WHY THE WEB WON�T REPLACE TRADITIONAL SOURCES ANYTIME SOON

James H. Sweetland, University of Wisconsin, Milwaukee, North American Convenor

A recent study reported in Reference and User Services Quarterly inspires these thoughts.

Connell and Tipple evaluated the quality of Web site content by a very simple and elegant test. They collected all the ready reference questions received by a public library reference desk for a week, eliminating library-specific questions, and verifying answers to the 60 remaining questions in two different printed reference sources.

Then, an experienced user of Alta Vista performed searches on the World Wide Web for each of these questions. This often required revising a strategy when the first did not yield any useful results, but the data analysis was conducted only on a strategy which produced at least one answer to the question.

For each of the sixty questions, the first two screens (totaling up to 20 web sites) were used to access all Web sites listed. The results are most interesting, if not alarming:

1160 sites were found in answer to the 60 questions.
144 of the cited sites were dead links
241 of the cited sites were duplicates (e.g. mirror sites, a different page from the same site)

Of the 1,010 sites which were, in fact, accessible (eliminating the duplicates).

160 (15.8% of the live sites) sites provided the correct answer;
115 (11.4%) provided a partially correct answer (such as the correct local phone exchange, but not the area or country code);
89 (8.8%) of the sites provided incorrect information;
646 (64%) of the sites did not provide any information to answer the question.

Or, given a trained librarian with experience in searching and familiar with the ins and outs of a given search engine, trying to find rather straight forward information, about 2/3 of the information retrieved was completely useless, and another 9% was worse than useless.

Another way of looking at the Web as an information source appears in the statistics collected by the OCLC (Online Computer Library Center) Office of Research . As of June 1999, they found that % of the Web sites (actually, the IP addresses for the sites) sampled a year earlier were no longer active. Compare this figure with the speed with which books go out of print, and consider that even when that happens, libraries will still keep the o.p. book, and retain a record for that book in their catalogs.

A third statistic is relevant here as well. Choice magazine, a U. S. reviewing source for academic library-oriented material, issued a Supplement of recommended Web sites, complete with reviews, in 1997 and 1998. All sites in the 1998 issue were checked in June of 1998. When I checked a sample of these recommended sites in September (three months later) four of them were nonexistent entirely, and four others had changed addresses (two of these had also changed titles).

The conclusion? At the present time, contrary to the impression one gets from the endless advertising and publicity about the World Wide Web, it will not replace printed material, CD-ROMs and, for that matter, what are now being called "traditional databases". It�s hard to find good content, when you do, it tends to disappear (or at least move). And, more frightening, it�s easy to get wholly incorrect information.

As commercial advertising and excited news reports continue to tell people that "print is dead" and that "you can get everything you need on the Web"; as a growing chorus of voices says that libraries are obsolete, and that librarians themselves are obsolete, the above needs to be remembered. Ease of access and amount of information available have little meaning if the information so readily obtained is wrong. A very old computer term continues to have meaning here: GIGO (Garbage in, Garbage out). Too many people, including too many who should know better, are assuming that computer-based information, readily mounted by anyone who wants to do so, must be good. It isn�t.

And, it would appear that the last bastion of quality control, often avoided by the Web surfer, is the librarian (or, if you prefer, the information professional, knowledge manager, search intermediary, etc.). It�s past time when everyone who works with information started a campaign to counteract the hype having good information is good; having bad information is often worse than having none at all.

Back to Electronic Resources Viewpoints