Analyzing the Web: Directions and Challenges

Past Research

The increasing popularity of the Internet and WWW has naturally attracted the attention of a vast body of researchers, ranging from philosophers to technocrats. In general, WWW research has progressed in two directions. First, there has been an attempt to examine the people who use the WWW. These “rating” studies have been empirical analyses of the extent of WWW use and the assessment of the behavior and opinions of the WWW users. Many of the major American polling companies such as Gallup and Nielson have entered into WWW ratings research. Given the dynamics of WWW use, with its rapid changes and the ongoing addition of innumerable users, any analysis of WWW usage yields conflicting results. Sometimes, however, such conflicting results are also the product of inadequate and untested (thus unreliable) methods of counting WWW users and obtaining data about their opinions and behaviors. Thus, researchers are often quick to point out that their results could be “dated” or slanted, depending on their methodology. One such example is the CommerceNet and Nielsen study that is updated frequently (see the information available at the address http://www.commerce.net/nielsen/press_97.html and at sites such as http://www.interlog.comEbxi/size.htm). In this case, there is an attempt to provide frequently updated counts of Internet users while recognizing that even the most carefully planned research can miss sections of users as they constantly move in and out of cyberspace.

The second kind of analysis has focused on the text exchanged by the users of the WWW. There is a presumption in this analysis that the WWW and the Internet can be considered a mass medium (Newhagen & Rafaeli, 1996). This perspective on the Internet, which legitimizes the analysis of the Internet's content, is supported by Morris and Ogan (1996), who contend that “the Internet is a multifaceted mass medium, that is, it contains many different configurations of communication” (p. 42). Clearly, the analysis of the message/text/discourse [Page 181]content has been more popular with exchanges observed in newsgroups.1 However, popular recognition of the “mass media” appeal of the WWW turns attention to the WWW text. This interest leads researchers to examine questions about the content of the WWW.2 In fact, debates around the censoring of the WWW text begin with the presumption that textual content can have specific effects on its audiences. However, how the effect might manifest and how the text can be analyzed has not been explored in great detail.

...