Magazine article Information Today

Some Advice for Google Answers

Magazine article Information Today

Some Advice for Google Answers

Article excerpt

Google. Well, well, well. So you've decided to answer questions directly, to come out from behind the protective cover of a "resource" and face the challenge of talking to requesters by asking them what they want to know and telling them what you think the answer is and why. And you're not alone. Besides existing "Ask a" or experts-on-call services like AskJeeves, it appears that the announcement of your new Google Answers service (http://answers.google.com) has inspired other monster engines to adopt, accelerate, or even revive their own "Ask a" offerings-such as Yahoo! Advice, which is conducted in partnership with LiveAdvice (http://advice.yahoo.com). In other words, you and your competitors have decided to go into the reference librarian business.

Welcome to our world. You will probably have considerable success, at least in terms of the quantity of questions. After all, the Google search engine service has handled over 100 million requests a day for years now. According to some library statistics, you answer in a day and a half more questions than all the libraries in the country tally in a year. You have always offered the 2417 service that only a handful of libraries have just begun to offer. Clearly, when it comes to marketing an answer service, you sit in the catbird seat. You already loom large in the sights of all the millions of users behind those 100 million-plus daily questions.

You may get a lot of questions all right, but now you also have to get all the answers right. Don't get me wrong, I didn't mean that sarcastically. You get lots of answers right all the time. Otherwise people wouldn't turn to you so often. But a search engine interaction places the burden of proof on the asker of the question. The asker plays on your turf. You set the rules. You define the game. If searchers don't get the information they want, they may go away thinking that it doesn't exist or can't be found, since you apparently couldn't find it. Or they may think they asked the question incorrectly. But when you take questions from users in conversational form with the potential for follow-up to clarify any confusion, you're going to allow for interactive dialogues, of course. Then it becomes a reference interview. The burden now falls on you to answer the question. The requesters feel that they have explained themselves sufficiently. In allowing them to ask their questions fully, in conversational prose, you have implicitly assured them that you understand what they want and will try to find it to the best of your abilities.

How do I know all this? Because people like me-reference librarians, professional searchers, information professionals, or whatever one calls us-have been busy finding answers on behalf of needy clients for years. We're a nervous lot, liable to jump when cars backfire. Taking on the challenge of answering any question that can pop up in the mind of any human would make anyone gun-shy. Mainly it makes us insecure and inclined to doubt our performance. For example, for several decades a series of studies claimed that our success rate for answering questions ran at around 55 percent. This rule-of-thumb standard-- called the "55-percent rule" or sometimes "half-right reference"-served as a depressing reminder of the fallibility of even trained professionals.

A recent study, which was conducted on a broader scale than previous ones and focused on more real-world reference requests, shows that our accuracy actually reaches around 90 percent. (For a quick overview, read John V. Richardson Jr.'s article, "Reference Is Better Than We Thought," in Library Journal, April 15, 2002, pp. 41-42. For the full version, turn to Richardson and Matthew Sexton's Understanding Reference Transactions: Transforming an Art into a Science, Academic Press, 2002. For a list of reference studies over the years, start at http://url.org/net/references. Pack a lunch. The literature begins with a seminal study that was conducted in 1876. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.