Bad news for advocates of traditional, book-based, brick-and-mortar library-centric research: the Web may be a better place to get your answers. A new study conducted by researchers at the University of Michigan compares the efficacy of on-line and off-line search methods, with on-line search providing answers a greater percentage of the and with 2/3 less time invested. From the study’s introduction:

Using a random sample of queries from a major search engine, we evaluate the amount of time participants spend when they use a search engine versus when they use the library without access to Web resources; any quality differences in the information found from the Web versus non-Web resources; and any differences in affective experiences between online and offline search processes.

Here is what the researchers found:

Searchers are significantly more likely to find an answer using web search. Of the 305 questions, 99.7% are answered in the web treatment, while 90.2% are answered in the non-web treatment.

Web search yields significantly more positive affective experiences than non-web search.

Online search takes significantly less time than offline search. Online search takes on average 1/3 of the time of offline search.

Searchers under the web condition looked at and used more sources than those under the non-web condition.

While non-web sources are judged to be significantly more trustworthy and authoritative than the corresponding web sources, web sources are judged to be significantly more relevant and more likely to contain enough information to answer the question. Furthermore, the accuracy of answer source is not significantly different across the two treatments. Balancing all factors, the overall source quality is not significantly different across the two venues.

I feel it necessary to point out a caveat to these conclusions, which even the researchers recognize. The research questions underpinning the study were all gleaned from the Web, with the implication that the Web would be the best place to answer Web questions. It is not certain that the same results would obtain for research questions first posed in a library setting. Nonetheless, the study certainly is sufficient to point out that for a subset of questions, the Web is definitely the place to go.

But, I still love libraries. 🙂