In some movies, a visitor arrives in a foreign country and meets someone who offers to be his personal guide. The tourist doesn't know his way around, so he could use the help, but can he trust the guide? At this point, depending on what sort of movie it is, one of two things happens: hilarity ensues, or the tourist meets a grisly end.
The programmers who build search engines recognize that most people who use their tools don't understand how they work, and so they try to find ways of helping them. Sometimes they succeed — but sometimes that help can steer you in exactly the wrong direction. Here are some things you need to know if you don't want your research to meet a grisly end (or even just a laughable one) on the internet.
First, search engines may alter your search query in an attempt to help you find what you're looking for. They may correct obvious misspellings (“reading onlin” will become “reading online”). But they may also “suggest” words that are more common or return more results than the one you're actually looking for — in a way that completely changes the meaning of your query! Typically you'll see a message at the top of the page, asking “Did you mean...?” or noting that your query was changed — but it's easy to overlook, so be careful.
Search engines may also change the way they work — at any time, without notice. For example, Google used to require every word in your query, but it no longer does for all searches. Even when you're used to a particular tool and know how it works, keep paying attention.
Google, in particular, also filters your search results based on everything it knows about you — based, that is, on all the data it has collected about you, which can be quite a lot. Millions of websites use Google Analytics to track their visitors; if you've visited any of them, then by placing a cookie in your web browser, Google is able to track all the websites you visit. Google also tracks what you search for and which results you click on. And if you're logged in with a Google account — to Google Docs, Gmail, or Google Plus — then it has even more data about you. In 2011, Google began using that data to display different sets of results to different users, to direct them to the web pages they were most likely looking for. As a result, two people searching Google in exactly the same way could get very different results. The biggest public concern with this practice has been the filtering of news articles — people who tend to click on liberal websites might wind up seeing primarily liberal-biased news results, while conservatives could see the opposite — but it can cause problems with research as well.
In response to public objections about this practice, Google now offers a Verbatim search: see http://support.google.com/websearch/bin/answer.py?hl=en&answer=1734130. This turns off all personalization, but it also turns off spelling corrections, synonym finding, and so on. If you're searching for information on something politically controversial (global warming, perhaps?) you might want to try the Verbatim search and see how your results differ.
It's important to realize that because all these query rewrites and search filters are based on what people actually click on after searching, they're probably quite successful in getting people to what they want. (If they're not, the engineers will see that people aren't clicking on the top search results, and they'll refine the algorithm.) The problem this poses for classroom research or professional development is that most web searches are not for serious, objective, scholarly research! Web search tools are designed to give people what most people think they want. Consider how that differs from the job of a reference librarian, who judges based on professional knowledge and experience what a patron actually needs! It's tempting to see Google as a kind of reference librarian, but it isn't guiding you: it's only reinforcing your own tendencies and choices (and those of searchers collectively). And it can be very difficult to break out of that loop.
Eli Pariser's TED talk on the “filter bubble”: http://www.ted.com/talks/eli_pariser_beware_online_filter_bubbles.html – this is CC]
Most of us use a favorite search engine all the time — or pick a new
one at random. You may want to take the time to learn more about how
the four biggest search engines work. Some provide more information
than others.
links to how each of the four major search engines work (Google, Bing, Yahoo, Ask)