of publicly accessible newsgroups. Is the researcher obligated to obtain consent from the poster? After all, researchers have traditionally been exempt from obtaining consent for data collected from public sources such as television, radio, books, and open records and spaces.
On the other hand, it has been argued (King, 1996) that some virtual communities have a high degree of perceived privacy. Despite the fact that messages are publicly available, members of the community participate with a belief that privacy will be maintained by other members, allowing for a limited public discussion of sensitive issues. The fact that the information can be perceived as both public and private is the ethical aspect of the issue. Now suppose that the researcher has concluded that the community to which the desired message was posted operates with a high level of perceived privacy, and to proceed without obtaining consent from the person who posted the message would be ethically unacceptable. By what means should the researcher inform the potential subject, and how can he or she verify the validity of the consent should it be given? Assuming that the researcher is not able to meet personally with the subject and is thus forced to obtain information through the Internet, many problems can arise. There may be difficulty in obtaining contact information, the wrong person may be contacted, the subject may be underage or not sufficiently mentally competent to give consent or the subject may falsify information (including age) in responding. The difficulty in obtaining consent ethically when the need for it has been accepted is a new technical problem that does not appear in conventional study and consequently is not dealt with in ethics procedures.
In this paper, we examine some of the problematic areas of Internet research, some ethical, some technical, and some that have both aspects, including informed consent. Another important issue is the varying degrees of public availability or group accessibility of different Internet forums. Much of the research that has been conducted regarding ethics on the Internet has mainly considered publicly accessible newsgroups and MUDs and MOOs. Many more forums exist, such as listservs, chat rooms, and bulletin boards, each with its own particular public or private character, and each with its own particular ethical concerns. For example, some forums, such as newsgroups and listservs, are archived, which means that researchers can have access to large amounts of data that are entirely naturalistic (free from experimental interference) and nearly impossible to obtain consent for.
Other issues addressed in this paper are confidentiality, data maintenance and security, how privacy and deception apply to naturalistic studies on the Internet, as well as the use of pseudonyms and personally identifiable information. We examine the ethics guidelines of the American Psychological Association and the Social Sciences and Humanities Research Council of Canada, as well as those of the researchers’ home university, concluding that these guidelines often fail to provide sufficient guidance for those conducting research on the Internet. Guidelines for off-line research simply assume that researchers have access to a certain amount of basic information about the subject: one of the biggest problems with Internet research is the difficulty of verifying subjective data. Any data related to the physical world (age, sex, location, and occupation) can be falsified by the subject with little trouble. In this sense Internet research is more akin to historical or archaeological research, except that attempting to verify subjective data can lead to further ethical complications.
Several examples of Internet research are considered, including two that have been criticized for their ethical approaches (Finn and Lavitt, 1994, and Rimm, 1995). We finish by examining some possible approaches to the suggested problems, concluding however that there are far more questions than answers. Nevertheless, the potential value of the Internet as an area for research is immense; new ways must be found to apply the essential concepts of ethical research to this important but problematic domain.