Home / Topics / Research Design / Theory and Method / Web-based / Opportunities and Challenges in Internet Research
Robert Kraut

Opportunities and Challenges in Internet Research

Posted on April 26, 2006

Robert Kraut (bio) discusses the benefits of online research and some of the differences between online research and traditional research.


Q: What are some of the advantages of Internet research?
A: Internet research can reduce the costs of research, and some types of research are easier to conduct online. In particular, the costs of recruiting large samples or specialized samples lessen when done online because a researcher can get many participants from a single posting on a website or a targeted email to a specific online community. And for researchers who are interested in observing social behavior, the Internet gives them a chance to be less noticeable, and the archived communication available online can be valuable for that purpose as well. In addition, Internet research allows automation and experimental control through online administration of surveys, which reduces expenses and time spent in administering measures and entering data.
Q: What are the concerns about data quality and generalizability with Internet research?
A: The concerns can be divided into two main areas: sample biases and control over the data-collection setting. We have no sampling frame to provide a random sample of Internet users, so generalizing to a larger population is difficult in Internet research. Bias is also a problem due to the self-selection nature of online surveys and non-response due to dropout during longitudinal studies. Maintaining contact with participants in Internet research can be more difficult because email addresses are more likely than phone numbers or mailing addresses to change over time.

By conducting research online, the researcher gives up some control of the research environment. The researcher has a much harder time verifying subjects' identities online and has more difficulty monitoring and/or intervening with the participants. Multiple submissions from participants can be a concern, but some of that can be prevented by assigning unique identifiers to invited participants and tracking Internet protocol (IP) addresses. Also, researchers should examine incoming data for inconsistent data patterns.
Q: What are the challenges of Internet research with regard to the protection of human subjects?
A: Some types of online research are not considered human subjects research with respect to federal regulations. It is not human subjects research if the research only involves the collection of pre-existing data that contains no identifiable private information (information that identifies an individual and for which the individual had reasonable expectation that no observation was taking place or that information collected for a specific purpose would remain private.)

Assessing whether the research is human subjects research or what types of regulations apply requires the researcher to clarify some basic terms such as anonymous, public behavior, and preexisting data for use in online research.

First, researchers need to address what makes a person identifiable or anonymous. Anonymity is not so complicated an issue with online surveys, but it becomes a more ambiguous concept when one wants to observe naturally occurring behavior online. Researchers cannot assume that pseudonyms make conversations anonymous because pseudonyms may contain real names or identifying information, and quoted sections can often be linked back to the original poster through search engines. To truly keep participants anonymous, researchers need to alter pseudonyms and any quoted text.

Second, researchers have to determine when online behavior can be regarded as public behavior. Private communication such as emails or instant messaging between individuals is protected communication. Online forums can usually be treated as public areas, but there are many factors to consider when determining the status of online communications. What are the applicable legal regulations? What are the social norms of the group? Is membership restricted or open? Is the membership relatively static, or is there frequent turnover? Are conversations archived? Does the forum have explicit recording policies?

Third, researchers need to clearly define preexisting public data because research involving the collection of that data is exempt from federal human-subjects regulations. To be classified as preexisting, all of the data must exist prior to the beginning of the research, so entries in an ongoing online journal or live discussion group cannot be considered preexisting.
Q: What are potential sources of risk for participants in Internet research?
A: Two key categories of potential risk are harm directly resulting from participating in research and harm resulting from confidentiality breaches. Most Internet research involves minimal risk to the participants. The potential risks from participation come primarily from the researcher's decreased ability to monitor participants and to intervene if necessary. Another factor to consider is the welfare of the online community, not just the individual participants.

The more likely source for risk in Internet research involves breaches of confidentiality. Data in transit such as email between a participant and researcher can be vulnerable, and data also needs to be stored securely. Identifying information needs to be stored in separate files away from the data, and researchers should protect directories containing data so that only researchers and their staff have access.

Another critical issue involves vulnerable populations, especially children. In Internet research, it is difficult to identify whether a potential participant is a child. If you want to include only adults in your study, there are certain actions you can take to discourage children from participating. For example, you can create displays that do not appeal to children, advertise in sites that children do not frequent, ask participants for their date of birth, or ask them for some identifying information that children are unlikely to have.



Based on published article and personal communication with the researcher in March 2006. Kraut, R. E., Olson, J., Manaji, M., Bruckman, A., Cohen, J., & Couper, M. (2003). Psychological research online: Opportunities and challenges. American Psychologist, 59(2), 105-117.


Please note that the feedback is viewed only by 4researchers staff and is not intended for communication with individual contributors.


Use the form below to submit feedback about this article. If you would like a response, please be sure to include your e-mail address.

More About "Web-based"


Benefits of Web-Based Data Collection

Designing Research Studies for eHealth Programs

Show All...